[tvm-site] branch asf-site updated: deploying docs (apache/tvm@bf4fedd06e2e60ee4a947740738949e8acf64400)

[tq...@apache.org]

This is an automated email from the ASF dual-hosted git repository.

tqchen pushed a commit to branch asf-site
in repository https://gitbox.apache.org/repos/asf/tvm-site.git


The following commit(s) were added to refs/heads/asf-site by this push:
     new 5ca0bb6d72 deploying docs (apache/tvm@bf4fedd06e2e60ee4a947740738949e8acf64400)
5ca0bb6d72 is described below

commit 5ca0bb6d7222d557b011f5276b187aeb1e312c37
Author: tvm-bot <95...@users.noreply.github.com>
AuthorDate: Wed Jan 11 02:02:17 2023 +0000

    deploying docs (apache/tvm@bf4fedd06e2e60ee4a947740738949e8acf64400)
---
 docs/_images/sphx_glr_micro_train_001.png          |  Bin 302170 -> 309078 bytes
 docs/_images/sphx_glr_micro_train_thumb.png        |  Bin 22480 -> 22719 bytes
 .../how_to/compile_models/from_darknet.rst.txt     |    2 +-
 .../how_to/compile_models/from_keras.rst.txt       |    2 +-
 .../how_to/compile_models/from_mxnet.rst.txt       |    2 +-
 .../how_to/compile_models/from_oneflow.rst.txt     |    2 +-
 .../how_to/compile_models/from_pytorch.rst.txt     |    2 +-
 .../how_to/compile_models/from_tensorflow.rst.txt  |    2 +-
 .../compile_models/sg_execution_times.rst.txt      |   22 +-
 .../deploy_models/deploy_model_on_adreno.rst.txt   |    2 +-
 .../deploy_models/deploy_model_on_android.rst.txt  |    2 +-
 .../deploy_object_detection_pytorch.rst.txt        |    4 +-
 .../deploy_models/deploy_prequantized.rst.txt      |    6 +-
 .../deploy_prequantized_tflite.rst.txt             |    4 +-
 .../how_to/deploy_models/deploy_quantized.rst.txt  |    2 +-
 .../deploy_models/deploy_ssd_gluoncv.rst.txt       |    4 +-
 .../deploy_models/sg_execution_times.rst.txt       |   20 +-
 .../extend_tvm/bring_your_own_datatypes.rst.txt    |    2 +-
 .../how_to/extend_tvm/sg_execution_times.rst.txt   |   10 +-
 .../how_to/extend_tvm/use_pass_instrument.rst.txt  |   16 +-
 .../optimize_operators/opt_conv_cuda.rst.txt       |    2 +-
 .../optimize_operators/opt_conv_tensorcore.rst.txt |    2 +-
 .../how_to/optimize_operators/opt_gemm.rst.txt     |   16 +-
 .../optimize_operators/sg_execution_times.rst.txt  |    8 +-
 .../sg_execution_times.rst.txt                     |   14 +-
 .../tune_conv2d_layer_cuda.rst.txt                 | 3405 +++++++++++++++-----
 .../tune_network_cuda.rst.txt                      |    4 +-
 .../tune_network_x86.rst.txt                       |    4 +-
 .../tune_sparse_x86.rst.txt                        |   86 +-
 .../tune_with_autotvm/sg_execution_times.rst.txt   |    6 +-
 .../tune_with_autotvm/tune_conv2d_cuda.rst.txt     |  256 +-
 .../work_with_microtvm/micro_autotune.rst.txt      |   16 +-
 .../work_with_microtvm/micro_pytorch.rst.txt       |    4 +-
 .../how_to/work_with_microtvm/micro_train.rst.txt  |   18 +-
 .../work_with_microtvm/sg_execution_times.rst.txt  |   12 +-
 .../work_with_relay/sg_execution_times.rst.txt     |    8 +-
 .../how_to/work_with_schedules/intrin_math.rst.txt |    2 +-
 .../work_with_schedules/sg_execution_times.rst.txt |   14 +-
 .../how_to/work_with_schedules/tensorize.rst.txt   |    2 +-
 .../tutorials/autotvm/sg_execution_times.rst.txt   |    4 +-
 .../frontend/deploy_classification.rst.txt         |    2 +-
 .../tutorials/frontend/deploy_detection.rst.txt    |    2 +-
 .../tutorials/frontend/sg_execution_times.rst.txt  |    6 +-
 .../tutorials/optimize/sg_execution_times.rst.txt  |    6 +-
 .../topic/vta/tutorials/sg_execution_times.rst.txt |    6 +-
 .../tutorial/auto_scheduler_matmul_x86.rst.txt     |    6 +-
 docs/_sources/tutorial/autotvm_matmul_x86.rst.txt  |   20 +-
 docs/_sources/tutorial/autotvm_relay_x86.rst.txt   |   57 +-
 .../tutorial/cross_compilation_and_rpc.rst.txt     |    2 +-
 docs/_sources/tutorial/intro_topi.rst.txt          |    2 +-
 docs/_sources/tutorial/sg_execution_times.rst.txt  |   20 +-
 .../tutorial/tensor_expr_get_started.rst.txt       |   45 +-
 docs/commit_hash                                   |    2 +-
 docs/how_to/compile_models/from_darknet.html       |    2 +-
 docs/how_to/compile_models/from_keras.html         |    2 +-
 docs/how_to/compile_models/from_mxnet.html         |    2 +-
 docs/how_to/compile_models/from_oneflow.html       |   11 +-
 docs/how_to/compile_models/from_pytorch.html       |    9 +-
 docs/how_to/compile_models/from_tensorflow.html    |    2 +-
 docs/how_to/compile_models/sg_execution_times.html |   22 +-
 .../deploy_models/deploy_model_on_adreno.html      |    2 +-
 .../deploy_models/deploy_model_on_android.html     |    2 +-
 .../deploy_object_detection_pytorch.html           |   40 +-
 docs/how_to/deploy_models/deploy_prequantized.html |    8 +-
 .../deploy_models/deploy_prequantized_tflite.html  |    4 +-
 docs/how_to/deploy_models/deploy_quantized.html    |    2 +-
 docs/how_to/deploy_models/deploy_ssd_gluoncv.html  |   34 +-
 docs/how_to/deploy_models/sg_execution_times.html  |   20 +-
 .../extend_tvm/bring_your_own_datatypes.html       |    2 +-
 docs/how_to/extend_tvm/sg_execution_times.html     |   10 +-
 docs/how_to/extend_tvm/use_pass_instrument.html    |   16 +-
 docs/how_to/optimize_operators/opt_conv_cuda.html  |    2 +-
 .../optimize_operators/opt_conv_tensorcore.html    |    2 +-
 docs/how_to/optimize_operators/opt_gemm.html       |   16 +-
 .../optimize_operators/sg_execution_times.html     |    8 +-
 .../sg_execution_times.html                        |   14 +-
 .../tune_conv2d_layer_cuda.html                    | 3401 ++++++++++++++-----
 .../tune_with_autoscheduler/tune_network_cuda.html |    4 +-
 .../tune_with_autoscheduler/tune_network_x86.html  |    4 +-
 .../tune_with_autoscheduler/tune_sparse_x86.html   |   86 +-
 .../tune_with_autotvm/sg_execution_times.html      |    6 +-
 .../how_to/tune_with_autotvm/tune_conv2d_cuda.html |  256 +-
 docs/how_to/work_with_microtvm/micro_autotune.html |   16 +-
 docs/how_to/work_with_microtvm/micro_pytorch.html  |    5 +-
 docs/how_to/work_with_microtvm/micro_train.html    |   16 +-
 .../work_with_microtvm/sg_execution_times.html     |   12 +-
 .../how_to/work_with_relay/sg_execution_times.html |    8 +-
 docs/how_to/work_with_schedules/intrin_math.html   |    2 +-
 .../work_with_schedules/sg_execution_times.html    |   14 +-
 docs/how_to/work_with_schedules/tensorize.html     |    2 +-
 docs/install/nnpack.html                           |   12 +-
 docs/reference/api/python/auto_scheduler.html      |    4 +-
 .../api/typedoc/classes/bytestreamreader.html      |   12 +-
 .../api/typedoc/classes/cachedcallstack.html       |   34 +-
 docs/reference/api/typedoc/classes/dldatatype.html |   12 +-
 docs/reference/api/typedoc/classes/dldevice.html   |   10 +-
 .../reference/api/typedoc/classes/environment.html |   12 +-
 docs/reference/api/typedoc/classes/ffilibrary.html |   20 +-
 .../api/typedoc/classes/graphexecutor.html         |   16 +-
 docs/reference/api/typedoc/classes/instance.html   |   40 +-
 docs/reference/api/typedoc/classes/memory.html     |   34 +-
 docs/reference/api/typedoc/classes/module.html     |   10 +-
 docs/reference/api/typedoc/classes/ndarray.html    |   22 +-
 .../api/typedoc/classes/packedfunccell.html        |    6 +-
 docs/reference/api/typedoc/classes/rpcserver.html  |   14 +-
 docs/reference/api/typedoc/classes/scalar.html     |    6 +-
 .../api/typedoc/classes/webgpucontext.html         |   12 +-
 docs/reference/api/typedoc/enums/argtypecode.html  |   30 +-
 .../api/typedoc/enums/aynccallbackcode.html        |    4 +-
 .../api/typedoc/enums/dldatatypecode.html          |    8 +-
 .../api/typedoc/enums/rpcserverstate.html          |   12 +-
 docs/reference/api/typedoc/enums/sizeof.html       |   18 +-
 docs/reference/api/typedoc/index.html              |  112 +-
 .../api/typedoc/interfaces/disposable.html         |    2 +-
 .../api/typedoc/interfaces/functioninfo.html       |    6 +-
 .../api/typedoc/interfaces/libraryprovider.html    |    4 +-
 docs/searchindex.js                                |    2 +-
 .../vta/tutorials/autotvm/sg_execution_times.html  |    4 +-
 .../tutorials/frontend/deploy_classification.html  |    2 +-
 .../vta/tutorials/frontend/deploy_detection.html   |    2 +-
 .../vta/tutorials/frontend/sg_execution_times.html |   10 +-
 .../vta/tutorials/optimize/sg_execution_times.html |    6 +-
 docs/topic/vta/tutorials/sg_execution_times.html   |    6 +-
 docs/tutorial/auto_scheduler_matmul_x86.html       |    4 +-
 docs/tutorial/autotvm_matmul_x86.html              |   20 +-
 docs/tutorial/autotvm_relay_x86.html               |  265 +-
 docs/tutorial/cross_compilation_and_rpc.html       |    2 +-
 docs/tutorial/intro_topi.html                      |    2 +-
 docs/tutorial/sg_execution_times.html              |   20 +-
 docs/tutorial/tensor_expr_get_started.html         |   41 +-
 130 files changed, 5960 insertions(+), 3127 deletions(-)

diff --git a/docs/_images/sphx_glr_micro_train_001.png b/docs/_images/sphx_glr_micro_train_001.png
index 2deab86823..c3da230c00 100644
Binary files a/docs/_images/sphx_glr_micro_train_001.png and b/docs/_images/sphx_glr_micro_train_001.png differ
diff --git a/docs/_images/sphx_glr_micro_train_thumb.png b/docs/_images/sphx_glr_micro_train_thumb.png
index 9d8a85810f..4541a1fa7e 100644
Binary files a/docs/_images/sphx_glr_micro_train_thumb.png and b/docs/_images/sphx_glr_micro_train_thumb.png differ
diff --git a/docs/_sources/how_to/compile_models/from_darknet.rst.txt b/docs/_sources/how_to/compile_models/from_darknet.rst.txt
index 6771b6b709..8634f59c02 100644
--- a/docs/_sources/how_to/compile_models/from_darknet.rst.txt
+++ b/docs/_sources/how_to/compile_models/from_darknet.rst.txt
@@ -319,7 +319,7 @@ The process is no different from other examples.
 
 .. rst-class:: sphx-glr-timing
 
-   **Total running time of the script:** ( 1 minutes  10.448 seconds)
+   **Total running time of the script:** ( 1 minutes  9.859 seconds)
 
 
 .. _sphx_glr_download_how_to_compile_models_from_darknet.py:
diff --git a/docs/_sources/how_to/compile_models/from_keras.rst.txt b/docs/_sources/how_to/compile_models/from_keras.rst.txt
index 5c95612143..85bc112709 100644
--- a/docs/_sources/how_to/compile_models/from_keras.rst.txt
+++ b/docs/_sources/how_to/compile_models/from_keras.rst.txt
@@ -232,7 +232,7 @@ Look up prediction top 1 index in 1000 class synset.
  .. code-block:: none
 
     Relay top-1 id: 285, class name: Egyptian cat
-
    1/1 [==============================] - ETA: 0s
    1/1 [==============================] - 1s 957ms/step
+
    1/1 [==============================] - ETA: 0s
    1/1 [==============================] - 1s 923ms/step
     Keras top-1 id: 285, class name: Egyptian cat
 
 
diff --git a/docs/_sources/how_to/compile_models/from_mxnet.rst.txt b/docs/_sources/how_to/compile_models/from_mxnet.rst.txt
index 71209611fa..686602d704 100644
--- a/docs/_sources/how_to/compile_models/from_mxnet.rst.txt
+++ b/docs/_sources/how_to/compile_models/from_mxnet.rst.txt
@@ -116,7 +116,7 @@ In this section, we download a pretrained imagenet model and classify an image.
 
  .. code-block:: none
 
-    Downloading /workspace/.mxnet/models/resnet18_v1-a0666292.zipf8026450-0129-4148-8d3d-7308ae4026b4 from https://apache-mxnet.s3-accelerate.dualstack.amazonaws.com/gluon/models/resnet18_v1-a0666292.zip...
+    Downloading /workspace/.mxnet/models/resnet18_v1-a0666292.zip7626fc43-7722-4bb2-b4c4-e3a0d2d4c363 from https://apache-mxnet.s3-accelerate.dualstack.amazonaws.com/gluon/models/resnet18_v1-a0666292.zip...
     x (1, 3, 224, 224)
 
 
diff --git a/docs/_sources/how_to/compile_models/from_oneflow.rst.txt b/docs/_sources/how_to/compile_models/from_oneflow.rst.txt
index 9bd9d5b859..582d3f2cdd 100644
--- a/docs/_sources/how_to/compile_models/from_oneflow.rst.txt
+++ b/docs/_sources/how_to/compile_models/from_oneflow.rst.txt
@@ -121,7 +121,7 @@ Load a pretrained OneFlow model and save model
  .. code-block:: none
 
     Downloading: "https://oneflow-public.oss-cn-beijing.aliyuncs.com/model_zoo/flowvision/classification/ResNet/resnet18.zip" to /workspace/.oneflow/flowvision_cache/resnet18.zip
-
      0%|          | 0.00/41.5M [00:00<?, ?B/s]
     21%|##        | 8.59M/41.5M [00:00<00:00, 90.0MB/s]
     41%|####1     | 17.2M/41.5M [00:00<00:00, 46.2MB/s]
     58%|#####7    | 24.0M/41.5M [00:00<00:00, 44.9MB/s]
     77%|#######7  | 32.0M/41.5M [00:00<00:00, 51.3MB/s]
    100%|##########| 41.5M/41.5M [00:00<00:00, 60.8MB/s]
+
      0%|          | 0.00/41.5M [00:00<?, ?B/s]
     19%|#9        | 7.99M/41.5M [00:00<00:00, 45.6MB/s]
     35%|###4      | 14.3M/41.5M [00:00<00:00, 41.6MB/s]
     53%|#####3    | 22.1M/41.5M [00:00<00:00, 54.4MB/s]
     77%|#######7  | 32.0M/41.5M [00:00<00:00, 53.0MB/s]
     92%|#########2| 38.3M/41.5M [00:00<00:00, 51.4MB/s]
    100%|##########| 41.5M/41.5M [00:00<00:00, 51.0MB/s]
 
 
 
diff --git a/docs/_sources/how_to/compile_models/from_pytorch.rst.txt b/docs/_sources/how_to/compile_models/from_pytorch.rst.txt
index 10c35f62ce..43e101e115 100644
--- a/docs/_sources/how_to/compile_models/from_pytorch.rst.txt
+++ b/docs/_sources/how_to/compile_models/from_pytorch.rst.txt
@@ -102,7 +102,7 @@ Load a pretrained PyTorch model
     /venv/apache-tvm-py3.7/lib/python3.7/site-packages/torchvision/models/_utils.py:223: UserWarning: Arguments other than a weight enum or `None` for 'weights' are deprecated since 0.13 and will be removed in 0.15. The current behavior is equivalent to passing `weights=ResNet18_Weights.IMAGENET1K_V1`. You can also use `weights=ResNet18_Weights.DEFAULT` to get the most up-to-date weights.
       warnings.warn(msg)
     Downloading: "https://download.pytorch.org/models/resnet18-f37072fd.pth" to /workspace/.cache/torch/hub/checkpoints/resnet18-f37072fd.pth
-
      0%|          | 0.00/44.7M [00:00<?, ?B/s]
     18%|#7        | 7.99M/44.7M [00:00<00:00, 66.9MB/s]
     32%|###2      | 14.4M/44.7M [00:00<00:00, 56.9MB/s]
     54%|#####3    | 24.0M/44.7M [00:00<00:00, 60.0MB/s]
     66%|######6   | 29.7M/44.7M [00:00<00:00, 59.4MB/s]
     90%|########9 | 40.0M/44.7M [00:00<00:00, 63.0MB/s]
    100%|##########| 44.7M/44.7M [00:00<00:00, 66.9MB/s]
+
      0%|          | 0.00/44.7M [00:00<?, ?B/s]
     30%|##9       | 13.2M/44.7M [00:00<00:00, 63.0MB/s]
     71%|#######1  | 31.9M/44.7M [00:00<00:00, 116MB/s] 
    100%|##########| 44.7M/44.7M [00:00<00:00, 118MB/s]
 
 
 
diff --git a/docs/_sources/how_to/compile_models/from_tensorflow.rst.txt b/docs/_sources/how_to/compile_models/from_tensorflow.rst.txt
index c9ac1ee764..6748b1f45f 100644
--- a/docs/_sources/how_to/compile_models/from_tensorflow.rst.txt
+++ b/docs/_sources/how_to/compile_models/from_tensorflow.rst.txt
@@ -425,7 +425,7 @@ Run the corresponding model on tensorflow
 
 .. rst-class:: sphx-glr-timing
 
-   **Total running time of the script:** ( 1 minutes  12.122 seconds)
+   **Total running time of the script:** ( 1 minutes  10.824 seconds)
 
 
 .. _sphx_glr_download_how_to_compile_models_from_tensorflow.py:
diff --git a/docs/_sources/how_to/compile_models/sg_execution_times.rst.txt b/docs/_sources/how_to/compile_models/sg_execution_times.rst.txt
index e2dfa4357c..326535aca6 100644
--- a/docs/_sources/how_to/compile_models/sg_execution_times.rst.txt
+++ b/docs/_sources/how_to/compile_models/sg_execution_times.rst.txt
@@ -5,26 +5,26 @@
 
 Computation times
 =================
-**05:47.551** total execution time for **how_to_compile_models** files:
+**05:38.207** total execution time for **how_to_compile_models** files:
 
 +-----------------------------------------------------------------------------------+-----------+--------+
-| :ref:`sphx_glr_how_to_compile_models_from_tensorflow.py` (``from_tensorflow.py``) | 01:12.122 | 0.0 MB |
+| :ref:`sphx_glr_how_to_compile_models_from_tensorflow.py` (``from_tensorflow.py``) | 01:10.824 | 0.0 MB |
 +-----------------------------------------------------------------------------------+-----------+--------+
-| :ref:`sphx_glr_how_to_compile_models_from_darknet.py` (``from_darknet.py``)       | 01:10.448 | 0.0 MB |
+| :ref:`sphx_glr_how_to_compile_models_from_darknet.py` (``from_darknet.py``)       | 01:09.859 | 0.0 MB |
 +-----------------------------------------------------------------------------------+-----------+--------+
-| :ref:`sphx_glr_how_to_compile_models_from_paddle.py` (``from_paddle.py``)         | 00:47.930 | 0.0 MB |
+| :ref:`sphx_glr_how_to_compile_models_from_paddle.py` (``from_paddle.py``)         | 00:45.970 | 0.0 MB |
 +-----------------------------------------------------------------------------------+-----------+--------+
-| :ref:`sphx_glr_how_to_compile_models_from_oneflow.py` (``from_oneflow.py``)       | 00:32.848 | 0.0 MB |
+| :ref:`sphx_glr_how_to_compile_models_from_oneflow.py` (``from_oneflow.py``)       | 00:31.237 | 0.0 MB |
 +-----------------------------------------------------------------------------------+-----------+--------+
-| :ref:`sphx_glr_how_to_compile_models_from_mxnet.py` (``from_mxnet.py``)           | 00:28.909 | 0.0 MB |
+| :ref:`sphx_glr_how_to_compile_models_from_mxnet.py` (``from_mxnet.py``)           | 00:27.711 | 0.0 MB |
 +-----------------------------------------------------------------------------------+-----------+--------+
-| :ref:`sphx_glr_how_to_compile_models_from_coreml.py` (``from_coreml.py``)         | 00:27.055 | 0.0 MB |
+| :ref:`sphx_glr_how_to_compile_models_from_coreml.py` (``from_coreml.py``)         | 00:26.331 | 0.0 MB |
 +-----------------------------------------------------------------------------------+-----------+--------+
-| :ref:`sphx_glr_how_to_compile_models_from_tflite.py` (``from_tflite.py``)         | 00:25.686 | 0.0 MB |
+| :ref:`sphx_glr_how_to_compile_models_from_tflite.py` (``from_tflite.py``)         | 00:25.096 | 0.0 MB |
 +-----------------------------------------------------------------------------------+-----------+--------+
-| :ref:`sphx_glr_how_to_compile_models_from_pytorch.py` (``from_pytorch.py``)       | 00:23.471 | 0.0 MB |
+| :ref:`sphx_glr_how_to_compile_models_from_pytorch.py` (``from_pytorch.py``)       | 00:22.291 | 0.0 MB |
 +-----------------------------------------------------------------------------------+-----------+--------+
-| :ref:`sphx_glr_how_to_compile_models_from_keras.py` (``from_keras.py``)           | 00:16.617 | 0.0 MB |
+| :ref:`sphx_glr_how_to_compile_models_from_keras.py` (``from_keras.py``)           | 00:16.429 | 0.0 MB |
 +-----------------------------------------------------------------------------------+-----------+--------+
-| :ref:`sphx_glr_how_to_compile_models_from_onnx.py` (``from_onnx.py``)             | 00:02.464 | 0.0 MB |
+| :ref:`sphx_glr_how_to_compile_models_from_onnx.py` (``from_onnx.py``)             | 00:02.459 | 0.0 MB |
 +-----------------------------------------------------------------------------------+-----------+--------+
diff --git a/docs/_sources/how_to/deploy_models/deploy_model_on_adreno.rst.txt b/docs/_sources/how_to/deploy_models/deploy_model_on_adreno.rst.txt
index 6fa31214bd..1282916112 100644
--- a/docs/_sources/how_to/deploy_models/deploy_model_on_adreno.rst.txt
+++ b/docs/_sources/how_to/deploy_models/deploy_model_on_adreno.rst.txt
@@ -728,7 +728,7 @@ well as provides information about the model's performance
     Evaluate inference time cost...
     Execution time summary:
      mean (ms)   median (ms)    max (ms)     min (ms)     std (ms)  
-     2755.8343    2754.5199    2766.7212    2752.4388      3.8831   
+     2754.4249    2754.0660    2757.6122    2753.0408      1.3760   
                
 
 
diff --git a/docs/_sources/how_to/deploy_models/deploy_model_on_android.rst.txt b/docs/_sources/how_to/deploy_models/deploy_model_on_android.rst.txt
index 113878785a..7b3cf5a46a 100644
--- a/docs/_sources/how_to/deploy_models/deploy_model_on_android.rst.txt
+++ b/docs/_sources/how_to/deploy_models/deploy_model_on_android.rst.txt
@@ -437,7 +437,7 @@ Execute on TVM
     Evaluate inference time cost...
     Execution time summary:
      mean (ms)   median (ms)    max (ms)     min (ms)     std (ms)  
-      16.5264      16.3593      17.2411      15.9385       0.5255   
+      15.9780      15.8666      16.8045      15.7341       0.3105   
                
 
 
diff --git a/docs/_sources/how_to/deploy_models/deploy_object_detection_pytorch.rst.txt b/docs/_sources/how_to/deploy_models/deploy_object_detection_pytorch.rst.txt
index 2b12ca8c29..54b36a09ad 100644
--- a/docs/_sources/how_to/deploy_models/deploy_object_detection_pytorch.rst.txt
+++ b/docs/_sources/how_to/deploy_models/deploy_object_detection_pytorch.rst.txt
@@ -131,7 +131,7 @@ Load pre-trained maskrcnn from torchvision and do tracing
     /venv/apache-tvm-py3.7/lib/python3.7/site-packages/torchvision/models/_utils.py:223: UserWarning: Arguments other than a weight enum or `None` for 'weights' are deprecated since 0.13 and will be removed in 0.15. The current behavior is equivalent to passing `weights=MaskRCNN_ResNet50_FPN_Weights.COCO_V1`. You can also use `weights=MaskRCNN_ResNet50_FPN_Weights.DEFAULT` to get the most up-to-date weights.
       warnings.warn(msg)
     Downloading: "https://download.pytorch.org/models/maskrcnn_resnet50_fpn_coco-bf2d0c1e.pth" to /workspace/.cache/torch/hub/checkpoints/maskrcnn_resnet50_fpn_coco-bf2d0c1e.pth
-
      0%|          | 0.00/170M [00:00<?, ?B/s]
      5%|4         | 7.99M/170M [00:00<00:02, 78.8MB/s]
      9%|9         | 16.0M/170M [00:00<00:02, 64.8MB/s]
     14%|#4        | 24.0M/170M [00:00<00:02, 55.3MB/s]
     19%|#8        | 32.0M/170M [00:00<00:02, 55.4MB/s]
     22%|##2       | 37.4M/170M [00:00<00:02, 48.5MB/s]
     25%|##4       | 42.1M/170M [00:00<00:02, 46.9MB/s]
     28%|##8       | 48.0M/170M [00:01<00:03, 40.1MB/s]
     37%|###6      | 62.3M/170M [00:01<00:02, 52.8MB/s]
     40%|####      | 68.8M/170M [00:01<00:01, 55.9MB/s]
     46%|####6     | 78.8M/170M [00:01<00:01, 67.1MB/s]
     52%|#####1    | 88.0M/170M [00:01<00:01, 65.6MB/s]
     57%|#####6    | 96.0M/170M [00:01<00:01, 67.1MB/s]
     61%|######1   | 104M/170M [00:01<00:01, 68.7MB/s] 
     66%|######6   | 112M/170M [00:01<00:00, 72.0MB/s]
     71%|#######   | 120M/170M [00:02<00:00, 73.4MB/s]
     75%|#######5  | 128M/170M [00:02<00:00, 69.8MB/s]
     80%|########  | 136M/170M [00:02<00:00, 69.8MB/s]
 
     87%|########6 | 147M/170M [00:02<00:00, 82.8MB/s]
     92%|#########1| 156M/170M [00:02<00:00, 74.1MB/s]
     96%|#########5| 163M/170M [00:02<00:00, 71.9MB/s]
    100%|##########| 170M/170M [00:02<00:00, 64.3MB/s]
+
      0%|          | 0.00/170M [00:00<?, ?B/s]
      5%|4         | 7.99M/170M [00:00<00:02, 79.9MB/s]
      9%|9         | 16.0M/170M [00:00<00:01, 81.3MB/s]
     17%|#7        | 28.9M/170M [00:00<00:01, 105MB/s] 
     24%|##3       | 40.0M/170M [00:00<00:01, 89.8MB/s]
     29%|##8       | 48.9M/170M [00:00<00:01, 87.5MB/s]
     34%|###3      | 57.5M/170M [00:00<00:01, 76.2MB/s]
     38%|###8      | 65.0M/170M [00:00<00:01, 68.9MB/s]
     47%|####7     | 80.0M/170M [00:01<00:01, 77.9MB/s]
     56%|#####5    | 94.9M/170M [00:01<00:00, 96.4MB/s]
     62%|######1   | 105M/170M [00:01<00:00, 86.3MB/s] 
     71%|#######   | 120M/170M [00:01<00:00, 95.8MB/s]
     76%|#######6  | 129M/170M [00:01<00:00, 92.9MB/s]
     82%|########1 | 139M/170M [00:01<00:00, 91.0MB/s]
     87%|########7 | 148M/170M [00:01<00:00, 94.1MB/s]
     93%|#########2| 157M/170M [00:01<00:00, 89.8MB/s]
     98%|#########7| 166M/170M [00:01<00:00, 89.3MB/s]
    100%|##########| 170M/170M [00:02<00:00, 88.0MB/s]
     /venv/apache-tvm-py3.7/lib/python3.7/site-packages/torch/nn/functional.py:3897: UserWarning: To copy construct from a tensor, it is recommended to use sourceTensor.clone().detach() or sourceTensor.clone().detach().requires_grad_(True), rather than torch.tensor(sourceTensor).
       for i in range(dim)
     /venv/apache-tvm-py3.7/lib/python3.7/site-packages/torchvision/models/detection/anchor_utils.py:124: UserWarning: __floordiv__ is deprecated, and its behavior will change in a future version of pytorch. It currently rounds toward 0 (like the 'trunc' function NOT 'floor'). This results in incorrect rounding for negative values. To keep the current behavior, use torch.div(a, b, rounding_mode='trunc'), or for actual floor division, use torch.div(a, b, rounding_mode='floor').
@@ -300,7 +300,7 @@ Get boxes with score larger than 0.9
 
 .. rst-class:: sphx-glr-timing
 
-   **Total running time of the script:** ( 3 minutes  19.797 seconds)
+   **Total running time of the script:** ( 3 minutes  14.579 seconds)
 
 
 .. _sphx_glr_download_how_to_deploy_models_deploy_object_detection_pytorch.py:
diff --git a/docs/_sources/how_to/deploy_models/deploy_prequantized.rst.txt b/docs/_sources/how_to/deploy_models/deploy_prequantized.rst.txt
index 35fae3c879..dfd84f2d28 100644
--- a/docs/_sources/how_to/deploy_models/deploy_prequantized.rst.txt
+++ b/docs/_sources/how_to/deploy_models/deploy_prequantized.rst.txt
@@ -240,7 +240,7 @@ training. Other models require a full post training calibration.
     /venv/apache-tvm-py3.7/lib/python3.7/site-packages/torchvision/models/_utils.py:223: UserWarning: Arguments other than a weight enum or `None` for 'weights' are deprecated since 0.13 and will be removed in 0.15. The current behavior is equivalent to passing `weights=MobileNet_V2_Weights.IMAGENET1K_V1`. You can also use `weights=MobileNet_V2_Weights.DEFAULT` to get the most up-to-date weights.
       warnings.warn(msg)
     Downloading: "https://download.pytorch.org/models/mobilenet_v2-b0353104.pth" to /workspace/.cache/torch/hub/checkpoints/mobilenet_v2-b0353104.pth
-
      0%|          | 0.00/13.6M [00:00<?, ?B/s]
     59%|#####8    | 7.99M/13.6M [00:00<00:00, 57.1MB/s]
    100%|##########| 13.6M/13.6M [00:00<00:00, 79.7MB/s]
+
      0%|          | 0.00/13.6M [00:00<?, ?B/s]
     60%|######    | 8.14M/13.6M [00:00<00:00, 85.3MB/s]
    100%|##########| 13.6M/13.6M [00:00<00:00, 112MB/s] 
 
 
 
@@ -422,7 +422,7 @@ Here we give an example of how to measure performance of TVM compiled models.
 
     Execution time summary:
      mean (ms)   median (ms)    max (ms)     min (ms)     std (ms)  
-      90.5098      90.4528      92.3407      90.1770       0.2807   
+      90.3708      90.2761      91.4479      90.0140       0.2925   
                
 
 
@@ -471,7 +471,7 @@ TODO
 
 .. rst-class:: sphx-glr-timing
 
-   **Total running time of the script:** ( 1 minutes  7.152 seconds)
+   **Total running time of the script:** ( 1 minutes  5.873 seconds)
 
 
 .. _sphx_glr_download_how_to_deploy_models_deploy_prequantized.py:
diff --git a/docs/_sources/how_to/deploy_models/deploy_prequantized_tflite.rst.txt b/docs/_sources/how_to/deploy_models/deploy_prequantized_tflite.rst.txt
index b5d5a51293..6927841b47 100644
--- a/docs/_sources/how_to/deploy_models/deploy_prequantized_tflite.rst.txt
+++ b/docs/_sources/how_to/deploy_models/deploy_prequantized_tflite.rst.txt
@@ -436,7 +436,7 @@ Here we give an example of how to measure performance of TVM compiled models.
 
     Execution time summary:
      mean (ms)   median (ms)    max (ms)     min (ms)     std (ms)  
-      121.3945     121.2771     125.1450     120.3455      0.6445   
+      120.5424     120.5388     122.1327     119.7057      0.3824   
                
 
 
@@ -473,7 +473,7 @@ Here we give an example of how to measure performance of TVM compiled models.
 
 .. rst-class:: sphx-glr-timing
 
-   **Total running time of the script:** ( 2 minutes  33.878 seconds)
+   **Total running time of the script:** ( 2 minutes  28.183 seconds)
 
 
 .. _sphx_glr_download_how_to_deploy_models_deploy_prequantized_tflite.py:
diff --git a/docs/_sources/how_to/deploy_models/deploy_quantized.rst.txt b/docs/_sources/how_to/deploy_models/deploy_quantized.rst.txt
index 2b08225791..a322b1554d 100644
--- a/docs/_sources/how_to/deploy_models/deploy_quantized.rst.txt
+++ b/docs/_sources/how_to/deploy_models/deploy_quantized.rst.txt
@@ -257,7 +257,7 @@ We create a Relay VM to build and execute the model.
 
 .. rst-class:: sphx-glr-timing
 
-   **Total running time of the script:** ( 1 minutes  27.833 seconds)
+   **Total running time of the script:** ( 1 minutes  32.385 seconds)
 
 
 .. _sphx_glr_download_how_to_deploy_models_deploy_quantized.py:
diff --git a/docs/_sources/how_to/deploy_models/deploy_ssd_gluoncv.rst.txt b/docs/_sources/how_to/deploy_models/deploy_ssd_gluoncv.rst.txt
index b59aaf0538..93623eb643 100644
--- a/docs/_sources/how_to/deploy_models/deploy_ssd_gluoncv.rst.txt
+++ b/docs/_sources/how_to/deploy_models/deploy_ssd_gluoncv.rst.txt
@@ -170,7 +170,7 @@ Convert and compile model for CPU.
             data: None
       input_sym_arg_type = in_param.infer_type()[0]
     Downloading /workspace/.mxnet/models/ssd_512_resnet50_v1_voc-9c8b225a.zip from https://apache-mxnet.s3-accelerate.dualstack.amazonaws.com/gluon/models/ssd_512_resnet50_v1_voc-9c8b225a.zip...
-
      0%|          | 0/132723 [00:00<?, ?KB/s]
      5%|5         | 6816/132723 [00:00<00:01, 68151.94KB/s]
     12%|#1        | 15413/132723 [00:00<00:01, 78503.22KB/s]
     18%|#8        | 24101/132723 [00:00<00:01, 82319.91KB/s]
     25%|##4       | 32802/132723 [00:00<00:01, 84165.21KB/s]
     31%|###1      | 41555/132723 [00:00<00:01, 85374.95KB/s]
     38%|###7      | 50208/132723 [00:00<00:00, 85764.79KB/s]
     44%|####4     | 58999/132723 [00:00<00:00, 86464.57KB/s]
     51%|#####1    | 67902/132723 [00:00<00:00, 87278.52KB/s]
     58%|#####7    | 76936/132723 [00:00<00:00, 88231.93KB/s]
     65%|######4   | 85905/132723 [00:01<00:00, 88680.67KB/s]
     72%|#######1  | 94935/132723 [00:01<00:00, 89173.39KB/s]
     78%|#######8  | 103923/132723 [00:01<00:00, 89386.54KB/s]
     85%|########5 | 112884/132723 [00:01<00:00, 89452.96KB/s]
     92%|#########1| 121856/132723 [00:01<00:00, 89528.03KB/s]
     99%|#########8| 130809/132723 [00:01<00:00, 89515.98KB/s]
    100%|#######
 ###| 132723/132723 [00:01<00:00, 87096.42KB/s]
+
      0%|          | 0/132723 [00:00<?, ?KB/s]
      5%|4         | 6005/132723 [00:00<00:02, 60036.75KB/s]
     11%|#         | 13948/132723 [00:00<00:01, 71441.04KB/s]
     17%|#6        | 22552/132723 [00:00<00:01, 78105.37KB/s]
     24%|##3       | 31193/132723 [00:00<00:01, 81382.58KB/s]
     30%|##9       | 39332/132723 [00:00<00:01, 80918.42KB/s]
     36%|###6      | 47973/132723 [00:00<00:01, 82776.58KB/s]
     43%|####2     | 56665/132723 [00:00<00:00, 84124.80KB/s]
     49%|####9     | 65314/132723 [00:00<00:00, 84874.39KB/s]
     56%|#####5    | 73903/132723 [00:00<00:00, 85188.53KB/s]
     62%|######2   | 82532/132723 [00:01<00:00, 85526.10KB/s]
     69%|######8   | 91169/132723 [00:01<00:00, 85781.17KB/s]
     75%|#######5  | 99909/132723 [00:01<00:00, 86272.37KB/s]
     82%|########1 | 108553/132723 [00:01<00:00, 86321.98KB/s]
     88%|########8 | 117186/132723 [00:01<00:00, 86139.24KB/s]
     95%|#########4| 125801/132723 [00:01<00:00, 86084.69KB/s]
    100%|########
 ##| 132723/132723 [00:01<00:00, 83737.23KB/s]
 
 
 
@@ -246,7 +246,7 @@ Display result
 
 .. rst-class:: sphx-glr-timing
 
-   **Total running time of the script:** ( 3 minutes  12.197 seconds)
+   **Total running time of the script:** ( 3 minutes  5.615 seconds)
 
 
 .. _sphx_glr_download_how_to_deploy_models_deploy_ssd_gluoncv.py:
diff --git a/docs/_sources/how_to/deploy_models/sg_execution_times.rst.txt b/docs/_sources/how_to/deploy_models/sg_execution_times.rst.txt
index a299b6dd4c..9a9fa23684 100644
--- a/docs/_sources/how_to/deploy_models/sg_execution_times.rst.txt
+++ b/docs/_sources/how_to/deploy_models/sg_execution_times.rst.txt
@@ -5,26 +5,26 @@
 
 Computation times
 =================
-**14:03.078** total execution time for **how_to_deploy_models** files:
+**13:46.850** total execution time for **how_to_deploy_models** files:
 
 +------------------------------------------------------------------------------------------------------------------+-----------+--------+
-| :ref:`sphx_glr_how_to_deploy_models_deploy_object_detection_pytorch.py` (``deploy_object_detection_pytorch.py``) | 03:19.797 | 0.0 MB |
+| :ref:`sphx_glr_how_to_deploy_models_deploy_object_detection_pytorch.py` (``deploy_object_detection_pytorch.py``) | 03:14.579 | 0.0 MB |
 +------------------------------------------------------------------------------------------------------------------+-----------+--------+
-| :ref:`sphx_glr_how_to_deploy_models_deploy_ssd_gluoncv.py` (``deploy_ssd_gluoncv.py``)                           | 03:12.197 | 0.0 MB |
+| :ref:`sphx_glr_how_to_deploy_models_deploy_ssd_gluoncv.py` (``deploy_ssd_gluoncv.py``)                           | 03:05.615 | 0.0 MB |
 +------------------------------------------------------------------------------------------------------------------+-----------+--------+
-| :ref:`sphx_glr_how_to_deploy_models_deploy_prequantized_tflite.py` (``deploy_prequantized_tflite.py``)           | 02:33.878 | 0.0 MB |
+| :ref:`sphx_glr_how_to_deploy_models_deploy_prequantized_tflite.py` (``deploy_prequantized_tflite.py``)           | 02:28.183 | 0.0 MB |
 +------------------------------------------------------------------------------------------------------------------+-----------+--------+
-| :ref:`sphx_glr_how_to_deploy_models_deploy_quantized.py` (``deploy_quantized.py``)                               | 01:27.833 | 0.0 MB |
+| :ref:`sphx_glr_how_to_deploy_models_deploy_quantized.py` (``deploy_quantized.py``)                               | 01:32.385 | 0.0 MB |
 +------------------------------------------------------------------------------------------------------------------+-----------+--------+
-| :ref:`sphx_glr_how_to_deploy_models_deploy_prequantized.py` (``deploy_prequantized.py``)                         | 01:07.152 | 0.0 MB |
+| :ref:`sphx_glr_how_to_deploy_models_deploy_prequantized.py` (``deploy_prequantized.py``)                         | 01:05.873 | 0.0 MB |
 +------------------------------------------------------------------------------------------------------------------+-----------+--------+
-| :ref:`sphx_glr_how_to_deploy_models_deploy_model_on_adreno.py` (``deploy_model_on_adreno.py``)                   | 00:54.484 | 0.0 MB |
+| :ref:`sphx_glr_how_to_deploy_models_deploy_model_on_adreno.py` (``deploy_model_on_adreno.py``)                   | 00:53.857 | 0.0 MB |
 +------------------------------------------------------------------------------------------------------------------+-----------+--------+
-| :ref:`sphx_glr_how_to_deploy_models_deploy_model_on_android.py` (``deploy_model_on_android.py``)                 | 00:36.060 | 0.0 MB |
+| :ref:`sphx_glr_how_to_deploy_models_deploy_model_on_android.py` (``deploy_model_on_android.py``)                 | 00:35.316 | 0.0 MB |
 +------------------------------------------------------------------------------------------------------------------+-----------+--------+
-| :ref:`sphx_glr_how_to_deploy_models_deploy_model_on_nano.py` (``deploy_model_on_nano.py``)                       | 00:26.002 | 0.0 MB |
+| :ref:`sphx_glr_how_to_deploy_models_deploy_model_on_nano.py` (``deploy_model_on_nano.py``)                       | 00:25.701 | 0.0 MB |
 +------------------------------------------------------------------------------------------------------------------+-----------+--------+
-| :ref:`sphx_glr_how_to_deploy_models_deploy_model_on_rasp.py` (``deploy_model_on_rasp.py``)                       | 00:25.669 | 0.0 MB |
+| :ref:`sphx_glr_how_to_deploy_models_deploy_model_on_rasp.py` (``deploy_model_on_rasp.py``)                       | 00:25.334 | 0.0 MB |
 +------------------------------------------------------------------------------------------------------------------+-----------+--------+
 | :ref:`sphx_glr_how_to_deploy_models_deploy_sparse.py` (``deploy_sparse.py``)                                     | 00:00.007 | 0.0 MB |
 +------------------------------------------------------------------------------------------------------------------+-----------+--------+
diff --git a/docs/_sources/how_to/extend_tvm/bring_your_own_datatypes.rst.txt b/docs/_sources/how_to/extend_tvm/bring_your_own_datatypes.rst.txt
index 33005de56e..c0e46f25d7 100644
--- a/docs/_sources/how_to/extend_tvm/bring_your_own_datatypes.rst.txt
+++ b/docs/_sources/how_to/extend_tvm/bring_your_own_datatypes.rst.txt
@@ -476,7 +476,7 @@ First let us define two helper functions to get the mobilenet model and a cat im
 
  .. code-block:: none
 
-    Downloading /workspace/.mxnet/models/mobilenet0.25-9f83e440.zipcba088a8-c84b-423d-b81b-c9e460cb7f68 from https://apache-mxnet.s3-accelerate.dualstack.amazonaws.com/gluon/models/mobilenet0.25-9f83e440.zip...
+    Downloading /workspace/.mxnet/models/mobilenet0.25-9f83e440.zip258d8e55-a0b4-49bf-8a93-02ba6b31ee53 from https://apache-mxnet.s3-accelerate.dualstack.amazonaws.com/gluon/models/mobilenet0.25-9f83e440.zip...
 
 
 
diff --git a/docs/_sources/how_to/extend_tvm/sg_execution_times.rst.txt b/docs/_sources/how_to/extend_tvm/sg_execution_times.rst.txt
index bb919f5034..ff5f141195 100644
--- a/docs/_sources/how_to/extend_tvm/sg_execution_times.rst.txt
+++ b/docs/_sources/how_to/extend_tvm/sg_execution_times.rst.txt
@@ -5,14 +5,14 @@
 
 Computation times
 =================
-**00:48.648** total execution time for **how_to_extend_tvm** files:
+**00:47.361** total execution time for **how_to_extend_tvm** files:
 
 +-------------------------------------------------------------------------------------------------+-----------+--------+
-| :ref:`sphx_glr_how_to_extend_tvm_bring_your_own_datatypes.py` (``bring_your_own_datatypes.py``) | 00:45.102 | 0.0 MB |
+| :ref:`sphx_glr_how_to_extend_tvm_bring_your_own_datatypes.py` (``bring_your_own_datatypes.py``) | 00:43.920 | 0.0 MB |
 +-------------------------------------------------------------------------------------------------+-----------+--------+
-| :ref:`sphx_glr_how_to_extend_tvm_use_pass_instrument.py` (``use_pass_instrument.py``)           | 00:02.478 | 0.0 MB |
+| :ref:`sphx_glr_how_to_extend_tvm_use_pass_instrument.py` (``use_pass_instrument.py``)           | 00:02.409 | 0.0 MB |
 +-------------------------------------------------------------------------------------------------+-----------+--------+
-| :ref:`sphx_glr_how_to_extend_tvm_use_pass_infra.py` (``use_pass_infra.py``)                     | 00:01.059 | 0.0 MB |
+| :ref:`sphx_glr_how_to_extend_tvm_use_pass_infra.py` (``use_pass_infra.py``)                     | 00:01.024 | 0.0 MB |
 +-------------------------------------------------------------------------------------------------+-----------+--------+
-| :ref:`sphx_glr_how_to_extend_tvm_low_level_custom_pass.py` (``low_level_custom_pass.py``)       | 00:00.009 | 0.0 MB |
+| :ref:`sphx_glr_how_to_extend_tvm_low_level_custom_pass.py` (``low_level_custom_pass.py``)       | 00:00.008 | 0.0 MB |
 +-------------------------------------------------------------------------------------------------+-----------+--------+
diff --git a/docs/_sources/how_to/extend_tvm/use_pass_instrument.rst.txt b/docs/_sources/how_to/extend_tvm/use_pass_instrument.rst.txt
index 2c1223371a..64eeab88d7 100644
--- a/docs/_sources/how_to/extend_tvm/use_pass_instrument.rst.txt
+++ b/docs/_sources/how_to/extend_tvm/use_pass_instrument.rst.txt
@@ -220,10 +220,10 @@ profile the execution time of each passes.
  .. code-block:: none
 
     Printing results of timing profile...
-    InferType: 7467us [7467us] (47.17%; 47.17%)
-    FoldScaleAxis: 8363us [7us] (52.83%; 52.83%)
-            FoldConstant: 8356us [1717us] (52.79%; 99.92%)
-                    InferType: 6639us [6639us] (41.94%; 79.45%)
+    InferType: 7196us [7196us] (46.44%; 46.44%)
+    FoldScaleAxis: 8299us [7us] (53.56%; 53.56%)
+            FoldConstant: 8292us [1698us] (53.52%; 99.92%)
+                    InferType: 6594us [6594us] (42.56%; 79.53%)
 
 
 
@@ -262,10 +262,10 @@ Refer to following sections and :py:func:`tvm.instrument.pass_instrument` for th
  .. code-block:: none
 
     Printing results of timing profile...
-    InferType: 6707us [6707us] (44.67%; 44.67%)
-    FoldScaleAxis: 8307us [6us] (55.33%; 55.33%)
-            FoldConstant: 8301us [1715us] (55.29%; 99.93%)
-                    InferType: 6586us [6586us] (43.87%; 79.34%)
+    InferType: 6608us [6608us] (44.86%; 44.86%)
+    FoldScaleAxis: 8123us [6us] (55.14%; 55.14%)
+            FoldConstant: 8118us [1676us] (55.10%; 99.93%)
+                    InferType: 6441us [6441us] (43.72%; 79.35%)
 
 
 
diff --git a/docs/_sources/how_to/optimize_operators/opt_conv_cuda.rst.txt b/docs/_sources/how_to/optimize_operators/opt_conv_cuda.rst.txt
index 1fc0c8ec13..9c432f9a3d 100644
--- a/docs/_sources/how_to/optimize_operators/opt_conv_cuda.rst.txt
+++ b/docs/_sources/how_to/optimize_operators/opt_conv_cuda.rst.txt
@@ -344,7 +344,7 @@ latency of convolution.
 
  .. code-block:: none
 
-    Convolution: 54.149791 ms
+    Convolution: 53.049343 ms
 
 
 
diff --git a/docs/_sources/how_to/optimize_operators/opt_conv_tensorcore.rst.txt b/docs/_sources/how_to/optimize_operators/opt_conv_tensorcore.rst.txt
index bb7883b64a..62b6d545d0 100644
--- a/docs/_sources/how_to/optimize_operators/opt_conv_tensorcore.rst.txt
+++ b/docs/_sources/how_to/optimize_operators/opt_conv_tensorcore.rst.txt
@@ -661,7 +661,7 @@ be able to run on our build server
 
  .. code-block:: none
 
-    conv2d with tensor core: 13.359639 ms
+    conv2d with tensor core: 13.356647 ms
 
 
 
diff --git a/docs/_sources/how_to/optimize_operators/opt_gemm.rst.txt b/docs/_sources/how_to/optimize_operators/opt_gemm.rst.txt
index 261403937f..6e97dcc8ec 100644
--- a/docs/_sources/how_to/optimize_operators/opt_gemm.rst.txt
+++ b/docs/_sources/how_to/optimize_operators/opt_gemm.rst.txt
@@ -147,8 +147,8 @@ Then we write a baseline implementation, the simplest way to write a matrix mult
 
  .. code-block:: none
 
-    Numpy running time: 0.019011
-    Baseline: 3.401171
+    Numpy running time: 0.018666
+    Baseline: 3.409708
 
 
 
@@ -242,7 +242,7 @@ fill 32 * 32 * sizeof(float) which is 4KB in the cache whose total size is 32KB
 
  .. code-block:: none
 
-    Opt1: 0.325391
+    Opt1: 0.303196
 
 
 
@@ -344,7 +344,7 @@ In this tutorial, we chose to vectorize the inner loop row data since it is cach
 
  .. code-block:: none
 
-    Opt2: 0.350191
+    Opt2: 0.339247
 
 
 
@@ -439,7 +439,7 @@ the access pattern for A matrix is more cache friendly.
 
  .. code-block:: none
 
-    Opt3: 0.119524
+    Opt3: 0.117116
 
 
 
@@ -563,7 +563,7 @@ flattening.
 
  .. code-block:: none
 
-    Opt4: 0.109798
+    Opt4: 0.109304
 
 
 
@@ -684,7 +684,7 @@ write to C when all the block results are ready.
 
  .. code-block:: none
 
-    Opt5: 0.111408
+    Opt5: 0.110906
 
 
 
@@ -808,7 +808,7 @@ Furthermore, we can also utilize multi-core processors to do the thread-level pa
 
  .. code-block:: none
 
-    Opt6: 0.147337
+    Opt6: 0.147578
 
 
 
diff --git a/docs/_sources/how_to/optimize_operators/sg_execution_times.rst.txt b/docs/_sources/how_to/optimize_operators/sg_execution_times.rst.txt
index 616b1a9080..1a0670b5c6 100644
--- a/docs/_sources/how_to/optimize_operators/sg_execution_times.rst.txt
+++ b/docs/_sources/how_to/optimize_operators/sg_execution_times.rst.txt
@@ -5,12 +5,12 @@
 
 Computation times
 =================
-**00:35.675** total execution time for **how_to_optimize_operators** files:
+**00:35.161** total execution time for **how_to_optimize_operators** files:
 
 +-----------------------------------------------------------------------------------------------+-----------+--------+
-| :ref:`sphx_glr_how_to_optimize_operators_opt_gemm.py` (``opt_gemm.py``)                       | 00:32.872 | 0.0 MB |
+| :ref:`sphx_glr_how_to_optimize_operators_opt_gemm.py` (``opt_gemm.py``)                       | 00:32.403 | 0.0 MB |
 +-----------------------------------------------------------------------------------------------+-----------+--------+
-| :ref:`sphx_glr_how_to_optimize_operators_opt_conv_tensorcore.py` (``opt_conv_tensorcore.py``) | 00:01.627 | 0.0 MB |
+| :ref:`sphx_glr_how_to_optimize_operators_opt_conv_tensorcore.py` (``opt_conv_tensorcore.py``) | 00:01.553 | 0.0 MB |
 +-----------------------------------------------------------------------------------------------+-----------+--------+
-| :ref:`sphx_glr_how_to_optimize_operators_opt_conv_cuda.py` (``opt_conv_cuda.py``)             | 00:01.175 | 0.0 MB |
+| :ref:`sphx_glr_how_to_optimize_operators_opt_conv_cuda.py` (``opt_conv_cuda.py``)             | 00:01.206 | 0.0 MB |
 +-----------------------------------------------------------------------------------------------+-----------+--------+
diff --git a/docs/_sources/how_to/tune_with_autoscheduler/sg_execution_times.rst.txt b/docs/_sources/how_to/tune_with_autoscheduler/sg_execution_times.rst.txt
index 8d94b3c7b9..c415b8ad0c 100644
--- a/docs/_sources/how_to/tune_with_autoscheduler/sg_execution_times.rst.txt
+++ b/docs/_sources/how_to/tune_with_autoscheduler/sg_execution_times.rst.txt
@@ -5,18 +5,18 @@
 
 Computation times
 =================
-**09:13.652** total execution time for **how_to_tune_with_autoscheduler** files:
+**09:03.580** total execution time for **how_to_tune_with_autoscheduler** files:
 
 +----------------------------------------------------------------------------------------------------------+-----------+--------+
-| :ref:`sphx_glr_how_to_tune_with_autoscheduler_tune_conv2d_layer_cuda.py` (``tune_conv2d_layer_cuda.py``) | 05:43.471 | 0.0 MB |
+| :ref:`sphx_glr_how_to_tune_with_autoscheduler_tune_conv2d_layer_cuda.py` (``tune_conv2d_layer_cuda.py``) | 05:37.856 | 0.0 MB |
 +----------------------------------------------------------------------------------------------------------+-----------+--------+
-| :ref:`sphx_glr_how_to_tune_with_autoscheduler_tune_network_x86.py` (``tune_network_x86.py``)             | 01:34.285 | 0.0 MB |
+| :ref:`sphx_glr_how_to_tune_with_autoscheduler_tune_network_x86.py` (``tune_network_x86.py``)             | 01:31.696 | 0.0 MB |
 +----------------------------------------------------------------------------------------------------------+-----------+--------+
-| :ref:`sphx_glr_how_to_tune_with_autoscheduler_tune_network_cuda.py` (``tune_network_cuda.py``)           | 01:02.866 | 0.0 MB |
+| :ref:`sphx_glr_how_to_tune_with_autoscheduler_tune_network_cuda.py` (``tune_network_cuda.py``)           | 01:01.645 | 0.0 MB |
 +----------------------------------------------------------------------------------------------------------+-----------+--------+
-| :ref:`sphx_glr_how_to_tune_with_autoscheduler_tune_sparse_x86.py` (``tune_sparse_x86.py``)               | 00:29.070 | 0.0 MB |
+| :ref:`sphx_glr_how_to_tune_with_autoscheduler_tune_sparse_x86.py` (``tune_sparse_x86.py``)               | 00:29.326 | 0.0 MB |
 +----------------------------------------------------------------------------------------------------------+-----------+--------+
-| :ref:`sphx_glr_how_to_tune_with_autoscheduler_tune_network_arm.py` (``tune_network_arm.py``)             | 00:12.384 | 0.0 MB |
+| :ref:`sphx_glr_how_to_tune_with_autoscheduler_tune_network_arm.py` (``tune_network_arm.py``)             | 00:12.006 | 0.0 MB |
 +----------------------------------------------------------------------------------------------------------+-----------+--------+
-| :ref:`sphx_glr_how_to_tune_with_autoscheduler_tune_network_mali.py` (``tune_network_mali.py``)           | 00:11.575 | 0.0 MB |
+| :ref:`sphx_glr_how_to_tune_with_autoscheduler_tune_network_mali.py` (``tune_network_mali.py``)           | 00:11.050 | 0.0 MB |
 +----------------------------------------------------------------------------------------------------------+-----------+--------+
diff --git a/docs/_sources/how_to/tune_with_autoscheduler/tune_conv2d_layer_cuda.rst.txt b/docs/_sources/how_to/tune_with_autoscheduler/tune_conv2d_layer_cuda.rst.txt
index 80af82107f..4b612a856a 100644
--- a/docs/_sources/how_to/tune_with_autoscheduler/tune_conv2d_layer_cuda.rst.txt
+++ b/docs/_sources/how_to/tune_with_autoscheduler/tune_conv2d_layer_cuda.rst.txt
@@ -210,13 +210,6 @@ file and apply it.
 
 
 
-.. rst-class:: sphx-glr-script-out
-
- .. code-block:: none
-
-    .T
-
-
 
 
 
@@ -252,10 +245,10 @@ cooperative fetching, unrolling and operator fusion.
       buffer_map = {data_1: data, kernel_1: kernel, bias_1: bias, compute_1: compute} {
       attr [IterVar(blockIdx.x: int32, (nullptr), "ThreadIndex", "blockIdx.x")] "thread_extent" = 28;
       allocate(conv2d_nchw: Pointer(local float32), float32, [14]), storage_scope = local;
-      allocate(pad_temp.shared: Pointer(shared float32), float32, [72]), storage_scope = shared;
-      allocate(kernel.shared: Pointer(shared float32), float32, [3072]), storage_scope = shared;
+      allocate(pad_temp.shared: Pointer(shared float32), float32, [216]), storage_scope = shared;
+      allocate(kernel.shared: Pointer(shared float32), float32, [9216]), storage_scope = shared;
       attr [IterVar(threadIdx.x: int32, (nullptr), "ThreadIndex", "threadIdx.x")] "thread_extent" = 64 {
-        conv2d_nchw_1: Buffer(conv2d_nchw, float32, [14], [], scope="local", align=32)[0] = 0f32
+        conv2d_nchw_1: Buffer(conv2d_nchw, float32, [1], [], scope="local", align=4)[0] = 0f32
         conv2d_nchw_1[1] = 0f32
         conv2d_nchw_1[2] = 0f32
         conv2d_nchw_1[3] = 0f32
@@ -270,464 +263,1331 @@ cooperative fetching, unrolling and operator fusion.
         conv2d_nchw_1[12] = 0f32
         conv2d_nchw_1[13] = 0f32
         for (rc.outer.outer: int32, 0, 64) {
-          for (ry.outer.outer: int32, 0, 3) {
-            let cse_var_2: int32 = (rc.outer.outer*72)
-            let cse_var_1: int32 = (ry.outer.outer*3)
-             {
-              attr [IterVar(threadIdx.x_1: int32, (nullptr), "ThreadIndex", "threadIdx.x")] "thread_extent" = 64 {
-                if @tir.likely((threadIdx.x_1 < 18), dtype=bool) {
-                  pad_temp.shared_1: Buffer(pad_temp.shared, float32, [72], [], scope="shared")[(threadIdx.x_1*4)] = @tir.if_then_else(((((1 <= (ry.outer.outer + floormod(blockIdx.x, 7))) && ((ry.outer.outer + floormod(blockIdx.x, 7)) < 8)) && (1 <= floormod((threadIdx.x_1*4), 9))) && (floormod((threadIdx.x_1*4), 9) < 8)), data_3: Buffer(data_2, float32, [25088], [])[((((((rc.outer.outer*392) + (floordiv((threadIdx.x_1*4), 9)*49)) + (ry.outer.outer*7)) + (floormod(blockIdx.x, 7)*7)) + fl [...]
-                }
-                if @tir.likely((threadIdx.x_1 < 18), dtype=bool) {
-                  pad_temp.shared_1[((threadIdx.x_1*4) + 1)] = @tir.if_then_else(((((1 <= (ry.outer.outer + floormod(blockIdx.x, 7))) && ((ry.outer.outer + floormod(blockIdx.x, 7)) < 8)) && (1 <= floormod(((threadIdx.x_1*4) + 1), 9))) && (floormod(((threadIdx.x_1*4) + 1), 9) < 8)), data_3[((((((rc.outer.outer*392) + (floordiv(((threadIdx.x_1*4) + 1), 9)*49)) + (ry.outer.outer*7)) + (floormod(blockIdx.x, 7)*7)) + floormod(((threadIdx.x_1*4) + 1), 9)) - 8)], 0f32, dtype=float32)
-                }
-                if @tir.likely((threadIdx.x_1 < 18), dtype=bool) {
-                  pad_temp.shared_1[((threadIdx.x_1*4) + 2)] = @tir.if_then_else(((((1 <= (ry.outer.outer + floormod(blockIdx.x, 7))) && ((ry.outer.outer + floormod(blockIdx.x, 7)) < 8)) && (1 <= floormod(((threadIdx.x_1*4) + 2), 9))) && (floormod(((threadIdx.x_1*4) + 2), 9) < 8)), data_3[((((((rc.outer.outer*392) + (floordiv(((threadIdx.x_1*4) + 2), 9)*49)) + (ry.outer.outer*7)) + (floormod(blockIdx.x, 7)*7)) + floormod(((threadIdx.x_1*4) + 2), 9)) - 8)], 0f32, dtype=float32)
-                }
-                if @tir.likely((threadIdx.x_1 < 18), dtype=bool) {
-                  pad_temp.shared_1[((threadIdx.x_1*4) + 3)] = @tir.if_then_else(((((1 <= (ry.outer.outer + floormod(blockIdx.x, 7))) && ((ry.outer.outer + floormod(blockIdx.x, 7)) < 8)) && (1 <= floormod(((threadIdx.x_1*4) + 3), 9))) && (floormod(((threadIdx.x_1*4) + 3), 9) < 8)), data_3[((((((rc.outer.outer*392) + (floordiv(((threadIdx.x_1*4) + 3), 9)*49)) + (ry.outer.outer*7)) + (floormod(blockIdx.x, 7)*7)) + floormod(((threadIdx.x_1*4) + 3), 9)) - 8)], 0f32, dtype=float32)
-                }
-              }
-              attr [IterVar(threadIdx.x_2: int32, (nullptr), "ThreadIndex", "threadIdx.x")] "thread_extent" = 64;
-              kernel.shared_1: Buffer(kernel.shared, float32, [3072], [], scope="shared")[threadIdx.x_2] = kernel_3: Buffer(kernel_2, float32, [2359296], [])[((((((floordiv(blockIdx.x, 7)*589824) + (floordiv(threadIdx.x_2, 24)*4608)) + cse_var_2) + (floordiv(floormod(threadIdx.x_2, 24), 3)*9)) + cse_var_1) + floormod(threadIdx.x_2, 3))]
-              attr [IterVar(threadIdx.x_2, (nullptr), "ThreadIndex", "threadIdx.x")] "thread_extent" = 64;
-              kernel.shared_1[(threadIdx.x_2 + 64)] = kernel_3[((((((floordiv(blockIdx.x, 7)*589824) + (floordiv((threadIdx.x_2 + 64), 24)*4608)) + cse_var_2) + (floordiv(floormod((threadIdx.x_2 + 16), 24), 3)*9)) + cse_var_1) + floormod((threadIdx.x_2 + 1), 3))]
-              attr [IterVar(threadIdx.x_2, (nullptr), "ThreadIndex", "threadIdx.x")] "thread_extent" = 64;
-              kernel.shared_1[(threadIdx.x_2 + 128)] = kernel_3[((((((floordiv(blockIdx.x, 7)*589824) + (floordiv((threadIdx.x_2 + 128), 24)*4608)) + cse_var_2) + (floordiv(floormod((threadIdx.x_2 + 8), 24), 3)*9)) + cse_var_1) + floormod((threadIdx.x_2 + 2), 3))]
-              attr [IterVar(threadIdx.x_2, (nullptr), "ThreadIndex", "threadIdx.x")] "thread_extent" = 64;
-              kernel.shared_1[(threadIdx.x_2 + 192)] = kernel_3[(((((((floordiv(blockIdx.x, 7)*589824) + (floordiv(threadIdx.x_2, 24)*4608)) + cse_var_2) + (floordiv(floormod(threadIdx.x_2, 24), 3)*9)) + cse_var_1) + floormod(threadIdx.x_2, 3)) + 36864)]
-              attr [IterVar(threadIdx.x_2, (nullptr), "ThreadIndex", "threadIdx.x")] "thread_extent" = 64;
-              kernel.shared_1[(threadIdx.x_2 + 256)] = kernel_3[((((((floordiv(blockIdx.x, 7)*589824) + (floordiv((threadIdx.x_2 + 256), 24)*4608)) + cse_var_2) + (floordiv(floormod((threadIdx.x_2 + 16), 24), 3)*9)) + cse_var_1) + floormod((threadIdx.x_2 + 1), 3))]
-              attr [IterVar(threadIdx.x_2, (nullptr), "ThreadIndex", "threadIdx.x")] "thread_extent" = 64;
-              kernel.shared_1[(threadIdx.x_2 + 320)] = kernel_3[((((((floordiv(blockIdx.x, 7)*589824) + (floordiv((threadIdx.x_2 + 320), 24)*4608)) + cse_var_2) + (floordiv(floormod((threadIdx.x_2 + 8), 24), 3)*9)) + cse_var_1) + floormod((threadIdx.x_2 + 2), 3))]
-              attr [IterVar(threadIdx.x_2, (nullptr), "ThreadIndex", "threadIdx.x")] "thread_extent" = 64;
-              kernel.shared_1[(threadIdx.x_2 + 384)] = kernel_3[(((((((floordiv(blockIdx.x, 7)*589824) + (floordiv(threadIdx.x_2, 24)*4608)) + cse_var_2) + (floordiv(floormod(threadIdx.x_2, 24), 3)*9)) + cse_var_1) + floormod(threadIdx.x_2, 3)) + 73728)]
-              attr [IterVar(threadIdx.x_2, (nullptr), "ThreadIndex", "threadIdx.x")] "thread_extent" = 64;
-              kernel.shared_1[(threadIdx.x_2 + 448)] = kernel_3[((((((floordiv(blockIdx.x, 7)*589824) + (floordiv((threadIdx.x_2 + 448), 24)*4608)) + cse_var_2) + (floordiv(floormod((threadIdx.x_2 + 16), 24), 3)*9)) + cse_var_1) + floormod((threadIdx.x_2 + 1), 3))]
-              attr [IterVar(threadIdx.x_2, (nullptr), "ThreadIndex", "threadIdx.x")] "thread_extent" = 64;
-              kernel.shared_1[(threadIdx.x_2 + 512)] = kernel_3[((((((floordiv(blockIdx.x, 7)*589824) + (floordiv((threadIdx.x_2 + 512), 24)*4608)) + cse_var_2) + (floordiv(floormod((threadIdx.x_2 + 8), 24), 3)*9)) + cse_var_1) + floormod((threadIdx.x_2 + 2), 3))]
-              attr [IterVar(threadIdx.x_2, (nullptr), "ThreadIndex", "threadIdx.x")] "thread_extent" = 64;
-              kernel.shared_1[(threadIdx.x_2 + 576)] = kernel_3[(((((((floordiv(blockIdx.x, 7)*589824) + (floordiv(threadIdx.x_2, 24)*4608)) + cse_var_2) + (floordiv(floormod(threadIdx.x_2, 24), 3)*9)) + cse_var_1) + floormod(threadIdx.x_2, 3)) + 110592)]
-              attr [IterVar(threadIdx.x_2, (nullptr), "ThreadIndex", "threadIdx.x")] "thread_extent" = 64;
-              kernel.shared_1[(threadIdx.x_2 + 640)] = kernel_3[((((((floordiv(blockIdx.x, 7)*589824) + (floordiv((threadIdx.x_2 + 640), 24)*4608)) + cse_var_2) + (floordiv(floormod((threadIdx.x_2 + 16), 24), 3)*9)) + cse_var_1) + floormod((threadIdx.x_2 + 1), 3))]
-              attr [IterVar(threadIdx.x_2, (nullptr), "ThreadIndex", "threadIdx.x")] "thread_extent" = 64;
-              kernel.shared_1[(threadIdx.x_2 + 704)] = kernel_3[((((((floordiv(blockIdx.x, 7)*589824) + (floordiv((threadIdx.x_2 + 704), 24)*4608)) + cse_var_2) + (floordiv(floormod((threadIdx.x_2 + 8), 24), 3)*9)) + cse_var_1) + floormod((threadIdx.x_2 + 2), 3))]
-              attr [IterVar(threadIdx.x_2, (nullptr), "ThreadIndex", "threadIdx.x")] "thread_extent" = 64;
-              kernel.shared_1[(threadIdx.x_2 + 768)] = kernel_3[(((((((floordiv(blockIdx.x, 7)*589824) + (floordiv(threadIdx.x_2, 24)*4608)) + cse_var_2) + (floordiv(floormod(threadIdx.x_2, 24), 3)*9)) + cse_var_1) + floormod(threadIdx.x_2, 3)) + 147456)]
-              attr [IterVar(threadIdx.x_2, (nullptr), "ThreadIndex", "threadIdx.x")] "thread_extent" = 64;
-              kernel.shared_1[(threadIdx.x_2 + 832)] = kernel_3[((((((floordiv(blockIdx.x, 7)*589824) + (floordiv((threadIdx.x_2 + 832), 24)*4608)) + cse_var_2) + (floordiv(floormod((threadIdx.x_2 + 16), 24), 3)*9)) + cse_var_1) + floormod((threadIdx.x_2 + 1), 3))]
-              attr [IterVar(threadIdx.x_2, (nullptr), "ThreadIndex", "threadIdx.x")] "thread_extent" = 64;
-              kernel.shared_1[(threadIdx.x_2 + 896)] = kernel_3[((((((floordiv(blockIdx.x, 7)*589824) + (floordiv((threadIdx.x_2 + 896), 24)*4608)) + cse_var_2) + (floordiv(floormod((threadIdx.x_2 + 8), 24), 3)*9)) + cse_var_1) + floormod((threadIdx.x_2 + 2), 3))]
-              attr [IterVar(threadIdx.x_2, (nullptr), "ThreadIndex", "threadIdx.x")] "thread_extent" = 64;
-              kernel.shared_1[(threadIdx.x_2 + 960)] = kernel_3[(((((((floordiv(blockIdx.x, 7)*589824) + (floordiv(threadIdx.x_2, 24)*4608)) + cse_var_2) + (floordiv(floormod(threadIdx.x_2, 24), 3)*9)) + cse_var_1) + floormod(threadIdx.x_2, 3)) + 184320)]
-              attr [IterVar(threadIdx.x_2, (nullptr), "ThreadIndex", "threadIdx.x")] "thread_extent" = 64;
-              kernel.shared_1[(threadIdx.x_2 + 1024)] = kernel_3[((((((floordiv(blockIdx.x, 7)*589824) + (floordiv((threadIdx.x_2 + 1024), 24)*4608)) + cse_var_2) + (floordiv(floormod((threadIdx.x_2 + 16), 24), 3)*9)) + cse_var_1) + floormod((threadIdx.x_2 + 1), 3))]
-              attr [IterVar(threadIdx.x_2, (nullptr), "ThreadIndex", "threadIdx.x")] "thread_extent" = 64;
-              kernel.shared_1[(threadIdx.x_2 + 1088)] = kernel_3[((((((floordiv(blockIdx.x, 7)*589824) + (floordiv((threadIdx.x_2 + 1088), 24)*4608)) + cse_var_2) + (floordiv(floormod((threadIdx.x_2 + 8), 24), 3)*9)) + cse_var_1) + floormod((threadIdx.x_2 + 2), 3))]
-              attr [IterVar(threadIdx.x_2, (nullptr), "ThreadIndex", "threadIdx.x")] "thread_extent" = 64;
-              kernel.shared_1[(threadIdx.x_2 + 1152)] = kernel_3[(((((((floordiv(blockIdx.x, 7)*589824) + (floordiv(threadIdx.x_2, 24)*4608)) + cse_var_2) + (floordiv(floormod(threadIdx.x_2, 24), 3)*9)) + cse_var_1) + floormod(threadIdx.x_2, 3)) + 221184)]
-              attr [IterVar(threadIdx.x_2, (nullptr), "ThreadIndex", "threadIdx.x")] "thread_extent" = 64;
-              kernel.shared_1[(threadIdx.x_2 + 1216)] = kernel_3[((((((floordiv(blockIdx.x, 7)*589824) + (floordiv((threadIdx.x_2 + 1216), 24)*4608)) + cse_var_2) + (floordiv(floormod((threadIdx.x_2 + 16), 24), 3)*9)) + cse_var_1) + floormod((threadIdx.x_2 + 1), 3))]
-              attr [IterVar(threadIdx.x_2, (nullptr), "ThreadIndex", "threadIdx.x")] "thread_extent" = 64;
-              kernel.shared_1[(threadIdx.x_2 + 1280)] = kernel_3[((((((floordiv(blockIdx.x, 7)*589824) + (floordiv((threadIdx.x_2 + 1280), 24)*4608)) + cse_var_2) + (floordiv(floormod((threadIdx.x_2 + 8), 24), 3)*9)) + cse_var_1) + floormod((threadIdx.x_2 + 2), 3))]
-              attr [IterVar(threadIdx.x_2, (nullptr), "ThreadIndex", "threadIdx.x")] "thread_extent" = 64;
-              kernel.shared_1[(threadIdx.x_2 + 1344)] = kernel_3[(((((((floordiv(blockIdx.x, 7)*589824) + (floordiv(threadIdx.x_2, 24)*4608)) + cse_var_2) + (floordiv(floormod(threadIdx.x_2, 24), 3)*9)) + cse_var_1) + floormod(threadIdx.x_2, 3)) + 258048)]
-              attr [IterVar(threadIdx.x_2, (nullptr), "ThreadIndex", "threadIdx.x")] "thread_extent" = 64;
-              kernel.shared_1[(threadIdx.x_2 + 1408)] = kernel_3[((((((floordiv(blockIdx.x, 7)*589824) + (floordiv((threadIdx.x_2 + 1408), 24)*4608)) + cse_var_2) + (floordiv(floormod((threadIdx.x_2 + 16), 24), 3)*9)) + cse_var_1) + floormod((threadIdx.x_2 + 1), 3))]
-              attr [IterVar(threadIdx.x_2, (nullptr), "ThreadIndex", "threadIdx.x")] "thread_extent" = 64;
-              kernel.shared_1[(threadIdx.x_2 + 1472)] = kernel_3[((((((floordiv(blockIdx.x, 7)*589824) + (floordiv((threadIdx.x_2 + 1472), 24)*4608)) + cse_var_2) + (floordiv(floormod((threadIdx.x_2 + 8), 24), 3)*9)) + cse_var_1) + floormod((threadIdx.x_2 + 2), 3))]
-              attr [IterVar(threadIdx.x_2, (nullptr), "ThreadIndex", "threadIdx.x")] "thread_extent" = 64;
-              kernel.shared_1[(threadIdx.x_2 + 1536)] = kernel_3[(((((((floordiv(blockIdx.x, 7)*589824) + (floordiv(threadIdx.x_2, 24)*4608)) + cse_var_2) + (floordiv(floormod(threadIdx.x_2, 24), 3)*9)) + cse_var_1) + floormod(threadIdx.x_2, 3)) + 294912)]
-              attr [IterVar(threadIdx.x_2, (nullptr), "ThreadIndex", "threadIdx.x")] "thread_extent" = 64;
-              kernel.shared_1[(threadIdx.x_2 + 1600)] = kernel_3[((((((floordiv(blockIdx.x, 7)*589824) + (floordiv((threadIdx.x_2 + 1600), 24)*4608)) + cse_var_2) + (floordiv(floormod((threadIdx.x_2 + 16), 24), 3)*9)) + cse_var_1) + floormod((threadIdx.x_2 + 1), 3))]
-              attr [IterVar(threadIdx.x_2, (nullptr), "ThreadIndex", "threadIdx.x")] "thread_extent" = 64;
-              kernel.shared_1[(threadIdx.x_2 + 1664)] = kernel_3[((((((floordiv(blockIdx.x, 7)*589824) + (floordiv((threadIdx.x_2 + 1664), 24)*4608)) + cse_var_2) + (floordiv(floormod((threadIdx.x_2 + 8), 24), 3)*9)) + cse_var_1) + floormod((threadIdx.x_2 + 2), 3))]
-              attr [IterVar(threadIdx.x_2, (nullptr), "ThreadIndex", "threadIdx.x")] "thread_extent" = 64;
-              kernel.shared_1[(threadIdx.x_2 + 1728)] = kernel_3[(((((((floordiv(blockIdx.x, 7)*589824) + (floordiv(threadIdx.x_2, 24)*4608)) + cse_var_2) + (floordiv(floormod(threadIdx.x_2, 24), 3)*9)) + cse_var_1) + floormod(threadIdx.x_2, 3)) + 331776)]
-              attr [IterVar(threadIdx.x_2, (nullptr), "ThreadIndex", "threadIdx.x")] "thread_extent" = 64;
-              kernel.shared_1[(threadIdx.x_2 + 1792)] = kernel_3[((((((floordiv(blockIdx.x, 7)*589824) + (floordiv((threadIdx.x_2 + 1792), 24)*4608)) + cse_var_2) + (floordiv(floormod((threadIdx.x_2 + 16), 24), 3)*9)) + cse_var_1) + floormod((threadIdx.x_2 + 1), 3))]
-              attr [IterVar(threadIdx.x_2, (nullptr), "ThreadIndex", "threadIdx.x")] "thread_extent" = 64;
-              kernel.shared_1[(threadIdx.x_2 + 1856)] = kernel_3[((((((floordiv(blockIdx.x, 7)*589824) + (floordiv((threadIdx.x_2 + 1856), 24)*4608)) + cse_var_2) + (floordiv(floormod((threadIdx.x_2 + 8), 24), 3)*9)) + cse_var_1) + floormod((threadIdx.x_2 + 2), 3))]
-              attr [IterVar(threadIdx.x_2, (nullptr), "ThreadIndex", "threadIdx.x")] "thread_extent" = 64;
-              kernel.shared_1[(threadIdx.x_2 + 1920)] = kernel_3[(((((((floordiv(blockIdx.x, 7)*589824) + (floordiv(threadIdx.x_2, 24)*4608)) + cse_var_2) + (floordiv(floormod(threadIdx.x_2, 24), 3)*9)) + cse_var_1) + floormod(threadIdx.x_2, 3)) + 368640)]
-              attr [IterVar(threadIdx.x_2, (nullptr), "ThreadIndex", "threadIdx.x")] "thread_extent" = 64;
-              kernel.shared_1[(threadIdx.x_2 + 1984)] = kernel_3[((((((floordiv(blockIdx.x, 7)*589824) + (floordiv((threadIdx.x_2 + 1984), 24)*4608)) + cse_var_2) + (floordiv(floormod((threadIdx.x_2 + 16), 24), 3)*9)) + cse_var_1) + floormod((threadIdx.x_2 + 1), 3))]
-              attr [IterVar(threadIdx.x_2, (nullptr), "ThreadIndex", "threadIdx.x")] "thread_extent" = 64;
-              kernel.shared_1[(threadIdx.x_2 + 2048)] = kernel_3[((((((floordiv(blockIdx.x, 7)*589824) + (floordiv((threadIdx.x_2 + 2048), 24)*4608)) + cse_var_2) + (floordiv(floormod((threadIdx.x_2 + 8), 24), 3)*9)) + cse_var_1) + floormod((threadIdx.x_2 + 2), 3))]
-              attr [IterVar(threadIdx.x_2, (nullptr), "ThreadIndex", "threadIdx.x")] "thread_extent" = 64;
-              kernel.shared_1[(threadIdx.x_2 + 2112)] = kernel_3[(((((((floordiv(blockIdx.x, 7)*589824) + (floordiv(threadIdx.x_2, 24)*4608)) + cse_var_2) + (floordiv(floormod(threadIdx.x_2, 24), 3)*9)) + cse_var_1) + floormod(threadIdx.x_2, 3)) + 405504)]
-              attr [IterVar(threadIdx.x_2, (nullptr), "ThreadIndex", "threadIdx.x")] "thread_extent" = 64;
-              kernel.shared_1[(threadIdx.x_2 + 2176)] = kernel_3[((((((floordiv(blockIdx.x, 7)*589824) + (floordiv((threadIdx.x_2 + 2176), 24)*4608)) + cse_var_2) + (floordiv(floormod((threadIdx.x_2 + 16), 24), 3)*9)) + cse_var_1) + floormod((threadIdx.x_2 + 1), 3))]
-              attr [IterVar(threadIdx.x_2, (nullptr), "ThreadIndex", "threadIdx.x")] "thread_extent" = 64;
-              kernel.shared_1[(threadIdx.x_2 + 2240)] = kernel_3[((((((floordiv(blockIdx.x, 7)*589824) + (floordiv((threadIdx.x_2 + 2240), 24)*4608)) + cse_var_2) + (floordiv(floormod((threadIdx.x_2 + 8), 24), 3)*9)) + cse_var_1) + floormod((threadIdx.x_2 + 2), 3))]
-              attr [IterVar(threadIdx.x_2, (nullptr), "ThreadIndex", "threadIdx.x")] "thread_extent" = 64;
-              kernel.shared_1[(threadIdx.x_2 + 2304)] = kernel_3[(((((((floordiv(blockIdx.x, 7)*589824) + (floordiv(threadIdx.x_2, 24)*4608)) + cse_var_2) + (floordiv(floormod(threadIdx.x_2, 24), 3)*9)) + cse_var_1) + floormod(threadIdx.x_2, 3)) + 442368)]
-              attr [IterVar(threadIdx.x_2, (nullptr), "ThreadIndex", "threadIdx.x")] "thread_extent" = 64;
-              kernel.shared_1[(threadIdx.x_2 + 2368)] = kernel_3[((((((floordiv(blockIdx.x, 7)*589824) + (floordiv((threadIdx.x_2 + 2368), 24)*4608)) + cse_var_2) + (floordiv(floormod((threadIdx.x_2 + 16), 24), 3)*9)) + cse_var_1) + floormod((threadIdx.x_2 + 1), 3))]
-              attr [IterVar(threadIdx.x_2, (nullptr), "ThreadIndex", "threadIdx.x")] "thread_extent" = 64;
-              kernel.shared_1[(threadIdx.x_2 + 2432)] = kernel_3[((((((floordiv(blockIdx.x, 7)*589824) + (floordiv((threadIdx.x_2 + 2432), 24)*4608)) + cse_var_2) + (floordiv(floormod((threadIdx.x_2 + 8), 24), 3)*9)) + cse_var_1) + floormod((threadIdx.x_2 + 2), 3))]
-              attr [IterVar(threadIdx.x_2, (nullptr), "ThreadIndex", "threadIdx.x")] "thread_extent" = 64;
-              kernel.shared_1[(threadIdx.x_2 + 2496)] = kernel_3[(((((((floordiv(blockIdx.x, 7)*589824) + (floordiv(threadIdx.x_2, 24)*4608)) + cse_var_2) + (floordiv(floormod(threadIdx.x_2, 24), 3)*9)) + cse_var_1) + floormod(threadIdx.x_2, 3)) + 479232)]
-              attr [IterVar(threadIdx.x_2, (nullptr), "ThreadIndex", "threadIdx.x")] "thread_extent" = 64;
-              kernel.shared_1[(threadIdx.x_2 + 2560)] = kernel_3[((((((floordiv(blockIdx.x, 7)*589824) + (floordiv((threadIdx.x_2 + 2560), 24)*4608)) + cse_var_2) + (floordiv(floormod((threadIdx.x_2 + 16), 24), 3)*9)) + cse_var_1) + floormod((threadIdx.x_2 + 1), 3))]
-              attr [IterVar(threadIdx.x_2, (nullptr), "ThreadIndex", "threadIdx.x")] "thread_extent" = 64;
-              kernel.shared_1[(threadIdx.x_2 + 2624)] = kernel_3[((((((floordiv(blockIdx.x, 7)*589824) + (floordiv((threadIdx.x_2 + 2624), 24)*4608)) + cse_var_2) + (floordiv(floormod((threadIdx.x_2 + 8), 24), 3)*9)) + cse_var_1) + floormod((threadIdx.x_2 + 2), 3))]
-              attr [IterVar(threadIdx.x_2, (nullptr), "ThreadIndex", "threadIdx.x")] "thread_extent" = 64;
-              kernel.shared_1[(threadIdx.x_2 + 2688)] = kernel_3[(((((((floordiv(blockIdx.x, 7)*589824) + (floordiv(threadIdx.x_2, 24)*4608)) + cse_var_2) + (floordiv(floormod(threadIdx.x_2, 24), 3)*9)) + cse_var_1) + floormod(threadIdx.x_2, 3)) + 516096)]
-              attr [IterVar(threadIdx.x_2, (nullptr), "ThreadIndex", "threadIdx.x")] "thread_extent" = 64;
-              kernel.shared_1[(threadIdx.x_2 + 2752)] = kernel_3[((((((floordiv(blockIdx.x, 7)*589824) + (floordiv((threadIdx.x_2 + 2752), 24)*4608)) + cse_var_2) + (floordiv(floormod((threadIdx.x_2 + 16), 24), 3)*9)) + cse_var_1) + floormod((threadIdx.x_2 + 1), 3))]
-              attr [IterVar(threadIdx.x_2, (nullptr), "ThreadIndex", "threadIdx.x")] "thread_extent" = 64;
-              kernel.shared_1[(threadIdx.x_2 + 2816)] = kernel_3[((((((floordiv(blockIdx.x, 7)*589824) + (floordiv((threadIdx.x_2 + 2816), 24)*4608)) + cse_var_2) + (floordiv(floormod((threadIdx.x_2 + 8), 24), 3)*9)) + cse_var_1) + floormod((threadIdx.x_2 + 2), 3))]
-              attr [IterVar(threadIdx.x_2, (nullptr), "ThreadIndex", "threadIdx.x")] "thread_extent" = 64;
-              kernel.shared_1[(threadIdx.x_2 + 2880)] = kernel_3[(((((((floordiv(blockIdx.x, 7)*589824) + (floordiv(threadIdx.x_2, 24)*4608)) + cse_var_2) + (floordiv(floormod(threadIdx.x_2, 24), 3)*9)) + cse_var_1) + floormod(threadIdx.x_2, 3)) + 552960)]
-              attr [IterVar(threadIdx.x_2, (nullptr), "ThreadIndex", "threadIdx.x")] "thread_extent" = 64;
-              kernel.shared_1[(threadIdx.x_2 + 2944)] = kernel_3[((((((floordiv(blockIdx.x, 7)*589824) + (floordiv((threadIdx.x_2 + 2944), 24)*4608)) + cse_var_2) + (floordiv(floormod((threadIdx.x_2 + 16), 24), 3)*9)) + cse_var_1) + floormod((threadIdx.x_2 + 1), 3))]
-              attr [IterVar(threadIdx.x_2, (nullptr), "ThreadIndex", "threadIdx.x")] "thread_extent" = 64;
-              kernel.shared_1[(threadIdx.x_2 + 3008)] = kernel_3[((((((floordiv(blockIdx.x, 7)*589824) + (floordiv((threadIdx.x_2 + 3008), 24)*4608)) + cse_var_2) + (floordiv(floormod((threadIdx.x_2 + 8), 24), 3)*9)) + cse_var_1) + floormod((threadIdx.x_2 + 2), 3))]
-              conv2d_nchw_1[0] = (conv2d_nchw_1[0] + (pad_temp.shared_1[0]*kernel.shared_1[(threadIdx.x*48)]))
-              conv2d_nchw_1[0] = (conv2d_nchw_1[0] + (pad_temp.shared_1[9]*kernel.shared_1[((threadIdx.x*48) + 3)]))
-              conv2d_nchw_1[1] = (conv2d_nchw_1[1] + (pad_temp.shared_1[1]*kernel.shared_1[(threadIdx.x*48)]))
-              conv2d_nchw_1[1] = (conv2d_nchw_1[1] + (pad_temp.shared_1[10]*kernel.shared_1[((threadIdx.x*48) + 3)]))
-              conv2d_nchw_1[2] = (conv2d_nchw_1[2] + (pad_temp.shared_1[2]*kernel.shared_1[(threadIdx.x*48)]))
-              conv2d_nchw_1[2] = (conv2d_nchw_1[2] + (pad_temp.shared_1[11]*kernel.shared_1[((threadIdx.x*48) + 3)]))
-              conv2d_nchw_1[3] = (conv2d_nchw_1[3] + (pad_temp.shared_1[3]*kernel.shared_1[(threadIdx.x*48)]))
-              conv2d_nchw_1[3] = (conv2d_nchw_1[3] + (pad_temp.shared_1[12]*kernel.shared_1[((threadIdx.x*48) + 3)]))
-              conv2d_nchw_1[4] = (conv2d_nchw_1[4] + (pad_temp.shared_1[4]*kernel.shared_1[(threadIdx.x*48)]))
-              conv2d_nchw_1[4] = (conv2d_nchw_1[4] + (pad_temp.shared_1[13]*kernel.shared_1[((threadIdx.x*48) + 3)]))
-              conv2d_nchw_1[5] = (conv2d_nchw_1[5] + (pad_temp.shared_1[5]*kernel.shared_1[(threadIdx.x*48)]))
-              conv2d_nchw_1[5] = (conv2d_nchw_1[5] + (pad_temp.shared_1[14]*kernel.shared_1[((threadIdx.x*48) + 3)]))
-              conv2d_nchw_1[6] = (conv2d_nchw_1[6] + (pad_temp.shared_1[6]*kernel.shared_1[(threadIdx.x*48)]))
-              conv2d_nchw_1[6] = (conv2d_nchw_1[6] + (pad_temp.shared_1[15]*kernel.shared_1[((threadIdx.x*48) + 3)]))
-              conv2d_nchw_1[7] = (conv2d_nchw_1[7] + (pad_temp.shared_1[0]*kernel.shared_1[((threadIdx.x*48) + 24)]))
-              conv2d_nchw_1[7] = (conv2d_nchw_1[7] + (pad_temp.shared_1[9]*kernel.shared_1[((threadIdx.x*48) + 27)]))
-              conv2d_nchw_1[8] = (conv2d_nchw_1[8] + (pad_temp.shared_1[1]*kernel.shared_1[((threadIdx.x*48) + 24)]))
-              conv2d_nchw_1[8] = (conv2d_nchw_1[8] + (pad_temp.shared_1[10]*kernel.shared_1[((threadIdx.x*48) + 27)]))
-              conv2d_nchw_1[9] = (conv2d_nchw_1[9] + (pad_temp.shared_1[2]*kernel.shared_1[((threadIdx.x*48) + 24)]))
-              conv2d_nchw_1[9] = (conv2d_nchw_1[9] + (pad_temp.shared_1[11]*kernel.shared_1[((threadIdx.x*48) + 27)]))
-              conv2d_nchw_1[10] = (conv2d_nchw_1[10] + (pad_temp.shared_1[3]*kernel.shared_1[((threadIdx.x*48) + 24)]))
-              conv2d_nchw_1[10] = (conv2d_nchw_1[10] + (pad_temp.shared_1[12]*kernel.shared_1[((threadIdx.x*48) + 27)]))
-              conv2d_nchw_1[11] = (conv2d_nchw_1[11] + (pad_temp.shared_1[4]*kernel.shared_1[((threadIdx.x*48) + 24)]))
-              conv2d_nchw_1[11] = (conv2d_nchw_1[11] + (pad_temp.shared_1[13]*kernel.shared_1[((threadIdx.x*48) + 27)]))
-              conv2d_nchw_1[12] = (conv2d_nchw_1[12] + (pad_temp.shared_1[5]*kernel.shared_1[((threadIdx.x*48) + 24)]))
-              conv2d_nchw_1[12] = (conv2d_nchw_1[12] + (pad_temp.shared_1[14]*kernel.shared_1[((threadIdx.x*48) + 27)]))
-              conv2d_nchw_1[13] = (conv2d_nchw_1[13] + (pad_temp.shared_1[6]*kernel.shared_1[((threadIdx.x*48) + 24)]))
-              conv2d_nchw_1[13] = (conv2d_nchw_1[13] + (pad_temp.shared_1[15]*kernel.shared_1[((threadIdx.x*48) + 27)]))
-              conv2d_nchw_1[0] = (conv2d_nchw_1[0] + (pad_temp.shared_1[1]*kernel.shared_1[((threadIdx.x*48) + 1)]))
-              conv2d_nchw_1[0] = (conv2d_nchw_1[0] + (pad_temp.shared_1[10]*kernel.shared_1[((threadIdx.x*48) + 4)]))
-              conv2d_nchw_1[1] = (conv2d_nchw_1[1] + (pad_temp.shared_1[2]*kernel.shared_1[((threadIdx.x*48) + 1)]))
-              conv2d_nchw_1[1] = (conv2d_nchw_1[1] + (pad_temp.shared_1[11]*kernel.shared_1[((threadIdx.x*48) + 4)]))
-              conv2d_nchw_1[2] = (conv2d_nchw_1[2] + (pad_temp.shared_1[3]*kernel.shared_1[((threadIdx.x*48) + 1)]))
-              conv2d_nchw_1[2] = (conv2d_nchw_1[2] + (pad_temp.shared_1[12]*kernel.shared_1[((threadIdx.x*48) + 4)]))
-              conv2d_nchw_1[3] = (conv2d_nchw_1[3] + (pad_temp.shared_1[4]*kernel.shared_1[((threadIdx.x*48) + 1)]))
-              conv2d_nchw_1[3] = (conv2d_nchw_1[3] + (pad_temp.shared_1[13]*kernel.shared_1[((threadIdx.x*48) + 4)]))
-              conv2d_nchw_1[4] = (conv2d_nchw_1[4] + (pad_temp.shared_1[5]*kernel.shared_1[((threadIdx.x*48) + 1)]))
-              conv2d_nchw_1[4] = (conv2d_nchw_1[4] + (pad_temp.shared_1[14]*kernel.shared_1[((threadIdx.x*48) + 4)]))
-              conv2d_nchw_1[5] = (conv2d_nchw_1[5] + (pad_temp.shared_1[6]*kernel.shared_1[((threadIdx.x*48) + 1)]))
-              conv2d_nchw_1[5] = (conv2d_nchw_1[5] + (pad_temp.shared_1[15]*kernel.shared_1[((threadIdx.x*48) + 4)]))
-              conv2d_nchw_1[6] = (conv2d_nchw_1[6] + (pad_temp.shared_1[7]*kernel.shared_1[((threadIdx.x*48) + 1)]))
-              conv2d_nchw_1[6] = (conv2d_nchw_1[6] + (pad_temp.shared_1[16]*kernel.shared_1[((threadIdx.x*48) + 4)]))
-              conv2d_nchw_1[7] = (conv2d_nchw_1[7] + (pad_temp.shared_1[1]*kernel.shared_1[((threadIdx.x*48) + 25)]))
-              conv2d_nchw_1[7] = (conv2d_nchw_1[7] + (pad_temp.shared_1[10]*kernel.shared_1[((threadIdx.x*48) + 28)]))
-              conv2d_nchw_1[8] = (conv2d_nchw_1[8] + (pad_temp.shared_1[2]*kernel.shared_1[((threadIdx.x*48) + 25)]))
-              conv2d_nchw_1[8] = (conv2d_nchw_1[8] + (pad_temp.shared_1[11]*kernel.shared_1[((threadIdx.x*48) + 28)]))
-              conv2d_nchw_1[9] = (conv2d_nchw_1[9] + (pad_temp.shared_1[3]*kernel.shared_1[((threadIdx.x*48) + 25)]))
-              conv2d_nchw_1[9] = (conv2d_nchw_1[9] + (pad_temp.shared_1[12]*kernel.shared_1[((threadIdx.x*48) + 28)]))
-              conv2d_nchw_1[10] = (conv2d_nchw_1[10] + (pad_temp.shared_1[4]*kernel.shared_1[((threadIdx.x*48) + 25)]))
-              conv2d_nchw_1[10] = (conv2d_nchw_1[10] + (pad_temp.shared_1[13]*kernel.shared_1[((threadIdx.x*48) + 28)]))
-              conv2d_nchw_1[11] = (conv2d_nchw_1[11] + (pad_temp.shared_1[5]*kernel.shared_1[((threadIdx.x*48) + 25)]))
-              conv2d_nchw_1[11] = (conv2d_nchw_1[11] + (pad_temp.shared_1[14]*kernel.shared_1[((threadIdx.x*48) + 28)]))
-              conv2d_nchw_1[12] = (conv2d_nchw_1[12] + (pad_temp.shared_1[6]*kernel.shared_1[((threadIdx.x*48) + 25)]))
-              conv2d_nchw_1[12] = (conv2d_nchw_1[12] + (pad_temp.shared_1[15]*kernel.shared_1[((threadIdx.x*48) + 28)]))
-              conv2d_nchw_1[13] = (conv2d_nchw_1[13] + (pad_temp.shared_1[7]*kernel.shared_1[((threadIdx.x*48) + 25)]))
-              conv2d_nchw_1[13] = (conv2d_nchw_1[13] + (pad_temp.shared_1[16]*kernel.shared_1[((threadIdx.x*48) + 28)]))
-              conv2d_nchw_1[0] = (conv2d_nchw_1[0] + (pad_temp.shared_1[2]*kernel.shared_1[((threadIdx.x*48) + 2)]))
-              conv2d_nchw_1[0] = (conv2d_nchw_1[0] + (pad_temp.shared_1[11]*kernel.shared_1[((threadIdx.x*48) + 5)]))
-              conv2d_nchw_1[1] = (conv2d_nchw_1[1] + (pad_temp.shared_1[3]*kernel.shared_1[((threadIdx.x*48) + 2)]))
-              conv2d_nchw_1[1] = (conv2d_nchw_1[1] + (pad_temp.shared_1[12]*kernel.shared_1[((threadIdx.x*48) + 5)]))
-              conv2d_nchw_1[2] = (conv2d_nchw_1[2] + (pad_temp.shared_1[4]*kernel.shared_1[((threadIdx.x*48) + 2)]))
-              conv2d_nchw_1[2] = (conv2d_nchw_1[2] + (pad_temp.shared_1[13]*kernel.shared_1[((threadIdx.x*48) + 5)]))
-              conv2d_nchw_1[3] = (conv2d_nchw_1[3] + (pad_temp.shared_1[5]*kernel.shared_1[((threadIdx.x*48) + 2)]))
-              conv2d_nchw_1[3] = (conv2d_nchw_1[3] + (pad_temp.shared_1[14]*kernel.shared_1[((threadIdx.x*48) + 5)]))
-              conv2d_nchw_1[4] = (conv2d_nchw_1[4] + (pad_temp.shared_1[6]*kernel.shared_1[((threadIdx.x*48) + 2)]))
-              conv2d_nchw_1[4] = (conv2d_nchw_1[4] + (pad_temp.shared_1[15]*kernel.shared_1[((threadIdx.x*48) + 5)]))
-              conv2d_nchw_1[5] = (conv2d_nchw_1[5] + (pad_temp.shared_1[7]*kernel.shared_1[((threadIdx.x*48) + 2)]))
-              conv2d_nchw_1[5] = (conv2d_nchw_1[5] + (pad_temp.shared_1[16]*kernel.shared_1[((threadIdx.x*48) + 5)]))
-              conv2d_nchw_1[6] = (conv2d_nchw_1[6] + (pad_temp.shared_1[8]*kernel.shared_1[((threadIdx.x*48) + 2)]))
-              conv2d_nchw_1[6] = (conv2d_nchw_1[6] + (pad_temp.shared_1[17]*kernel.shared_1[((threadIdx.x*48) + 5)]))
-              conv2d_nchw_1[7] = (conv2d_nchw_1[7] + (pad_temp.shared_1[2]*kernel.shared_1[((threadIdx.x*48) + 26)]))
-              conv2d_nchw_1[7] = (conv2d_nchw_1[7] + (pad_temp.shared_1[11]*kernel.shared_1[((threadIdx.x*48) + 29)]))
-              conv2d_nchw_1[8] = (conv2d_nchw_1[8] + (pad_temp.shared_1[3]*kernel.shared_1[((threadIdx.x*48) + 26)]))
-              conv2d_nchw_1[8] = (conv2d_nchw_1[8] + (pad_temp.shared_1[12]*kernel.shared_1[((threadIdx.x*48) + 29)]))
-              conv2d_nchw_1[9] = (conv2d_nchw_1[9] + (pad_temp.shared_1[4]*kernel.shared_1[((threadIdx.x*48) + 26)]))
-              conv2d_nchw_1[9] = (conv2d_nchw_1[9] + (pad_temp.shared_1[13]*kernel.shared_1[((threadIdx.x*48) + 29)]))
-              conv2d_nchw_1[10] = (conv2d_nchw_1[10] + (pad_temp.shared_1[5]*kernel.shared_1[((threadIdx.x*48) + 26)]))
-              conv2d_nchw_1[10] = (conv2d_nchw_1[10] + (pad_temp.shared_1[14]*kernel.shared_1[((threadIdx.x*48) + 29)]))
-              conv2d_nchw_1[11] = (conv2d_nchw_1[11] + (pad_temp.shared_1[6]*kernel.shared_1[((threadIdx.x*48) + 26)]))
-              conv2d_nchw_1[11] = (conv2d_nchw_1[11] + (pad_temp.shared_1[15]*kernel.shared_1[((threadIdx.x*48) + 29)]))
-              conv2d_nchw_1[12] = (conv2d_nchw_1[12] + (pad_temp.shared_1[7]*kernel.shared_1[((threadIdx.x*48) + 26)]))
-              conv2d_nchw_1[12] = (conv2d_nchw_1[12] + (pad_temp.shared_1[16]*kernel.shared_1[((threadIdx.x*48) + 29)]))
-              conv2d_nchw_1[13] = (conv2d_nchw_1[13] + (pad_temp.shared_1[8]*kernel.shared_1[((threadIdx.x*48) + 26)]))
-              conv2d_nchw_1[13] = (conv2d_nchw_1[13] + (pad_temp.shared_1[17]*kernel.shared_1[((threadIdx.x*48) + 29)]))
-              conv2d_nchw_1[0] = (conv2d_nchw_1[0] + (pad_temp.shared_1[18]*kernel.shared_1[((threadIdx.x*48) + 6)]))
-              conv2d_nchw_1[0] = (conv2d_nchw_1[0] + (pad_temp.shared_1[27]*kernel.shared_1[((threadIdx.x*48) + 9)]))
-              conv2d_nchw_1[1] = (conv2d_nchw_1[1] + (pad_temp.shared_1[19]*kernel.shared_1[((threadIdx.x*48) + 6)]))
-              conv2d_nchw_1[1] = (conv2d_nchw_1[1] + (pad_temp.shared_1[28]*kernel.shared_1[((threadIdx.x*48) + 9)]))
-              conv2d_nchw_1[2] = (conv2d_nchw_1[2] + (pad_temp.shared_1[20]*kernel.shared_1[((threadIdx.x*48) + 6)]))
-              conv2d_nchw_1[2] = (conv2d_nchw_1[2] + (pad_temp.shared_1[29]*kernel.shared_1[((threadIdx.x*48) + 9)]))
-              conv2d_nchw_1[3] = (conv2d_nchw_1[3] + (pad_temp.shared_1[21]*kernel.shared_1[((threadIdx.x*48) + 6)]))
-              conv2d_nchw_1[3] = (conv2d_nchw_1[3] + (pad_temp.shared_1[30]*kernel.shared_1[((threadIdx.x*48) + 9)]))
-              conv2d_nchw_1[4] = (conv2d_nchw_1[4] + (pad_temp.shared_1[22]*kernel.shared_1[((threadIdx.x*48) + 6)]))
-              conv2d_nchw_1[4] = (conv2d_nchw_1[4] + (pad_temp.shared_1[31]*kernel.shared_1[((threadIdx.x*48) + 9)]))
-              conv2d_nchw_1[5] = (conv2d_nchw_1[5] + (pad_temp.shared_1[23]*kernel.shared_1[((threadIdx.x*48) + 6)]))
-              conv2d_nchw_1[5] = (conv2d_nchw_1[5] + (pad_temp.shared_1[32]*kernel.shared_1[((threadIdx.x*48) + 9)]))
-              conv2d_nchw_1[6] = (conv2d_nchw_1[6] + (pad_temp.shared_1[24]*kernel.shared_1[((threadIdx.x*48) + 6)]))
-              conv2d_nchw_1[6] = (conv2d_nchw_1[6] + (pad_temp.shared_1[33]*kernel.shared_1[((threadIdx.x*48) + 9)]))
-              conv2d_nchw_1[7] = (conv2d_nchw_1[7] + (pad_temp.shared_1[18]*kernel.shared_1[((threadIdx.x*48) + 30)]))
-              conv2d_nchw_1[7] = (conv2d_nchw_1[7] + (pad_temp.shared_1[27]*kernel.shared_1[((threadIdx.x*48) + 33)]))
-              conv2d_nchw_1[8] = (conv2d_nchw_1[8] + (pad_temp.shared_1[19]*kernel.shared_1[((threadIdx.x*48) + 30)]))
-              conv2d_nchw_1[8] = (conv2d_nchw_1[8] + (pad_temp.shared_1[28]*kernel.shared_1[((threadIdx.x*48) + 33)]))
-              conv2d_nchw_1[9] = (conv2d_nchw_1[9] + (pad_temp.shared_1[20]*kernel.shared_1[((threadIdx.x*48) + 30)]))
-              conv2d_nchw_1[9] = (conv2d_nchw_1[9] + (pad_temp.shared_1[29]*kernel.shared_1[((threadIdx.x*48) + 33)]))
-              conv2d_nchw_1[10] = (conv2d_nchw_1[10] + (pad_temp.shared_1[21]*kernel.shared_1[((threadIdx.x*48) + 30)]))
-              conv2d_nchw_1[10] = (conv2d_nchw_1[10] + (pad_temp.shared_1[30]*kernel.shared_1[((threadIdx.x*48) + 33)]))
-              conv2d_nchw_1[11] = (conv2d_nchw_1[11] + (pad_temp.shared_1[22]*kernel.shared_1[((threadIdx.x*48) + 30)]))
-              conv2d_nchw_1[11] = (conv2d_nchw_1[11] + (pad_temp.shared_1[31]*kernel.shared_1[((threadIdx.x*48) + 33)]))
-              conv2d_nchw_1[12] = (conv2d_nchw_1[12] + (pad_temp.shared_1[23]*kernel.shared_1[((threadIdx.x*48) + 30)]))
-              conv2d_nchw_1[12] = (conv2d_nchw_1[12] + (pad_temp.shared_1[32]*kernel.shared_1[((threadIdx.x*48) + 33)]))
-              conv2d_nchw_1[13] = (conv2d_nchw_1[13] + (pad_temp.shared_1[24]*kernel.shared_1[((threadIdx.x*48) + 30)]))
-              conv2d_nchw_1[13] = (conv2d_nchw_1[13] + (pad_temp.shared_1[33]*kernel.shared_1[((threadIdx.x*48) + 33)]))
-              conv2d_nchw_1[0] = (conv2d_nchw_1[0] + (pad_temp.shared_1[19]*kernel.shared_1[((threadIdx.x*48) + 7)]))
-              conv2d_nchw_1[0] = (conv2d_nchw_1[0] + (pad_temp.shared_1[28]*kernel.shared_1[((threadIdx.x*48) + 10)]))
-              conv2d_nchw_1[1] = (conv2d_nchw_1[1] + (pad_temp.shared_1[20]*kernel.shared_1[((threadIdx.x*48) + 7)]))
-              conv2d_nchw_1[1] = (conv2d_nchw_1[1] + (pad_temp.shared_1[29]*kernel.shared_1[((threadIdx.x*48) + 10)]))
-              conv2d_nchw_1[2] = (conv2d_nchw_1[2] + (pad_temp.shared_1[21]*kernel.shared_1[((threadIdx.x*48) + 7)]))
-              conv2d_nchw_1[2] = (conv2d_nchw_1[2] + (pad_temp.shared_1[30]*kernel.shared_1[((threadIdx.x*48) + 10)]))
-              conv2d_nchw_1[3] = (conv2d_nchw_1[3] + (pad_temp.shared_1[22]*kernel.shared_1[((threadIdx.x*48) + 7)]))
-              conv2d_nchw_1[3] = (conv2d_nchw_1[3] + (pad_temp.shared_1[31]*kernel.shared_1[((threadIdx.x*48) + 10)]))
-              conv2d_nchw_1[4] = (conv2d_nchw_1[4] + (pad_temp.shared_1[23]*kernel.shared_1[((threadIdx.x*48) + 7)]))
-              conv2d_nchw_1[4] = (conv2d_nchw_1[4] + (pad_temp.shared_1[32]*kernel.shared_1[((threadIdx.x*48) + 10)]))
-              conv2d_nchw_1[5] = (conv2d_nchw_1[5] + (pad_temp.shared_1[24]*kernel.shared_1[((threadIdx.x*48) + 7)]))
-              conv2d_nchw_1[5] = (conv2d_nchw_1[5] + (pad_temp.shared_1[33]*kernel.shared_1[((threadIdx.x*48) + 10)]))
-              conv2d_nchw_1[6] = (conv2d_nchw_1[6] + (pad_temp.shared_1[25]*kernel.shared_1[((threadIdx.x*48) + 7)]))
-              conv2d_nchw_1[6] = (conv2d_nchw_1[6] + (pad_temp.shared_1[34]*kernel.shared_1[((threadIdx.x*48) + 10)]))
-              conv2d_nchw_1[7] = (conv2d_nchw_1[7] + (pad_temp.shared_1[19]*kernel.shared_1[((threadIdx.x*48) + 31)]))
-              conv2d_nchw_1[7] = (conv2d_nchw_1[7] + (pad_temp.shared_1[28]*kernel.shared_1[((threadIdx.x*48) + 34)]))
-              conv2d_nchw_1[8] = (conv2d_nchw_1[8] + (pad_temp.shared_1[20]*kernel.shared_1[((threadIdx.x*48) + 31)]))
-              conv2d_nchw_1[8] = (conv2d_nchw_1[8] + (pad_temp.shared_1[29]*kernel.shared_1[((threadIdx.x*48) + 34)]))
-              conv2d_nchw_1[9] = (conv2d_nchw_1[9] + (pad_temp.shared_1[21]*kernel.shared_1[((threadIdx.x*48) + 31)]))
-              conv2d_nchw_1[9] = (conv2d_nchw_1[9] + (pad_temp.shared_1[30]*kernel.shared_1[((threadIdx.x*48) + 34)]))
-              conv2d_nchw_1[10] = (conv2d_nchw_1[10] + (pad_temp.shared_1[22]*kernel.shared_1[((threadIdx.x*48) + 31)]))
-              conv2d_nchw_1[10] = (conv2d_nchw_1[10] + (pad_temp.shared_1[31]*kernel.shared_1[((threadIdx.x*48) + 34)]))
-              conv2d_nchw_1[11] = (conv2d_nchw_1[11] + (pad_temp.shared_1[23]*kernel.shared_1[((threadIdx.x*48) + 31)]))
-              conv2d_nchw_1[11] = (conv2d_nchw_1[11] + (pad_temp.shared_1[32]*kernel.shared_1[((threadIdx.x*48) + 34)]))
-              conv2d_nchw_1[12] = (conv2d_nchw_1[12] + (pad_temp.shared_1[24]*kernel.shared_1[((threadIdx.x*48) + 31)]))
-              conv2d_nchw_1[12] = (conv2d_nchw_1[12] + (pad_temp.shared_1[33]*kernel.shared_1[((threadIdx.x*48) + 34)]))
-              conv2d_nchw_1[13] = (conv2d_nchw_1[13] + (pad_temp.shared_1[25]*kernel.shared_1[((threadIdx.x*48) + 31)]))
-              conv2d_nchw_1[13] = (conv2d_nchw_1[13] + (pad_temp.shared_1[34]*kernel.shared_1[((threadIdx.x*48) + 34)]))
-              conv2d_nchw_1[0] = (conv2d_nchw_1[0] + (pad_temp.shared_1[20]*kernel.shared_1[((threadIdx.x*48) + 8)]))
-              conv2d_nchw_1[0] = (conv2d_nchw_1[0] + (pad_temp.shared_1[29]*kernel.shared_1[((threadIdx.x*48) + 11)]))
-              conv2d_nchw_1[1] = (conv2d_nchw_1[1] + (pad_temp.shared_1[21]*kernel.shared_1[((threadIdx.x*48) + 8)]))
-              conv2d_nchw_1[1] = (conv2d_nchw_1[1] + (pad_temp.shared_1[30]*kernel.shared_1[((threadIdx.x*48) + 11)]))
-              conv2d_nchw_1[2] = (conv2d_nchw_1[2] + (pad_temp.shared_1[22]*kernel.shared_1[((threadIdx.x*48) + 8)]))
-              conv2d_nchw_1[2] = (conv2d_nchw_1[2] + (pad_temp.shared_1[31]*kernel.shared_1[((threadIdx.x*48) + 11)]))
-              conv2d_nchw_1[3] = (conv2d_nchw_1[3] + (pad_temp.shared_1[23]*kernel.shared_1[((threadIdx.x*48) + 8)]))
-              conv2d_nchw_1[3] = (conv2d_nchw_1[3] + (pad_temp.shared_1[32]*kernel.shared_1[((threadIdx.x*48) + 11)]))
-              conv2d_nchw_1[4] = (conv2d_nchw_1[4] + (pad_temp.shared_1[24]*kernel.shared_1[((threadIdx.x*48) + 8)]))
-              conv2d_nchw_1[4] = (conv2d_nchw_1[4] + (pad_temp.shared_1[33]*kernel.shared_1[((threadIdx.x*48) + 11)]))
-              conv2d_nchw_1[5] = (conv2d_nchw_1[5] + (pad_temp.shared_1[25]*kernel.shared_1[((threadIdx.x*48) + 8)]))
-              conv2d_nchw_1[5] = (conv2d_nchw_1[5] + (pad_temp.shared_1[34]*kernel.shared_1[((threadIdx.x*48) + 11)]))
-              conv2d_nchw_1[6] = (conv2d_nchw_1[6] + (pad_temp.shared_1[26]*kernel.shared_1[((threadIdx.x*48) + 8)]))
-              conv2d_nchw_1[6] = (conv2d_nchw_1[6] + (pad_temp.shared_1[35]*kernel.shared_1[((threadIdx.x*48) + 11)]))
-              conv2d_nchw_1[7] = (conv2d_nchw_1[7] + (pad_temp.shared_1[20]*kernel.shared_1[((threadIdx.x*48) + 32)]))
-              conv2d_nchw_1[7] = (conv2d_nchw_1[7] + (pad_temp.shared_1[29]*kernel.shared_1[((threadIdx.x*48) + 35)]))
-              conv2d_nchw_1[8] = (conv2d_nchw_1[8] + (pad_temp.shared_1[21]*kernel.shared_1[((threadIdx.x*48) + 32)]))
-              conv2d_nchw_1[8] = (conv2d_nchw_1[8] + (pad_temp.shared_1[30]*kernel.shared_1[((threadIdx.x*48) + 35)]))
-              conv2d_nchw_1[9] = (conv2d_nchw_1[9] + (pad_temp.shared_1[22]*kernel.shared_1[((threadIdx.x*48) + 32)]))
-              conv2d_nchw_1[9] = (conv2d_nchw_1[9] + (pad_temp.shared_1[31]*kernel.shared_1[((threadIdx.x*48) + 35)]))
-              conv2d_nchw_1[10] = (conv2d_nchw_1[10] + (pad_temp.shared_1[23]*kernel.shared_1[((threadIdx.x*48) + 32)]))
-              conv2d_nchw_1[10] = (conv2d_nchw_1[10] + (pad_temp.shared_1[32]*kernel.shared_1[((threadIdx.x*48) + 35)]))
-              conv2d_nchw_1[11] = (conv2d_nchw_1[11] + (pad_temp.shared_1[24]*kernel.shared_1[((threadIdx.x*48) + 32)]))
-              conv2d_nchw_1[11] = (conv2d_nchw_1[11] + (pad_temp.shared_1[33]*kernel.shared_1[((threadIdx.x*48) + 35)]))
-              conv2d_nchw_1[12] = (conv2d_nchw_1[12] + (pad_temp.shared_1[25]*kernel.shared_1[((threadIdx.x*48) + 32)]))
-              conv2d_nchw_1[12] = (conv2d_nchw_1[12] + (pad_temp.shared_1[34]*kernel.shared_1[((threadIdx.x*48) + 35)]))
-              conv2d_nchw_1[13] = (conv2d_nchw_1[13] + (pad_temp.shared_1[26]*kernel.shared_1[((threadIdx.x*48) + 32)]))
-              conv2d_nchw_1[13] = (conv2d_nchw_1[13] + (pad_temp.shared_1[35]*kernel.shared_1[((threadIdx.x*48) + 35)]))
-              conv2d_nchw_1[0] = (conv2d_nchw_1[0] + (pad_temp.shared_1[36]*kernel.shared_1[((threadIdx.x*48) + 12)]))
-              conv2d_nchw_1[0] = (conv2d_nchw_1[0] + (pad_temp.shared_1[45]*kernel.shared_1[((threadIdx.x*48) + 15)]))
-              conv2d_nchw_1[1] = (conv2d_nchw_1[1] + (pad_temp.shared_1[37]*kernel.shared_1[((threadIdx.x*48) + 12)]))
-              conv2d_nchw_1[1] = (conv2d_nchw_1[1] + (pad_temp.shared_1[46]*kernel.shared_1[((threadIdx.x*48) + 15)]))
-              conv2d_nchw_1[2] = (conv2d_nchw_1[2] + (pad_temp.shared_1[38]*kernel.shared_1[((threadIdx.x*48) + 12)]))
-              conv2d_nchw_1[2] = (conv2d_nchw_1[2] + (pad_temp.shared_1[47]*kernel.shared_1[((threadIdx.x*48) + 15)]))
-              conv2d_nchw_1[3] = (conv2d_nchw_1[3] + (pad_temp.shared_1[39]*kernel.shared_1[((threadIdx.x*48) + 12)]))
-              conv2d_nchw_1[3] = (conv2d_nchw_1[3] + (pad_temp.shared_1[48]*kernel.shared_1[((threadIdx.x*48) + 15)]))
-              conv2d_nchw_1[4] = (conv2d_nchw_1[4] + (pad_temp.shared_1[40]*kernel.shared_1[((threadIdx.x*48) + 12)]))
-              conv2d_nchw_1[4] = (conv2d_nchw_1[4] + (pad_temp.shared_1[49]*kernel.shared_1[((threadIdx.x*48) + 15)]))
-              conv2d_nchw_1[5] = (conv2d_nchw_1[5] + (pad_temp.shared_1[41]*kernel.shared_1[((threadIdx.x*48) + 12)]))
-              conv2d_nchw_1[5] = (conv2d_nchw_1[5] + (pad_temp.shared_1[50]*kernel.shared_1[((threadIdx.x*48) + 15)]))
-              conv2d_nchw_1[6] = (conv2d_nchw_1[6] + (pad_temp.shared_1[42]*kernel.shared_1[((threadIdx.x*48) + 12)]))
-              conv2d_nchw_1[6] = (conv2d_nchw_1[6] + (pad_temp.shared_1[51]*kernel.shared_1[((threadIdx.x*48) + 15)]))
-              conv2d_nchw_1[7] = (conv2d_nchw_1[7] + (pad_temp.shared_1[36]*kernel.shared_1[((threadIdx.x*48) + 36)]))
-              conv2d_nchw_1[7] = (conv2d_nchw_1[7] + (pad_temp.shared_1[45]*kernel.shared_1[((threadIdx.x*48) + 39)]))
-              conv2d_nchw_1[8] = (conv2d_nchw_1[8] + (pad_temp.shared_1[37]*kernel.shared_1[((threadIdx.x*48) + 36)]))
-              conv2d_nchw_1[8] = (conv2d_nchw_1[8] + (pad_temp.shared_1[46]*kernel.shared_1[((threadIdx.x*48) + 39)]))
-              conv2d_nchw_1[9] = (conv2d_nchw_1[9] + (pad_temp.shared_1[38]*kernel.shared_1[((threadIdx.x*48) + 36)]))
-              conv2d_nchw_1[9] = (conv2d_nchw_1[9] + (pad_temp.shared_1[47]*kernel.shared_1[((threadIdx.x*48) + 39)]))
-              conv2d_nchw_1[10] = (conv2d_nchw_1[10] + (pad_temp.shared_1[39]*kernel.shared_1[((threadIdx.x*48) + 36)]))
-              conv2d_nchw_1[10] = (conv2d_nchw_1[10] + (pad_temp.shared_1[48]*kernel.shared_1[((threadIdx.x*48) + 39)]))
-              conv2d_nchw_1[11] = (conv2d_nchw_1[11] + (pad_temp.shared_1[40]*kernel.shared_1[((threadIdx.x*48) + 36)]))
-              conv2d_nchw_1[11] = (conv2d_nchw_1[11] + (pad_temp.shared_1[49]*kernel.shared_1[((threadIdx.x*48) + 39)]))
-              conv2d_nchw_1[12] = (conv2d_nchw_1[12] + (pad_temp.shared_1[41]*kernel.shared_1[((threadIdx.x*48) + 36)]))
-              conv2d_nchw_1[12] = (conv2d_nchw_1[12] + (pad_temp.shared_1[50]*kernel.shared_1[((threadIdx.x*48) + 39)]))
-              conv2d_nchw_1[13] = (conv2d_nchw_1[13] + (pad_temp.shared_1[42]*kernel.shared_1[((threadIdx.x*48) + 36)]))
-              conv2d_nchw_1[13] = (conv2d_nchw_1[13] + (pad_temp.shared_1[51]*kernel.shared_1[((threadIdx.x*48) + 39)]))
-              conv2d_nchw_1[0] = (conv2d_nchw_1[0] + (pad_temp.shared_1[37]*kernel.shared_1[((threadIdx.x*48) + 13)]))
-              conv2d_nchw_1[0] = (conv2d_nchw_1[0] + (pad_temp.shared_1[46]*kernel.shared_1[((threadIdx.x*48) + 16)]))
-              conv2d_nchw_1[1] = (conv2d_nchw_1[1] + (pad_temp.shared_1[38]*kernel.shared_1[((threadIdx.x*48) + 13)]))
-              conv2d_nchw_1[1] = (conv2d_nchw_1[1] + (pad_temp.shared_1[47]*kernel.shared_1[((threadIdx.x*48) + 16)]))
-              conv2d_nchw_1[2] = (conv2d_nchw_1[2] + (pad_temp.shared_1[39]*kernel.shared_1[((threadIdx.x*48) + 13)]))
-              conv2d_nchw_1[2] = (conv2d_nchw_1[2] + (pad_temp.shared_1[48]*kernel.shared_1[((threadIdx.x*48) + 16)]))
-              conv2d_nchw_1[3] = (conv2d_nchw_1[3] + (pad_temp.shared_1[40]*kernel.shared_1[((threadIdx.x*48) + 13)]))
-              conv2d_nchw_1[3] = (conv2d_nchw_1[3] + (pad_temp.shared_1[49]*kernel.shared_1[((threadIdx.x*48) + 16)]))
-              conv2d_nchw_1[4] = (conv2d_nchw_1[4] + (pad_temp.shared_1[41]*kernel.shared_1[((threadIdx.x*48) + 13)]))
-              conv2d_nchw_1[4] = (conv2d_nchw_1[4] + (pad_temp.shared_1[50]*kernel.shared_1[((threadIdx.x*48) + 16)]))
-              conv2d_nchw_1[5] = (conv2d_nchw_1[5] + (pad_temp.shared_1[42]*kernel.shared_1[((threadIdx.x*48) + 13)]))
-              conv2d_nchw_1[5] = (conv2d_nchw_1[5] + (pad_temp.shared_1[51]*kernel.shared_1[((threadIdx.x*48) + 16)]))
-              conv2d_nchw_1[6] = (conv2d_nchw_1[6] + (pad_temp.shared_1[43]*kernel.shared_1[((threadIdx.x*48) + 13)]))
-              conv2d_nchw_1[6] = (conv2d_nchw_1[6] + (pad_temp.shared_1[52]*kernel.shared_1[((threadIdx.x*48) + 16)]))
-              conv2d_nchw_1[7] = (conv2d_nchw_1[7] + (pad_temp.shared_1[37]*kernel.shared_1[((threadIdx.x*48) + 37)]))
-              conv2d_nchw_1[7] = (conv2d_nchw_1[7] + (pad_temp.shared_1[46]*kernel.shared_1[((threadIdx.x*48) + 40)]))
-              conv2d_nchw_1[8] = (conv2d_nchw_1[8] + (pad_temp.shared_1[38]*kernel.shared_1[((threadIdx.x*48) + 37)]))
-              conv2d_nchw_1[8] = (conv2d_nchw_1[8] + (pad_temp.shared_1[47]*kernel.shared_1[((threadIdx.x*48) + 40)]))
-              conv2d_nchw_1[9] = (conv2d_nchw_1[9] + (pad_temp.shared_1[39]*kernel.shared_1[((threadIdx.x*48) + 37)]))
-              conv2d_nchw_1[9] = (conv2d_nchw_1[9] + (pad_temp.shared_1[48]*kernel.shared_1[((threadIdx.x*48) + 40)]))
-              conv2d_nchw_1[10] = (conv2d_nchw_1[10] + (pad_temp.shared_1[40]*kernel.shared_1[((threadIdx.x*48) + 37)]))
-              conv2d_nchw_1[10] = (conv2d_nchw_1[10] + (pad_temp.shared_1[49]*kernel.shared_1[((threadIdx.x*48) + 40)]))
-              conv2d_nchw_1[11] = (conv2d_nchw_1[11] + (pad_temp.shared_1[41]*kernel.shared_1[((threadIdx.x*48) + 37)]))
-              conv2d_nchw_1[11] = (conv2d_nchw_1[11] + (pad_temp.shared_1[50]*kernel.shared_1[((threadIdx.x*48) + 40)]))
-              conv2d_nchw_1[12] = (conv2d_nchw_1[12] + (pad_temp.shared_1[42]*kernel.shared_1[((threadIdx.x*48) + 37)]))
-              conv2d_nchw_1[12] = (conv2d_nchw_1[12] + (pad_temp.shared_1[51]*kernel.shared_1[((threadIdx.x*48) + 40)]))
-              conv2d_nchw_1[13] = (conv2d_nchw_1[13] + (pad_temp.shared_1[43]*kernel.shared_1[((threadIdx.x*48) + 37)]))
-              conv2d_nchw_1[13] = (conv2d_nchw_1[13] + (pad_temp.shared_1[52]*kernel.shared_1[((threadIdx.x*48) + 40)]))
-              conv2d_nchw_1[0] = (conv2d_nchw_1[0] + (pad_temp.shared_1[38]*kernel.shared_1[((threadIdx.x*48) + 14)]))
-              conv2d_nchw_1[0] = (conv2d_nchw_1[0] + (pad_temp.shared_1[47]*kernel.shared_1[((threadIdx.x*48) + 17)]))
-              conv2d_nchw_1[1] = (conv2d_nchw_1[1] + (pad_temp.shared_1[39]*kernel.shared_1[((threadIdx.x*48) + 14)]))
-              conv2d_nchw_1[1] = (conv2d_nchw_1[1] + (pad_temp.shared_1[48]*kernel.shared_1[((threadIdx.x*48) + 17)]))
-              conv2d_nchw_1[2] = (conv2d_nchw_1[2] + (pad_temp.shared_1[40]*kernel.shared_1[((threadIdx.x*48) + 14)]))
-              conv2d_nchw_1[2] = (conv2d_nchw_1[2] + (pad_temp.shared_1[49]*kernel.shared_1[((threadIdx.x*48) + 17)]))
-              conv2d_nchw_1[3] = (conv2d_nchw_1[3] + (pad_temp.shared_1[41]*kernel.shared_1[((threadIdx.x*48) + 14)]))
-              conv2d_nchw_1[3] = (conv2d_nchw_1[3] + (pad_temp.shared_1[50]*kernel.shared_1[((threadIdx.x*48) + 17)]))
-              conv2d_nchw_1[4] = (conv2d_nchw_1[4] + (pad_temp.shared_1[42]*kernel.shared_1[((threadIdx.x*48) + 14)]))
-              conv2d_nchw_1[4] = (conv2d_nchw_1[4] + (pad_temp.shared_1[51]*kernel.shared_1[((threadIdx.x*48) + 17)]))
-              conv2d_nchw_1[5] = (conv2d_nchw_1[5] + (pad_temp.shared_1[43]*kernel.shared_1[((threadIdx.x*48) + 14)]))
-              conv2d_nchw_1[5] = (conv2d_nchw_1[5] + (pad_temp.shared_1[52]*kernel.shared_1[((threadIdx.x*48) + 17)]))
-              conv2d_nchw_1[6] = (conv2d_nchw_1[6] + (pad_temp.shared_1[44]*kernel.shared_1[((threadIdx.x*48) + 14)]))
-              conv2d_nchw_1[6] = (conv2d_nchw_1[6] + (pad_temp.shared_1[53]*kernel.shared_1[((threadIdx.x*48) + 17)]))
-              conv2d_nchw_1[7] = (conv2d_nchw_1[7] + (pad_temp.shared_1[38]*kernel.shared_1[((threadIdx.x*48) + 38)]))
-              conv2d_nchw_1[7] = (conv2d_nchw_1[7] + (pad_temp.shared_1[47]*kernel.shared_1[((threadIdx.x*48) + 41)]))
-              conv2d_nchw_1[8] = (conv2d_nchw_1[8] + (pad_temp.shared_1[39]*kernel.shared_1[((threadIdx.x*48) + 38)]))
-              conv2d_nchw_1[8] = (conv2d_nchw_1[8] + (pad_temp.shared_1[48]*kernel.shared_1[((threadIdx.x*48) + 41)]))
-              conv2d_nchw_1[9] = (conv2d_nchw_1[9] + (pad_temp.shared_1[40]*kernel.shared_1[((threadIdx.x*48) + 38)]))
-              conv2d_nchw_1[9] = (conv2d_nchw_1[9] + (pad_temp.shared_1[49]*kernel.shared_1[((threadIdx.x*48) + 41)]))
-              conv2d_nchw_1[10] = (conv2d_nchw_1[10] + (pad_temp.shared_1[41]*kernel.shared_1[((threadIdx.x*48) + 38)]))
-              conv2d_nchw_1[10] = (conv2d_nchw_1[10] + (pad_temp.shared_1[50]*kernel.shared_1[((threadIdx.x*48) + 41)]))
-              conv2d_nchw_1[11] = (conv2d_nchw_1[11] + (pad_temp.shared_1[42]*kernel.shared_1[((threadIdx.x*48) + 38)]))
-              conv2d_nchw_1[11] = (conv2d_nchw_1[11] + (pad_temp.shared_1[51]*kernel.shared_1[((threadIdx.x*48) + 41)]))
-              conv2d_nchw_1[12] = (conv2d_nchw_1[12] + (pad_temp.shared_1[43]*kernel.shared_1[((threadIdx.x*48) + 38)]))
-              conv2d_nchw_1[12] = (conv2d_nchw_1[12] + (pad_temp.shared_1[52]*kernel.shared_1[((threadIdx.x*48) + 41)]))
-              conv2d_nchw_1[13] = (conv2d_nchw_1[13] + (pad_temp.shared_1[44]*kernel.shared_1[((threadIdx.x*48) + 38)]))
-              conv2d_nchw_1[13] = (conv2d_nchw_1[13] + (pad_temp.shared_1[53]*kernel.shared_1[((threadIdx.x*48) + 41)]))
-              conv2d_nchw_1[0] = (conv2d_nchw_1[0] + (pad_temp.shared_1[54]*kernel.shared_1[((threadIdx.x*48) + 18)]))
-              conv2d_nchw_1[0] = (conv2d_nchw_1[0] + (pad_temp.shared_1[63]*kernel.shared_1[((threadIdx.x*48) + 21)]))
-              conv2d_nchw_1[1] = (conv2d_nchw_1[1] + (pad_temp.shared_1[55]*kernel.shared_1[((threadIdx.x*48) + 18)]))
-              conv2d_nchw_1[1] = (conv2d_nchw_1[1] + (pad_temp.shared_1[64]*kernel.shared_1[((threadIdx.x*48) + 21)]))
-              conv2d_nchw_1[2] = (conv2d_nchw_1[2] + (pad_temp.shared_1[56]*kernel.shared_1[((threadIdx.x*48) + 18)]))
-              conv2d_nchw_1[2] = (conv2d_nchw_1[2] + (pad_temp.shared_1[65]*kernel.shared_1[((threadIdx.x*48) + 21)]))
-              conv2d_nchw_1[3] = (conv2d_nchw_1[3] + (pad_temp.shared_1[57]*kernel.shared_1[((threadIdx.x*48) + 18)]))
-              conv2d_nchw_1[3] = (conv2d_nchw_1[3] + (pad_temp.shared_1[66]*kernel.shared_1[((threadIdx.x*48) + 21)]))
-              conv2d_nchw_1[4] = (conv2d_nchw_1[4] + (pad_temp.shared_1[58]*kernel.shared_1[((threadIdx.x*48) + 18)]))
-              conv2d_nchw_1[4] = (conv2d_nchw_1[4] + (pad_temp.shared_1[67]*kernel.shared_1[((threadIdx.x*48) + 21)]))
-              conv2d_nchw_1[5] = (conv2d_nchw_1[5] + (pad_temp.shared_1[59]*kernel.shared_1[((threadIdx.x*48) + 18)]))
-              conv2d_nchw_1[5] = (conv2d_nchw_1[5] + (pad_temp.shared_1[68]*kernel.shared_1[((threadIdx.x*48) + 21)]))
-              conv2d_nchw_1[6] = (conv2d_nchw_1[6] + (pad_temp.shared_1[60]*kernel.shared_1[((threadIdx.x*48) + 18)]))
-              conv2d_nchw_1[6] = (conv2d_nchw_1[6] + (pad_temp.shared_1[69]*kernel.shared_1[((threadIdx.x*48) + 21)]))
-              conv2d_nchw_1[7] = (conv2d_nchw_1[7] + (pad_temp.shared_1[54]*kernel.shared_1[((threadIdx.x*48) + 42)]))
-              conv2d_nchw_1[7] = (conv2d_nchw_1[7] + (pad_temp.shared_1[63]*kernel.shared_1[((threadIdx.x*48) + 45)]))
-              conv2d_nchw_1[8] = (conv2d_nchw_1[8] + (pad_temp.shared_1[55]*kernel.shared_1[((threadIdx.x*48) + 42)]))
-              conv2d_nchw_1[8] = (conv2d_nchw_1[8] + (pad_temp.shared_1[64]*kernel.shared_1[((threadIdx.x*48) + 45)]))
-              conv2d_nchw_1[9] = (conv2d_nchw_1[9] + (pad_temp.shared_1[56]*kernel.shared_1[((threadIdx.x*48) + 42)]))
-              conv2d_nchw_1[9] = (conv2d_nchw_1[9] + (pad_temp.shared_1[65]*kernel.shared_1[((threadIdx.x*48) + 45)]))
-              conv2d_nchw_1[10] = (conv2d_nchw_1[10] + (pad_temp.shared_1[57]*kernel.shared_1[((threadIdx.x*48) + 42)]))
-              conv2d_nchw_1[10] = (conv2d_nchw_1[10] + (pad_temp.shared_1[66]*kernel.shared_1[((threadIdx.x*48) + 45)]))
-              conv2d_nchw_1[11] = (conv2d_nchw_1[11] + (pad_temp.shared_1[58]*kernel.shared_1[((threadIdx.x*48) + 42)]))
-              conv2d_nchw_1[11] = (conv2d_nchw_1[11] + (pad_temp.shared_1[67]*kernel.shared_1[((threadIdx.x*48) + 45)]))
-              conv2d_nchw_1[12] = (conv2d_nchw_1[12] + (pad_temp.shared_1[59]*kernel.shared_1[((threadIdx.x*48) + 42)]))
-              conv2d_nchw_1[12] = (conv2d_nchw_1[12] + (pad_temp.shared_1[68]*kernel.shared_1[((threadIdx.x*48) + 45)]))
-              conv2d_nchw_1[13] = (conv2d_nchw_1[13] + (pad_temp.shared_1[60]*kernel.shared_1[((threadIdx.x*48) + 42)]))
-              conv2d_nchw_1[13] = (conv2d_nchw_1[13] + (pad_temp.shared_1[69]*kernel.shared_1[((threadIdx.x*48) + 45)]))
-              conv2d_nchw_1[0] = (conv2d_nchw_1[0] + (pad_temp.shared_1[55]*kernel.shared_1[((threadIdx.x*48) + 19)]))
-              conv2d_nchw_1[0] = (conv2d_nchw_1[0] + (pad_temp.shared_1[64]*kernel.shared_1[((threadIdx.x*48) + 22)]))
-              conv2d_nchw_1[1] = (conv2d_nchw_1[1] + (pad_temp.shared_1[56]*kernel.shared_1[((threadIdx.x*48) + 19)]))
-              conv2d_nchw_1[1] = (conv2d_nchw_1[1] + (pad_temp.shared_1[65]*kernel.shared_1[((threadIdx.x*48) + 22)]))
-              conv2d_nchw_1[2] = (conv2d_nchw_1[2] + (pad_temp.shared_1[57]*kernel.shared_1[((threadIdx.x*48) + 19)]))
-              conv2d_nchw_1[2] = (conv2d_nchw_1[2] + (pad_temp.shared_1[66]*kernel.shared_1[((threadIdx.x*48) + 22)]))
-              conv2d_nchw_1[3] = (conv2d_nchw_1[3] + (pad_temp.shared_1[58]*kernel.shared_1[((threadIdx.x*48) + 19)]))
-              conv2d_nchw_1[3] = (conv2d_nchw_1[3] + (pad_temp.shared_1[67]*kernel.shared_1[((threadIdx.x*48) + 22)]))
-              conv2d_nchw_1[4] = (conv2d_nchw_1[4] + (pad_temp.shared_1[59]*kernel.shared_1[((threadIdx.x*48) + 19)]))
-              conv2d_nchw_1[4] = (conv2d_nchw_1[4] + (pad_temp.shared_1[68]*kernel.shared_1[((threadIdx.x*48) + 22)]))
-              conv2d_nchw_1[5] = (conv2d_nchw_1[5] + (pad_temp.shared_1[60]*kernel.shared_1[((threadIdx.x*48) + 19)]))
-              conv2d_nchw_1[5] = (conv2d_nchw_1[5] + (pad_temp.shared_1[69]*kernel.shared_1[((threadIdx.x*48) + 22)]))
-              conv2d_nchw_1[6] = (conv2d_nchw_1[6] + (pad_temp.shared_1[61]*kernel.shared_1[((threadIdx.x*48) + 19)]))
-              conv2d_nchw_1[6] = (conv2d_nchw_1[6] + (pad_temp.shared_1[70]*kernel.shared_1[((threadIdx.x*48) + 22)]))
-              conv2d_nchw_1[7] = (conv2d_nchw_1[7] + (pad_temp.shared_1[55]*kernel.shared_1[((threadIdx.x*48) + 43)]))
-              conv2d_nchw_1[7] = (conv2d_nchw_1[7] + (pad_temp.shared_1[64]*kernel.shared_1[((threadIdx.x*48) + 46)]))
-              conv2d_nchw_1[8] = (conv2d_nchw_1[8] + (pad_temp.shared_1[56]*kernel.shared_1[((threadIdx.x*48) + 43)]))
-              conv2d_nchw_1[8] = (conv2d_nchw_1[8] + (pad_temp.shared_1[65]*kernel.shared_1[((threadIdx.x*48) + 46)]))
-              conv2d_nchw_1[9] = (conv2d_nchw_1[9] + (pad_temp.shared_1[57]*kernel.shared_1[((threadIdx.x*48) + 43)]))
-              conv2d_nchw_1[9] = (conv2d_nchw_1[9] + (pad_temp.shared_1[66]*kernel.shared_1[((threadIdx.x*48) + 46)]))
-              conv2d_nchw_1[10] = (conv2d_nchw_1[10] + (pad_temp.shared_1[58]*kernel.shared_1[((threadIdx.x*48) + 43)]))
-              conv2d_nchw_1[10] = (conv2d_nchw_1[10] + (pad_temp.shared_1[67]*kernel.shared_1[((threadIdx.x*48) + 46)]))
-              conv2d_nchw_1[11] = (conv2d_nchw_1[11] + (pad_temp.shared_1[59]*kernel.shared_1[((threadIdx.x*48) + 43)]))
-              conv2d_nchw_1[11] = (conv2d_nchw_1[11] + (pad_temp.shared_1[68]*kernel.shared_1[((threadIdx.x*48) + 46)]))
-              conv2d_nchw_1[12] = (conv2d_nchw_1[12] + (pad_temp.shared_1[60]*kernel.shared_1[((threadIdx.x*48) + 43)]))
-              conv2d_nchw_1[12] = (conv2d_nchw_1[12] + (pad_temp.shared_1[69]*kernel.shared_1[((threadIdx.x*48) + 46)]))
-              conv2d_nchw_1[13] = (conv2d_nchw_1[13] + (pad_temp.shared_1[61]*kernel.shared_1[((threadIdx.x*48) + 43)]))
-              conv2d_nchw_1[13] = (conv2d_nchw_1[13] + (pad_temp.shared_1[70]*kernel.shared_1[((threadIdx.x*48) + 46)]))
-              conv2d_nchw_1[0] = (conv2d_nchw_1[0] + (pad_temp.shared_1[56]*kernel.shared_1[((threadIdx.x*48) + 20)]))
-              conv2d_nchw_1[0] = (conv2d_nchw_1[0] + (pad_temp.shared_1[65]*kernel.shared_1[((threadIdx.x*48) + 23)]))
-              conv2d_nchw_1[1] = (conv2d_nchw_1[1] + (pad_temp.shared_1[57]*kernel.shared_1[((threadIdx.x*48) + 20)]))
-              conv2d_nchw_1[1] = (conv2d_nchw_1[1] + (pad_temp.shared_1[66]*kernel.shared_1[((threadIdx.x*48) + 23)]))
-              conv2d_nchw_1[2] = (conv2d_nchw_1[2] + (pad_temp.shared_1[58]*kernel.shared_1[((threadIdx.x*48) + 20)]))
-              conv2d_nchw_1[2] = (conv2d_nchw_1[2] + (pad_temp.shared_1[67]*kernel.shared_1[((threadIdx.x*48) + 23)]))
-              conv2d_nchw_1[3] = (conv2d_nchw_1[3] + (pad_temp.shared_1[59]*kernel.shared_1[((threadIdx.x*48) + 20)]))
-              conv2d_nchw_1[3] = (conv2d_nchw_1[3] + (pad_temp.shared_1[68]*kernel.shared_1[((threadIdx.x*48) + 23)]))
-              conv2d_nchw_1[4] = (conv2d_nchw_1[4] + (pad_temp.shared_1[60]*kernel.shared_1[((threadIdx.x*48) + 20)]))
-              conv2d_nchw_1[4] = (conv2d_nchw_1[4] + (pad_temp.shared_1[69]*kernel.shared_1[((threadIdx.x*48) + 23)]))
-              conv2d_nchw_1[5] = (conv2d_nchw_1[5] + (pad_temp.shared_1[61]*kernel.shared_1[((threadIdx.x*48) + 20)]))
-              conv2d_nchw_1[5] = (conv2d_nchw_1[5] + (pad_temp.shared_1[70]*kernel.shared_1[((threadIdx.x*48) + 23)]))
-              conv2d_nchw_1[6] = (conv2d_nchw_1[6] + (pad_temp.shared_1[62]*kernel.shared_1[((threadIdx.x*48) + 20)]))
-              conv2d_nchw_1[6] = (conv2d_nchw_1[6] + (pad_temp.shared_1[71]*kernel.shared_1[((threadIdx.x*48) + 23)]))
-              conv2d_nchw_1[7] = (conv2d_nchw_1[7] + (pad_temp.shared_1[56]*kernel.shared_1[((threadIdx.x*48) + 44)]))
-              conv2d_nchw_1[7] = (conv2d_nchw_1[7] + (pad_temp.shared_1[65]*kernel.shared_1[((threadIdx.x*48) + 47)]))
-              conv2d_nchw_1[8] = (conv2d_nchw_1[8] + (pad_temp.shared_1[57]*kernel.shared_1[((threadIdx.x*48) + 44)]))
-              conv2d_nchw_1[8] = (conv2d_nchw_1[8] + (pad_temp.shared_1[66]*kernel.shared_1[((threadIdx.x*48) + 47)]))
-              conv2d_nchw_1[9] = (conv2d_nchw_1[9] + (pad_temp.shared_1[58]*kernel.shared_1[((threadIdx.x*48) + 44)]))
-              conv2d_nchw_1[9] = (conv2d_nchw_1[9] + (pad_temp.shared_1[67]*kernel.shared_1[((threadIdx.x*48) + 47)]))
-              conv2d_nchw_1[10] = (conv2d_nchw_1[10] + (pad_temp.shared_1[59]*kernel.shared_1[((threadIdx.x*48) + 44)]))
-              conv2d_nchw_1[10] = (conv2d_nchw_1[10] + (pad_temp.shared_1[68]*kernel.shared_1[((threadIdx.x*48) + 47)]))
-              conv2d_nchw_1[11] = (conv2d_nchw_1[11] + (pad_temp.shared_1[60]*kernel.shared_1[((threadIdx.x*48) + 44)]))
-              conv2d_nchw_1[11] = (conv2d_nchw_1[11] + (pad_temp.shared_1[69]*kernel.shared_1[((threadIdx.x*48) + 47)]))
-              conv2d_nchw_1[12] = (conv2d_nchw_1[12] + (pad_temp.shared_1[61]*kernel.shared_1[((threadIdx.x*48) + 44)]))
-              conv2d_nchw_1[12] = (conv2d_nchw_1[12] + (pad_temp.shared_1[70]*kernel.shared_1[((threadIdx.x*48) + 47)]))
-              conv2d_nchw_1[13] = (conv2d_nchw_1[13] + (pad_temp.shared_1[62]*kernel.shared_1[((threadIdx.x*48) + 44)]))
-              conv2d_nchw_1[13] = (conv2d_nchw_1[13] + (pad_temp.shared_1[71]*kernel.shared_1[((threadIdx.x*48) + 47)]))
+          let cse_var_2: int32 = (rc.outer.outer*392)
+          let cse_var_1: int32 = (rc.outer.outer*72)
+           {
+            attr [IterVar(threadIdx.x_1: int32, (nullptr), "ThreadIndex", "threadIdx.x")] "thread_extent" = 64;
+            pad_temp.shared_1: Buffer(pad_temp.shared, float32, [216], [], scope="shared")[threadIdx.x_1] = @tir.if_then_else(((((1 <= (floordiv(floormod(threadIdx.x_1, 27), 9) + floormod(blockIdx.x, 7))) && ((floordiv(floormod(threadIdx.x_1, 27), 9) + floormod(blockIdx.x, 7)) < 8)) && (1 <= floormod(threadIdx.x_1, 9))) && (floormod(threadIdx.x_1, 9) < 8)), data_3: Buffer(data_2, float32, [25088], [])[(((((cse_var_2 + (floordiv(threadIdx.x_1, 27)*49)) + (floordiv(floormod(threadIdx.x_1,  [...]
+            attr [IterVar(threadIdx.x_1, (nullptr), "ThreadIndex", "threadIdx.x")] "thread_extent" = 64;
+            pad_temp.shared_1[(threadIdx.x_1 + 64)] = @tir.if_then_else(((((1 <= (floordiv(floormod((threadIdx.x_1 + 10), 27), 9) + floormod(blockIdx.x, 7))) && ((floordiv(floormod((threadIdx.x_1 + 10), 27), 9) + floormod(blockIdx.x, 7)) < 8)) && (1 <= floormod((threadIdx.x_1 + 1), 9))) && (floormod((threadIdx.x_1 + 1), 9) < 8)), data_3[(((((cse_var_2 + (floordiv((threadIdx.x_1 + 64), 27)*49)) + (floordiv(floormod((threadIdx.x_1 + 10), 27), 9)*7)) + (floormod(blockIdx.x, 7)*7)) + floormo [...]
+            attr [IterVar(threadIdx.x_1, (nullptr), "ThreadIndex", "threadIdx.x")] "thread_extent" = 64;
+            pad_temp.shared_1[(threadIdx.x_1 + 128)] = @tir.if_then_else(((((1 <= (floordiv(floormod((threadIdx.x_1 + 20), 27), 9) + floormod(blockIdx.x, 7))) && ((floordiv(floormod((threadIdx.x_1 + 20), 27), 9) + floormod(blockIdx.x, 7)) < 8)) && (1 <= floormod((threadIdx.x_1 + 2), 9))) && (floormod((threadIdx.x_1 + 2), 9) < 8)), data_3[(((((cse_var_2 + (floordiv((threadIdx.x_1 + 128), 27)*49)) + (floordiv(floormod((threadIdx.x_1 + 20), 27), 9)*7)) + (floormod(blockIdx.x, 7)*7)) + floor [...]
+            attr [IterVar(threadIdx.x_1, (nullptr), "ThreadIndex", "threadIdx.x")] "thread_extent" = 64;
+            if @tir.likely((threadIdx.x_1 < 24), dtype=bool) {
+              pad_temp.shared_1[(threadIdx.x_1 + 192)] = @tir.if_then_else(((((1 <= (floordiv(floormod((threadIdx.x_1 + 3), 27), 9) + floormod(blockIdx.x, 7))) && ((floordiv(floormod((threadIdx.x_1 + 3), 27), 9) + floormod(blockIdx.x, 7)) < 8)) && (1 <= floormod((threadIdx.x_1 + 3), 9))) && (floormod((threadIdx.x_1 + 3), 9) < 8)), data_3[(((((cse_var_2 + (floordiv((threadIdx.x_1 + 192), 27)*49)) + (floordiv(floormod((threadIdx.x_1 + 3), 27), 9)*7)) + (floormod(blockIdx.x, 7)*7)) + floorm [...]
             }
+            attr [IterVar(threadIdx.x_2: int32, (nullptr), "ThreadIndex", "threadIdx.x")] "thread_extent" = 64;
+            kernel.shared_1: Buffer(kernel.shared, float32, [9216], [], scope="shared")[threadIdx.x_2] = kernel_3: Buffer(kernel_2, float32, [2359296], [])[(((floordiv(blockIdx.x, 7)*589824) + cse_var_1) + threadIdx.x_2)]
+            attr [IterVar(threadIdx.x_2, (nullptr), "ThreadIndex", "threadIdx.x")] "thread_extent" = 64;
+            kernel.shared_1[(threadIdx.x_2 + 64)] = kernel_3[(((((floordiv(blockIdx.x, 7)*589824) + (floordiv((threadIdx.x_2 + 64), 72)*4608)) + cse_var_1) + (floordiv(floormod((threadIdx.x_2 + 64), 72), 3)*3)) + floormod((threadIdx.x_2 + 1), 3))]
+            attr [IterVar(threadIdx.x_2, (nullptr), "ThreadIndex", "threadIdx.x")] "thread_extent" = 64;
+            kernel.shared_1[(threadIdx.x_2 + 128)] = kernel_3[(((((floordiv(blockIdx.x, 7)*589824) + (floordiv((threadIdx.x_2 + 128), 72)*4608)) + cse_var_1) + (floordiv(floormod((threadIdx.x_2 + 56), 72), 3)*3)) + floormod((threadIdx.x_2 + 2), 3))]
+            attr [IterVar(threadIdx.x_2, (nullptr), "ThreadIndex", "threadIdx.x")] "thread_extent" = 64;
+            kernel.shared_1[(threadIdx.x_2 + 192)] = kernel_3[(((((floordiv(blockIdx.x, 7)*589824) + (floordiv((threadIdx.x_2 + 192), 72)*4608)) + cse_var_1) + (floormod((floordiv(threadIdx.x_2, 3) + 16), 24)*3)) + floormod(threadIdx.x_2, 3))]
+            attr [IterVar(threadIdx.x_2, (nullptr), "ThreadIndex", "threadIdx.x")] "thread_extent" = 64;
+            kernel.shared_1[(threadIdx.x_2 + 256)] = kernel_3[(((((floordiv(blockIdx.x, 7)*589824) + (floordiv((threadIdx.x_2 + 256), 72)*4608)) + cse_var_1) + (floordiv(floormod((threadIdx.x_2 + 40), 72), 3)*3)) + floormod((threadIdx.x_2 + 1), 3))]
+            attr [IterVar(threadIdx.x_2, (nullptr), "ThreadIndex", "threadIdx.x")] "thread_extent" = 64;
+            kernel.shared_1[(threadIdx.x_2 + 320)] = kernel_3[(((((floordiv(blockIdx.x, 7)*589824) + (floordiv((threadIdx.x_2 + 320), 72)*4608)) + cse_var_1) + (floordiv(floormod((threadIdx.x_2 + 32), 72), 3)*3)) + floormod((threadIdx.x_2 + 2), 3))]
+            attr [IterVar(threadIdx.x_2, (nullptr), "ThreadIndex", "threadIdx.x")] "thread_extent" = 64;
+            kernel.shared_1[(threadIdx.x_2 + 384)] = kernel_3[(((((floordiv(blockIdx.x, 7)*589824) + (floordiv((threadIdx.x_2 + 384), 72)*4608)) + cse_var_1) + (floormod((floordiv(threadIdx.x_2, 3) + 8), 24)*3)) + floormod(threadIdx.x_2, 3))]
+            attr [IterVar(threadIdx.x_2, (nullptr), "ThreadIndex", "threadIdx.x")] "thread_extent" = 64;
+            kernel.shared_1[(threadIdx.x_2 + 448)] = kernel_3[(((((floordiv(blockIdx.x, 7)*589824) + (floordiv((threadIdx.x_2 + 448), 72)*4608)) + cse_var_1) + (floordiv(floormod((threadIdx.x_2 + 16), 72), 3)*3)) + floormod((threadIdx.x_2 + 1), 3))]
+            attr [IterVar(threadIdx.x_2, (nullptr), "ThreadIndex", "threadIdx.x")] "thread_extent" = 64;
+            kernel.shared_1[(threadIdx.x_2 + 512)] = kernel_3[(((((floordiv(blockIdx.x, 7)*589824) + (floordiv((threadIdx.x_2 + 512), 72)*4608)) + cse_var_1) + (floordiv(floormod((threadIdx.x_2 + 8), 72), 3)*3)) + floormod((threadIdx.x_2 + 2), 3))]
+            attr [IterVar(threadIdx.x_2, (nullptr), "ThreadIndex", "threadIdx.x")] "thread_extent" = 64;
+            kernel.shared_1[(threadIdx.x_2 + 576)] = kernel_3[((((floordiv(blockIdx.x, 7)*589824) + cse_var_1) + threadIdx.x_2) + 36864)]
+            attr [IterVar(threadIdx.x_2, (nullptr), "ThreadIndex", "threadIdx.x")] "thread_extent" = 64;
+            kernel.shared_1[(threadIdx.x_2 + 640)] = kernel_3[(((((floordiv(blockIdx.x, 7)*589824) + (floordiv((threadIdx.x_2 + 640), 72)*4608)) + cse_var_1) + (floordiv(floormod((threadIdx.x_2 + 64), 72), 3)*3)) + floormod((threadIdx.x_2 + 1), 3))]
+            attr [IterVar(threadIdx.x_2, (nullptr), "ThreadIndex", "threadIdx.x")] "thread_extent" = 64;
+            kernel.shared_1[(threadIdx.x_2 + 704)] = kernel_3[(((((floordiv(blockIdx.x, 7)*589824) + (floordiv((threadIdx.x_2 + 704), 72)*4608)) + cse_var_1) + (floordiv(floormod((threadIdx.x_2 + 56), 72), 3)*3)) + floormod((threadIdx.x_2 + 2), 3))]
+            attr [IterVar(threadIdx.x_2, (nullptr), "ThreadIndex", "threadIdx.x")] "thread_extent" = 64;
+            kernel.shared_1[(threadIdx.x_2 + 768)] = kernel_3[(((((floordiv(blockIdx.x, 7)*589824) + (floordiv((threadIdx.x_2 + 768), 72)*4608)) + cse_var_1) + (floormod((floordiv(threadIdx.x_2, 3) + 16), 24)*3)) + floormod(threadIdx.x_2, 3))]
+            attr [IterVar(threadIdx.x_2, (nullptr), "ThreadIndex", "threadIdx.x")] "thread_extent" = 64;
+            kernel.shared_1[(threadIdx.x_2 + 832)] = kernel_3[(((((floordiv(blockIdx.x, 7)*589824) + (floordiv((threadIdx.x_2 + 832), 72)*4608)) + cse_var_1) + (floordiv(floormod((threadIdx.x_2 + 40), 72), 3)*3)) + floormod((threadIdx.x_2 + 1), 3))]
+            attr [IterVar(threadIdx.x_2, (nullptr), "ThreadIndex", "threadIdx.x")] "thread_extent" = 64;
+            kernel.shared_1[(threadIdx.x_2 + 896)] = kernel_3[(((((floordiv(blockIdx.x, 7)*589824) + (floordiv((threadIdx.x_2 + 896), 72)*4608)) + cse_var_1) + (floordiv(floormod((threadIdx.x_2 + 32), 72), 3)*3)) + floormod((threadIdx.x_2 + 2), 3))]
+            attr [IterVar(threadIdx.x_2, (nullptr), "ThreadIndex", "threadIdx.x")] "thread_extent" = 64;
+            kernel.shared_1[(threadIdx.x_2 + 960)] = kernel_3[(((((floordiv(blockIdx.x, 7)*589824) + (floordiv((threadIdx.x_2 + 960), 72)*4608)) + cse_var_1) + (floormod((floordiv(threadIdx.x_2, 3) + 8), 24)*3)) + floormod(threadIdx.x_2, 3))]
+            attr [IterVar(threadIdx.x_2, (nullptr), "ThreadIndex", "threadIdx.x")] "thread_extent" = 64;
+            kernel.shared_1[(threadIdx.x_2 + 1024)] = kernel_3[(((((floordiv(blockIdx.x, 7)*589824) + (floordiv((threadIdx.x_2 + 1024), 72)*4608)) + cse_var_1) + (floordiv(floormod((threadIdx.x_2 + 16), 72), 3)*3)) + floormod((threadIdx.x_2 + 1), 3))]
+            attr [IterVar(threadIdx.x_2, (nullptr), "ThreadIndex", "threadIdx.x")] "thread_extent" = 64;
+            kernel.shared_1[(threadIdx.x_2 + 1088)] = kernel_3[(((((floordiv(blockIdx.x, 7)*589824) + (floordiv((threadIdx.x_2 + 1088), 72)*4608)) + cse_var_1) + (floordiv(floormod((threadIdx.x_2 + 8), 72), 3)*3)) + floormod((threadIdx.x_2 + 2), 3))]
+            attr [IterVar(threadIdx.x_2, (nullptr), "ThreadIndex", "threadIdx.x")] "thread_extent" = 64;
+            kernel.shared_1[(threadIdx.x_2 + 1152)] = kernel_3[((((floordiv(blockIdx.x, 7)*589824) + cse_var_1) + threadIdx.x_2) + 73728)]
+            attr [IterVar(threadIdx.x_2, (nullptr), "ThreadIndex", "threadIdx.x")] "thread_extent" = 64;
+            kernel.shared_1[(threadIdx.x_2 + 1216)] = kernel_3[(((((floordiv(blockIdx.x, 7)*589824) + (floordiv((threadIdx.x_2 + 1216), 72)*4608)) + cse_var_1) + (floordiv(floormod((threadIdx.x_2 + 64), 72), 3)*3)) + floormod((threadIdx.x_2 + 1), 3))]
+            attr [IterVar(threadIdx.x_2, (nullptr), "ThreadIndex", "threadIdx.x")] "thread_extent" = 64;
+            kernel.shared_1[(threadIdx.x_2 + 1280)] = kernel_3[(((((floordiv(blockIdx.x, 7)*589824) + (floordiv((threadIdx.x_2 + 1280), 72)*4608)) + cse_var_1) + (floordiv(floormod((threadIdx.x_2 + 56), 72), 3)*3)) + floormod((threadIdx.x_2 + 2), 3))]
+            attr [IterVar(threadIdx.x_2, (nullptr), "ThreadIndex", "threadIdx.x")] "thread_extent" = 64;
+            kernel.shared_1[(threadIdx.x_2 + 1344)] = kernel_3[(((((floordiv(blockIdx.x, 7)*589824) + (floordiv((threadIdx.x_2 + 1344), 72)*4608)) + cse_var_1) + (floormod((floordiv(threadIdx.x_2, 3) + 16), 24)*3)) + floormod(threadIdx.x_2, 3))]
+            attr [IterVar(threadIdx.x_2, (nullptr), "ThreadIndex", "threadIdx.x")] "thread_extent" = 64;
+            kernel.shared_1[(threadIdx.x_2 + 1408)] = kernel_3[(((((floordiv(blockIdx.x, 7)*589824) + (floordiv((threadIdx.x_2 + 1408), 72)*4608)) + cse_var_1) + (floordiv(floormod((threadIdx.x_2 + 40), 72), 3)*3)) + floormod((threadIdx.x_2 + 1), 3))]
+            attr [IterVar(threadIdx.x_2, (nullptr), "ThreadIndex", "threadIdx.x")] "thread_extent" = 64;
+            kernel.shared_1[(threadIdx.x_2 + 1472)] = kernel_3[(((((floordiv(blockIdx.x, 7)*589824) + (floordiv((threadIdx.x_2 + 1472), 72)*4608)) + cse_var_1) + (floordiv(floormod((threadIdx.x_2 + 32), 72), 3)*3)) + floormod((threadIdx.x_2 + 2), 3))]
+            attr [IterVar(threadIdx.x_2, (nullptr), "ThreadIndex", "threadIdx.x")] "thread_extent" = 64;
+            kernel.shared_1[(threadIdx.x_2 + 1536)] = kernel_3[(((((floordiv(blockIdx.x, 7)*589824) + (floordiv((threadIdx.x_2 + 1536), 72)*4608)) + cse_var_1) + (floormod((floordiv(threadIdx.x_2, 3) + 8), 24)*3)) + floormod(threadIdx.x_2, 3))]
+            attr [IterVar(threadIdx.x_2, (nullptr), "ThreadIndex", "threadIdx.x")] "thread_extent" = 64;
+            kernel.shared_1[(threadIdx.x_2 + 1600)] = kernel_3[(((((floordiv(blockIdx.x, 7)*589824) + (floordiv((threadIdx.x_2 + 1600), 72)*4608)) + cse_var_1) + (floordiv(floormod((threadIdx.x_2 + 16), 72), 3)*3)) + floormod((threadIdx.x_2 + 1), 3))]
+            attr [IterVar(threadIdx.x_2, (nullptr), "ThreadIndex", "threadIdx.x")] "thread_extent" = 64;
+            kernel.shared_1[(threadIdx.x_2 + 1664)] = kernel_3[(((((floordiv(blockIdx.x, 7)*589824) + (floordiv((threadIdx.x_2 + 1664), 72)*4608)) + cse_var_1) + (floordiv(floormod((threadIdx.x_2 + 8), 72), 3)*3)) + floormod((threadIdx.x_2 + 2), 3))]
+            attr [IterVar(threadIdx.x_2, (nullptr), "ThreadIndex", "threadIdx.x")] "thread_extent" = 64;
+            kernel.shared_1[(threadIdx.x_2 + 1728)] = kernel_3[((((floordiv(blockIdx.x, 7)*589824) + cse_var_1) + threadIdx.x_2) + 110592)]
+            attr [IterVar(threadIdx.x_2, (nullptr), "ThreadIndex", "threadIdx.x")] "thread_extent" = 64;
+            kernel.shared_1[(threadIdx.x_2 + 1792)] = kernel_3[(((((floordiv(blockIdx.x, 7)*589824) + (floordiv((threadIdx.x_2 + 1792), 72)*4608)) + cse_var_1) + (floordiv(floormod((threadIdx.x_2 + 64), 72), 3)*3)) + floormod((threadIdx.x_2 + 1), 3))]
+            attr [IterVar(threadIdx.x_2, (nullptr), "ThreadIndex", "threadIdx.x")] "thread_extent" = 64;
+            kernel.shared_1[(threadIdx.x_2 + 1856)] = kernel_3[(((((floordiv(blockIdx.x, 7)*589824) + (floordiv((threadIdx.x_2 + 1856), 72)*4608)) + cse_var_1) + (floordiv(floormod((threadIdx.x_2 + 56), 72), 3)*3)) + floormod((threadIdx.x_2 + 2), 3))]
+            attr [IterVar(threadIdx.x_2, (nullptr), "ThreadIndex", "threadIdx.x")] "thread_extent" = 64;
+            kernel.shared_1[(threadIdx.x_2 + 1920)] = kernel_3[(((((floordiv(blockIdx.x, 7)*589824) + (floordiv((threadIdx.x_2 + 1920), 72)*4608)) + cse_var_1) + (floormod((floordiv(threadIdx.x_2, 3) + 16), 24)*3)) + floormod(threadIdx.x_2, 3))]
+            attr [IterVar(threadIdx.x_2, (nullptr), "ThreadIndex", "threadIdx.x")] "thread_extent" = 64;
+            kernel.shared_1[(threadIdx.x_2 + 1984)] = kernel_3[(((((floordiv(blockIdx.x, 7)*589824) + (floordiv((threadIdx.x_2 + 1984), 72)*4608)) + cse_var_1) + (floordiv(floormod((threadIdx.x_2 + 40), 72), 3)*3)) + floormod((threadIdx.x_2 + 1), 3))]
+            attr [IterVar(threadIdx.x_2, (nullptr), "ThreadIndex", "threadIdx.x")] "thread_extent" = 64;
+            kernel.shared_1[(threadIdx.x_2 + 2048)] = kernel_3[(((((floordiv(blockIdx.x, 7)*589824) + (floordiv((threadIdx.x_2 + 2048), 72)*4608)) + cse_var_1) + (floordiv(floormod((threadIdx.x_2 + 32), 72), 3)*3)) + floormod((threadIdx.x_2 + 2), 3))]
+            attr [IterVar(threadIdx.x_2, (nullptr), "ThreadIndex", "threadIdx.x")] "thread_extent" = 64;
+            kernel.shared_1[(threadIdx.x_2 + 2112)] = kernel_3[(((((floordiv(blockIdx.x, 7)*589824) + (floordiv((threadIdx.x_2 + 2112), 72)*4608)) + cse_var_1) + (floormod((floordiv(threadIdx.x_2, 3) + 8), 24)*3)) + floormod(threadIdx.x_2, 3))]
+            attr [IterVar(threadIdx.x_2, (nullptr), "ThreadIndex", "threadIdx.x")] "thread_extent" = 64;
+            kernel.shared_1[(threadIdx.x_2 + 2176)] = kernel_3[(((((floordiv(blockIdx.x, 7)*589824) + (floordiv((threadIdx.x_2 + 2176), 72)*4608)) + cse_var_1) + (floordiv(floormod((threadIdx.x_2 + 16), 72), 3)*3)) + floormod((threadIdx.x_2 + 1), 3))]
+            attr [IterVar(threadIdx.x_2, (nullptr), "ThreadIndex", "threadIdx.x")] "thread_extent" = 64;
+            kernel.shared_1[(threadIdx.x_2 + 2240)] = kernel_3[(((((floordiv(blockIdx.x, 7)*589824) + (floordiv((threadIdx.x_2 + 2240), 72)*4608)) + cse_var_1) + (floordiv(floormod((threadIdx.x_2 + 8), 72), 3)*3)) + floormod((threadIdx.x_2 + 2), 3))]
+            attr [IterVar(threadIdx.x_2, (nullptr), "ThreadIndex", "threadIdx.x")] "thread_extent" = 64;
+            kernel.shared_1[(threadIdx.x_2 + 2304)] = kernel_3[((((floordiv(blockIdx.x, 7)*589824) + cse_var_1) + threadIdx.x_2) + 147456)]
+            attr [IterVar(threadIdx.x_2, (nullptr), "ThreadIndex", "threadIdx.x")] "thread_extent" = 64;
+            kernel.shared_1[(threadIdx.x_2 + 2368)] = kernel_3[(((((floordiv(blockIdx.x, 7)*589824) + (floordiv((threadIdx.x_2 + 2368), 72)*4608)) + cse_var_1) + (floordiv(floormod((threadIdx.x_2 + 64), 72), 3)*3)) + floormod((threadIdx.x_2 + 1), 3))]
+            attr [IterVar(threadIdx.x_2, (nullptr), "ThreadIndex", "threadIdx.x")] "thread_extent" = 64;
+            kernel.shared_1[(threadIdx.x_2 + 2432)] = kernel_3[(((((floordiv(blockIdx.x, 7)*589824) + (floordiv((threadIdx.x_2 + 2432), 72)*4608)) + cse_var_1) + (floordiv(floormod((threadIdx.x_2 + 56), 72), 3)*3)) + floormod((threadIdx.x_2 + 2), 3))]
+            attr [IterVar(threadIdx.x_2, (nullptr), "ThreadIndex", "threadIdx.x")] "thread_extent" = 64;
+            kernel.shared_1[(threadIdx.x_2 + 2496)] = kernel_3[(((((floordiv(blockIdx.x, 7)*589824) + (floordiv((threadIdx.x_2 + 2496), 72)*4608)) + cse_var_1) + (floormod((floordiv(threadIdx.x_2, 3) + 16), 24)*3)) + floormod(threadIdx.x_2, 3))]
+            attr [IterVar(threadIdx.x_2, (nullptr), "ThreadIndex", "threadIdx.x")] "thread_extent" = 64;
+            kernel.shared_1[(threadIdx.x_2 + 2560)] = kernel_3[(((((floordiv(blockIdx.x, 7)*589824) + (floordiv((threadIdx.x_2 + 2560), 72)*4608)) + cse_var_1) + (floordiv(floormod((threadIdx.x_2 + 40), 72), 3)*3)) + floormod((threadIdx.x_2 + 1), 3))]
+            attr [IterVar(threadIdx.x_2, (nullptr), "ThreadIndex", "threadIdx.x")] "thread_extent" = 64;
+            kernel.shared_1[(threadIdx.x_2 + 2624)] = kernel_3[(((((floordiv(blockIdx.x, 7)*589824) + (floordiv((threadIdx.x_2 + 2624), 72)*4608)) + cse_var_1) + (floordiv(floormod((threadIdx.x_2 + 32), 72), 3)*3)) + floormod((threadIdx.x_2 + 2), 3))]
+            attr [IterVar(threadIdx.x_2, (nullptr), "ThreadIndex", "threadIdx.x")] "thread_extent" = 64;
+            kernel.shared_1[(threadIdx.x_2 + 2688)] = kernel_3[(((((floordiv(blockIdx.x, 7)*589824) + (floordiv((threadIdx.x_2 + 2688), 72)*4608)) + cse_var_1) + (floormod((floordiv(threadIdx.x_2, 3) + 8), 24)*3)) + floormod(threadIdx.x_2, 3))]
+            attr [IterVar(threadIdx.x_2, (nullptr), "ThreadIndex", "threadIdx.x")] "thread_extent" = 64;
+            kernel.shared_1[(threadIdx.x_2 + 2752)] = kernel_3[(((((floordiv(blockIdx.x, 7)*589824) + (floordiv((threadIdx.x_2 + 2752), 72)*4608)) + cse_var_1) + (floordiv(floormod((threadIdx.x_2 + 16), 72), 3)*3)) + floormod((threadIdx.x_2 + 1), 3))]
+            attr [IterVar(threadIdx.x_2, (nullptr), "ThreadIndex", "threadIdx.x")] "thread_extent" = 64;
+            kernel.shared_1[(threadIdx.x_2 + 2816)] = kernel_3[(((((floordiv(blockIdx.x, 7)*589824) + (floordiv((threadIdx.x_2 + 2816), 72)*4608)) + cse_var_1) + (floordiv(floormod((threadIdx.x_2 + 8), 72), 3)*3)) + floormod((threadIdx.x_2 + 2), 3))]
+            attr [IterVar(threadIdx.x_2, (nullptr), "ThreadIndex", "threadIdx.x")] "thread_extent" = 64;
+            kernel.shared_1[(threadIdx.x_2 + 2880)] = kernel_3[((((floordiv(blockIdx.x, 7)*589824) + cse_var_1) + threadIdx.x_2) + 184320)]
+            attr [IterVar(threadIdx.x_2, (nullptr), "ThreadIndex", "threadIdx.x")] "thread_extent" = 64;
+            kernel.shared_1[(threadIdx.x_2 + 2944)] = kernel_3[(((((floordiv(blockIdx.x, 7)*589824) + (floordiv((threadIdx.x_2 + 2944), 72)*4608)) + cse_var_1) + (floordiv(floormod((threadIdx.x_2 + 64), 72), 3)*3)) + floormod((threadIdx.x_2 + 1), 3))]
+            attr [IterVar(threadIdx.x_2, (nullptr), "ThreadIndex", "threadIdx.x")] "thread_extent" = 64;
+            kernel.shared_1[(threadIdx.x_2 + 3008)] = kernel_3[(((((floordiv(blockIdx.x, 7)*589824) + (floordiv((threadIdx.x_2 + 3008), 72)*4608)) + cse_var_1) + (floordiv(floormod((threadIdx.x_2 + 56), 72), 3)*3)) + floormod((threadIdx.x_2 + 2), 3))]
+            attr [IterVar(threadIdx.x_2, (nullptr), "ThreadIndex", "threadIdx.x")] "thread_extent" = 64;
+            kernel.shared_1[(threadIdx.x_2 + 3072)] = kernel_3[(((((floordiv(blockIdx.x, 7)*589824) + (floordiv((threadIdx.x_2 + 3072), 72)*4608)) + cse_var_1) + (floormod((floordiv(threadIdx.x_2, 3) + 16), 24)*3)) + floormod(threadIdx.x_2, 3))]
+            attr [IterVar(threadIdx.x_2, (nullptr), "ThreadIndex", "threadIdx.x")] "thread_extent" = 64;
+            kernel.shared_1[(threadIdx.x_2 + 3136)] = kernel_3[(((((floordiv(blockIdx.x, 7)*589824) + (floordiv((threadIdx.x_2 + 3136), 72)*4608)) + cse_var_1) + (floordiv(floormod((threadIdx.x_2 + 40), 72), 3)*3)) + floormod((threadIdx.x_2 + 1), 3))]
+            attr [IterVar(threadIdx.x_2, (nullptr), "ThreadIndex", "threadIdx.x")] "thread_extent" = 64;
+            kernel.shared_1[(threadIdx.x_2 + 3200)] = kernel_3[(((((floordiv(blockIdx.x, 7)*589824) + (floordiv((threadIdx.x_2 + 3200), 72)*4608)) + cse_var_1) + (floordiv(floormod((threadIdx.x_2 + 32), 72), 3)*3)) + floormod((threadIdx.x_2 + 2), 3))]
+            attr [IterVar(threadIdx.x_2, (nullptr), "ThreadIndex", "threadIdx.x")] "thread_extent" = 64;
+            kernel.shared_1[(threadIdx.x_2 + 3264)] = kernel_3[(((((floordiv(blockIdx.x, 7)*589824) + (floordiv((threadIdx.x_2 + 3264), 72)*4608)) + cse_var_1) + (floormod((floordiv(threadIdx.x_2, 3) + 8), 24)*3)) + floormod(threadIdx.x_2, 3))]
+            attr [IterVar(threadIdx.x_2, (nullptr), "ThreadIndex", "threadIdx.x")] "thread_extent" = 64;
+            kernel.shared_1[(threadIdx.x_2 + 3328)] = kernel_3[(((((floordiv(blockIdx.x, 7)*589824) + (floordiv((threadIdx.x_2 + 3328), 72)*4608)) + cse_var_1) + (floordiv(floormod((threadIdx.x_2 + 16), 72), 3)*3)) + floormod((threadIdx.x_2 + 1), 3))]
+            attr [IterVar(threadIdx.x_2, (nullptr), "ThreadIndex", "threadIdx.x")] "thread_extent" = 64;
+            kernel.shared_1[(threadIdx.x_2 + 3392)] = kernel_3[(((((floordiv(blockIdx.x, 7)*589824) + (floordiv((threadIdx.x_2 + 3392), 72)*4608)) + cse_var_1) + (floordiv(floormod((threadIdx.x_2 + 8), 72), 3)*3)) + floormod((threadIdx.x_2 + 2), 3))]
+            attr [IterVar(threadIdx.x_2, (nullptr), "ThreadIndex", "threadIdx.x")] "thread_extent" = 64;
+            kernel.shared_1[(threadIdx.x_2 + 3456)] = kernel_3[((((floordiv(blockIdx.x, 7)*589824) + cse_var_1) + threadIdx.x_2) + 221184)]
+            attr [IterVar(threadIdx.x_2, (nullptr), "ThreadIndex", "threadIdx.x")] "thread_extent" = 64;
+            kernel.shared_1[(threadIdx.x_2 + 3520)] = kernel_3[(((((floordiv(blockIdx.x, 7)*589824) + (floordiv((threadIdx.x_2 + 3520), 72)*4608)) + cse_var_1) + (floordiv(floormod((threadIdx.x_2 + 64), 72), 3)*3)) + floormod((threadIdx.x_2 + 1), 3))]
+            attr [IterVar(threadIdx.x_2, (nullptr), "ThreadIndex", "threadIdx.x")] "thread_extent" = 64;
+            kernel.shared_1[(threadIdx.x_2 + 3584)] = kernel_3[(((((floordiv(blockIdx.x, 7)*589824) + (floordiv((threadIdx.x_2 + 3584), 72)*4608)) + cse_var_1) + (floordiv(floormod((threadIdx.x_2 + 56), 72), 3)*3)) + floormod((threadIdx.x_2 + 2), 3))]
+            attr [IterVar(threadIdx.x_2, (nullptr), "ThreadIndex", "threadIdx.x")] "thread_extent" = 64;
+            kernel.shared_1[(threadIdx.x_2 + 3648)] = kernel_3[(((((floordiv(blockIdx.x, 7)*589824) + (floordiv((threadIdx.x_2 + 3648), 72)*4608)) + cse_var_1) + (floormod((floordiv(threadIdx.x_2, 3) + 16), 24)*3)) + floormod(threadIdx.x_2, 3))]
+            attr [IterVar(threadIdx.x_2, (nullptr), "ThreadIndex", "threadIdx.x")] "thread_extent" = 64;
+            kernel.shared_1[(threadIdx.x_2 + 3712)] = kernel_3[(((((floordiv(blockIdx.x, 7)*589824) + (floordiv((threadIdx.x_2 + 3712), 72)*4608)) + cse_var_1) + (floordiv(floormod((threadIdx.x_2 + 40), 72), 3)*3)) + floormod((threadIdx.x_2 + 1), 3))]
+            attr [IterVar(threadIdx.x_2, (nullptr), "ThreadIndex", "threadIdx.x")] "thread_extent" = 64;
+            kernel.shared_1[(threadIdx.x_2 + 3776)] = kernel_3[(((((floordiv(blockIdx.x, 7)*589824) + (floordiv((threadIdx.x_2 + 3776), 72)*4608)) + cse_var_1) + (floordiv(floormod((threadIdx.x_2 + 32), 72), 3)*3)) + floormod((threadIdx.x_2 + 2), 3))]
+            attr [IterVar(threadIdx.x_2, (nullptr), "ThreadIndex", "threadIdx.x")] "thread_extent" = 64;
+            kernel.shared_1[(threadIdx.x_2 + 3840)] = kernel_3[(((((floordiv(blockIdx.x, 7)*589824) + (floordiv((threadIdx.x_2 + 3840), 72)*4608)) + cse_var_1) + (floormod((floordiv(threadIdx.x_2, 3) + 8), 24)*3)) + floormod(threadIdx.x_2, 3))]
+            attr [IterVar(threadIdx.x_2, (nullptr), "ThreadIndex", "threadIdx.x")] "thread_extent" = 64;
+            kernel.shared_1[(threadIdx.x_2 + 3904)] = kernel_3[(((((floordiv(blockIdx.x, 7)*589824) + (floordiv((threadIdx.x_2 + 3904), 72)*4608)) + cse_var_1) + (floordiv(floormod((threadIdx.x_2 + 16), 72), 3)*3)) + floormod((threadIdx.x_2 + 1), 3))]
+            attr [IterVar(threadIdx.x_2, (nullptr), "ThreadIndex", "threadIdx.x")] "thread_extent" = 64;
+            kernel.shared_1[(threadIdx.x_2 + 3968)] = kernel_3[(((((floordiv(blockIdx.x, 7)*589824) + (floordiv((threadIdx.x_2 + 3968), 72)*4608)) + cse_var_1) + (floordiv(floormod((threadIdx.x_2 + 8), 72), 3)*3)) + floormod((threadIdx.x_2 + 2), 3))]
+            attr [IterVar(threadIdx.x_2, (nullptr), "ThreadIndex", "threadIdx.x")] "thread_extent" = 64;
+            kernel.shared_1[(threadIdx.x_2 + 4032)] = kernel_3[((((floordiv(blockIdx.x, 7)*589824) + cse_var_1) + threadIdx.x_2) + 258048)]
+            attr [IterVar(threadIdx.x_2, (nullptr), "ThreadIndex", "threadIdx.x")] "thread_extent" = 64;
+            kernel.shared_1[(threadIdx.x_2 + 4096)] = kernel_3[(((((floordiv(blockIdx.x, 7)*589824) + (floordiv((threadIdx.x_2 + 4096), 72)*4608)) + cse_var_1) + (floordiv(floormod((threadIdx.x_2 + 64), 72), 3)*3)) + floormod((threadIdx.x_2 + 1), 3))]
+            attr [IterVar(threadIdx.x_2, (nullptr), "ThreadIndex", "threadIdx.x")] "thread_extent" = 64;
+            kernel.shared_1[(threadIdx.x_2 + 4160)] = kernel_3[(((((floordiv(blockIdx.x, 7)*589824) + (floordiv((threadIdx.x_2 + 4160), 72)*4608)) + cse_var_1) + (floordiv(floormod((threadIdx.x_2 + 56), 72), 3)*3)) + floormod((threadIdx.x_2 + 2), 3))]
+            attr [IterVar(threadIdx.x_2, (nullptr), "ThreadIndex", "threadIdx.x")] "thread_extent" = 64;
+            kernel.shared_1[(threadIdx.x_2 + 4224)] = kernel_3[(((((floordiv(blockIdx.x, 7)*589824) + (floordiv((threadIdx.x_2 + 4224), 72)*4608)) + cse_var_1) + (floormod((floordiv(threadIdx.x_2, 3) + 16), 24)*3)) + floormod(threadIdx.x_2, 3))]
+            attr [IterVar(threadIdx.x_2, (nullptr), "ThreadIndex", "threadIdx.x")] "thread_extent" = 64;
+            kernel.shared_1[(threadIdx.x_2 + 4288)] = kernel_3[(((((floordiv(blockIdx.x, 7)*589824) + (floordiv((threadIdx.x_2 + 4288), 72)*4608)) + cse_var_1) + (floordiv(floormod((threadIdx.x_2 + 40), 72), 3)*3)) + floormod((threadIdx.x_2 + 1), 3))]
+            attr [IterVar(threadIdx.x_2, (nullptr), "ThreadIndex", "threadIdx.x")] "thread_extent" = 64;
+            kernel.shared_1[(threadIdx.x_2 + 4352)] = kernel_3[(((((floordiv(blockIdx.x, 7)*589824) + (floordiv((threadIdx.x_2 + 4352), 72)*4608)) + cse_var_1) + (floordiv(floormod((threadIdx.x_2 + 32), 72), 3)*3)) + floormod((threadIdx.x_2 + 2), 3))]
+            attr [IterVar(threadIdx.x_2, (nullptr), "ThreadIndex", "threadIdx.x")] "thread_extent" = 64;
+            kernel.shared_1[(threadIdx.x_2 + 4416)] = kernel_3[(((((floordiv(blockIdx.x, 7)*589824) + (floordiv((threadIdx.x_2 + 4416), 72)*4608)) + cse_var_1) + (floormod((floordiv(threadIdx.x_2, 3) + 8), 24)*3)) + floormod(threadIdx.x_2, 3))]
+            attr [IterVar(threadIdx.x_2, (nullptr), "ThreadIndex", "threadIdx.x")] "thread_extent" = 64;
+            kernel.shared_1[(threadIdx.x_2 + 4480)] = kernel_3[(((((floordiv(blockIdx.x, 7)*589824) + (floordiv((threadIdx.x_2 + 4480), 72)*4608)) + cse_var_1) + (floordiv(floormod((threadIdx.x_2 + 16), 72), 3)*3)) + floormod((threadIdx.x_2 + 1), 3))]
+            attr [IterVar(threadIdx.x_2, (nullptr), "ThreadIndex", "threadIdx.x")] "thread_extent" = 64;
+            kernel.shared_1[(threadIdx.x_2 + 4544)] = kernel_3[(((((floordiv(blockIdx.x, 7)*589824) + (floordiv((threadIdx.x_2 + 4544), 72)*4608)) + cse_var_1) + (floordiv(floormod((threadIdx.x_2 + 8), 72), 3)*3)) + floormod((threadIdx.x_2 + 2), 3))]
+            attr [IterVar(threadIdx.x_2, (nullptr), "ThreadIndex", "threadIdx.x")] "thread_extent" = 64;
+            kernel.shared_1[(threadIdx.x_2 + 4608)] = kernel_3[((((floordiv(blockIdx.x, 7)*589824) + cse_var_1) + threadIdx.x_2) + 294912)]
+            attr [IterVar(threadIdx.x_2, (nullptr), "ThreadIndex", "threadIdx.x")] "thread_extent" = 64;
+            kernel.shared_1[(threadIdx.x_2 + 4672)] = kernel_3[(((((floordiv(blockIdx.x, 7)*589824) + (floordiv((threadIdx.x_2 + 4672), 72)*4608)) + cse_var_1) + (floordiv(floormod((threadIdx.x_2 + 64), 72), 3)*3)) + floormod((threadIdx.x_2 + 1), 3))]
+            attr [IterVar(threadIdx.x_2, (nullptr), "ThreadIndex", "threadIdx.x")] "thread_extent" = 64;
+            kernel.shared_1[(threadIdx.x_2 + 4736)] = kernel_3[(((((floordiv(blockIdx.x, 7)*589824) + (floordiv((threadIdx.x_2 + 4736), 72)*4608)) + cse_var_1) + (floordiv(floormod((threadIdx.x_2 + 56), 72), 3)*3)) + floormod((threadIdx.x_2 + 2), 3))]
+            attr [IterVar(threadIdx.x_2, (nullptr), "ThreadIndex", "threadIdx.x")] "thread_extent" = 64;
+            kernel.shared_1[(threadIdx.x_2 + 4800)] = kernel_3[(((((floordiv(blockIdx.x, 7)*589824) + (floordiv((threadIdx.x_2 + 4800), 72)*4608)) + cse_var_1) + (floormod((floordiv(threadIdx.x_2, 3) + 16), 24)*3)) + floormod(threadIdx.x_2, 3))]
+            attr [IterVar(threadIdx.x_2, (nullptr), "ThreadIndex", "threadIdx.x")] "thread_extent" = 64;
+            kernel.shared_1[(threadIdx.x_2 + 4864)] = kernel_3[(((((floordiv(blockIdx.x, 7)*589824) + (floordiv((threadIdx.x_2 + 4864), 72)*4608)) + cse_var_1) + (floordiv(floormod((threadIdx.x_2 + 40), 72), 3)*3)) + floormod((threadIdx.x_2 + 1), 3))]
+            attr [IterVar(threadIdx.x_2, (nullptr), "ThreadIndex", "threadIdx.x")] "thread_extent" = 64;
+            kernel.shared_1[(threadIdx.x_2 + 4928)] = kernel_3[(((((floordiv(blockIdx.x, 7)*589824) + (floordiv((threadIdx.x_2 + 4928), 72)*4608)) + cse_var_1) + (floordiv(floormod((threadIdx.x_2 + 32), 72), 3)*3)) + floormod((threadIdx.x_2 + 2), 3))]
+            attr [IterVar(threadIdx.x_2, (nullptr), "ThreadIndex", "threadIdx.x")] "thread_extent" = 64;
+            kernel.shared_1[(threadIdx.x_2 + 4992)] = kernel_3[(((((floordiv(blockIdx.x, 7)*589824) + (floordiv((threadIdx.x_2 + 4992), 72)*4608)) + cse_var_1) + (floormod((floordiv(threadIdx.x_2, 3) + 8), 24)*3)) + floormod(threadIdx.x_2, 3))]
+            attr [IterVar(threadIdx.x_2, (nullptr), "ThreadIndex", "threadIdx.x")] "thread_extent" = 64;
+            kernel.shared_1[(threadIdx.x_2 + 5056)] = kernel_3[(((((floordiv(blockIdx.x, 7)*589824) + (floordiv((threadIdx.x_2 + 5056), 72)*4608)) + cse_var_1) + (floordiv(floormod((threadIdx.x_2 + 16), 72), 3)*3)) + floormod((threadIdx.x_2 + 1), 3))]
+            attr [IterVar(threadIdx.x_2, (nullptr), "ThreadIndex", "threadIdx.x")] "thread_extent" = 64;
+            kernel.shared_1[(threadIdx.x_2 + 5120)] = kernel_3[(((((floordiv(blockIdx.x, 7)*589824) + (floordiv((threadIdx.x_2 + 5120), 72)*4608)) + cse_var_1) + (floordiv(floormod((threadIdx.x_2 + 8), 72), 3)*3)) + floormod((threadIdx.x_2 + 2), 3))]
+            attr [IterVar(threadIdx.x_2, (nullptr), "ThreadIndex", "threadIdx.x")] "thread_extent" = 64;
+            kernel.shared_1[(threadIdx.x_2 + 5184)] = kernel_3[((((floordiv(blockIdx.x, 7)*589824) + cse_var_1) + threadIdx.x_2) + 331776)]
+            attr [IterVar(threadIdx.x_2, (nullptr), "ThreadIndex", "threadIdx.x")] "thread_extent" = 64;
+            kernel.shared_1[(threadIdx.x_2 + 5248)] = kernel_3[(((((floordiv(blockIdx.x, 7)*589824) + (floordiv((threadIdx.x_2 + 5248), 72)*4608)) + cse_var_1) + (floordiv(floormod((threadIdx.x_2 + 64), 72), 3)*3)) + floormod((threadIdx.x_2 + 1), 3))]
+            attr [IterVar(threadIdx.x_2, (nullptr), "ThreadIndex", "threadIdx.x")] "thread_extent" = 64;
+            kernel.shared_1[(threadIdx.x_2 + 5312)] = kernel_3[(((((floordiv(blockIdx.x, 7)*589824) + (floordiv((threadIdx.x_2 + 5312), 72)*4608)) + cse_var_1) + (floordiv(floormod((threadIdx.x_2 + 56), 72), 3)*3)) + floormod((threadIdx.x_2 + 2), 3))]
+            attr [IterVar(threadIdx.x_2, (nullptr), "ThreadIndex", "threadIdx.x")] "thread_extent" = 64;
+            kernel.shared_1[(threadIdx.x_2 + 5376)] = kernel_3[(((((floordiv(blockIdx.x, 7)*589824) + (floordiv((threadIdx.x_2 + 5376), 72)*4608)) + cse_var_1) + (floormod((floordiv(threadIdx.x_2, 3) + 16), 24)*3)) + floormod(threadIdx.x_2, 3))]
+            attr [IterVar(threadIdx.x_2, (nullptr), "ThreadIndex", "threadIdx.x")] "thread_extent" = 64;
+            kernel.shared_1[(threadIdx.x_2 + 5440)] = kernel_3[(((((floordiv(blockIdx.x, 7)*589824) + (floordiv((threadIdx.x_2 + 5440), 72)*4608)) + cse_var_1) + (floordiv(floormod((threadIdx.x_2 + 40), 72), 3)*3)) + floormod((threadIdx.x_2 + 1), 3))]
+            attr [IterVar(threadIdx.x_2, (nullptr), "ThreadIndex", "threadIdx.x")] "thread_extent" = 64;
+            kernel.shared_1[(threadIdx.x_2 + 5504)] = kernel_3[(((((floordiv(blockIdx.x, 7)*589824) + (floordiv((threadIdx.x_2 + 5504), 72)*4608)) + cse_var_1) + (floordiv(floormod((threadIdx.x_2 + 32), 72), 3)*3)) + floormod((threadIdx.x_2 + 2), 3))]
+            attr [IterVar(threadIdx.x_2, (nullptr), "ThreadIndex", "threadIdx.x")] "thread_extent" = 64;
+            kernel.shared_1[(threadIdx.x_2 + 5568)] = kernel_3[(((((floordiv(blockIdx.x, 7)*589824) + (floordiv((threadIdx.x_2 + 5568), 72)*4608)) + cse_var_1) + (floormod((floordiv(threadIdx.x_2, 3) + 8), 24)*3)) + floormod(threadIdx.x_2, 3))]
+            attr [IterVar(threadIdx.x_2, (nullptr), "ThreadIndex", "threadIdx.x")] "thread_extent" = 64;
+            kernel.shared_1[(threadIdx.x_2 + 5632)] = kernel_3[(((((floordiv(blockIdx.x, 7)*589824) + (floordiv((threadIdx.x_2 + 5632), 72)*4608)) + cse_var_1) + (floordiv(floormod((threadIdx.x_2 + 16), 72), 3)*3)) + floormod((threadIdx.x_2 + 1), 3))]
+            attr [IterVar(threadIdx.x_2, (nullptr), "ThreadIndex", "threadIdx.x")] "thread_extent" = 64;
+            kernel.shared_1[(threadIdx.x_2 + 5696)] = kernel_3[(((((floordiv(blockIdx.x, 7)*589824) + (floordiv((threadIdx.x_2 + 5696), 72)*4608)) + cse_var_1) + (floordiv(floormod((threadIdx.x_2 + 8), 72), 3)*3)) + floormod((threadIdx.x_2 + 2), 3))]
+            attr [IterVar(threadIdx.x_2, (nullptr), "ThreadIndex", "threadIdx.x")] "thread_extent" = 64;
+            kernel.shared_1[(threadIdx.x_2 + 5760)] = kernel_3[((((floordiv(blockIdx.x, 7)*589824) + cse_var_1) + threadIdx.x_2) + 368640)]
+            attr [IterVar(threadIdx.x_2, (nullptr), "ThreadIndex", "threadIdx.x")] "thread_extent" = 64;
+            kernel.shared_1[(threadIdx.x_2 + 5824)] = kernel_3[(((((floordiv(blockIdx.x, 7)*589824) + (floordiv((threadIdx.x_2 + 5824), 72)*4608)) + cse_var_1) + (floordiv(floormod((threadIdx.x_2 + 64), 72), 3)*3)) + floormod((threadIdx.x_2 + 1), 3))]
+            attr [IterVar(threadIdx.x_2, (nullptr), "ThreadIndex", "threadIdx.x")] "thread_extent" = 64;
+            kernel.shared_1[(threadIdx.x_2 + 5888)] = kernel_3[(((((floordiv(blockIdx.x, 7)*589824) + (floordiv((threadIdx.x_2 + 5888), 72)*4608)) + cse_var_1) + (floordiv(floormod((threadIdx.x_2 + 56), 72), 3)*3)) + floormod((threadIdx.x_2 + 2), 3))]
+            attr [IterVar(threadIdx.x_2, (nullptr), "ThreadIndex", "threadIdx.x")] "thread_extent" = 64;
+            kernel.shared_1[(threadIdx.x_2 + 5952)] = kernel_3[(((((floordiv(blockIdx.x, 7)*589824) + (floordiv((threadIdx.x_2 + 5952), 72)*4608)) + cse_var_1) + (floormod((floordiv(threadIdx.x_2, 3) + 16), 24)*3)) + floormod(threadIdx.x_2, 3))]
+            attr [IterVar(threadIdx.x_2, (nullptr), "ThreadIndex", "threadIdx.x")] "thread_extent" = 64;
+            kernel.shared_1[(threadIdx.x_2 + 6016)] = kernel_3[(((((floordiv(blockIdx.x, 7)*589824) + (floordiv((threadIdx.x_2 + 6016), 72)*4608)) + cse_var_1) + (floordiv(floormod((threadIdx.x_2 + 40), 72), 3)*3)) + floormod((threadIdx.x_2 + 1), 3))]
+            attr [IterVar(threadIdx.x_2, (nullptr), "ThreadIndex", "threadIdx.x")] "thread_extent" = 64;
+            kernel.shared_1[(threadIdx.x_2 + 6080)] = kernel_3[(((((floordiv(blockIdx.x, 7)*589824) + (floordiv((threadIdx.x_2 + 6080), 72)*4608)) + cse_var_1) + (floordiv(floormod((threadIdx.x_2 + 32), 72), 3)*3)) + floormod((threadIdx.x_2 + 2), 3))]
+            attr [IterVar(threadIdx.x_2, (nullptr), "ThreadIndex", "threadIdx.x")] "thread_extent" = 64;
+            kernel.shared_1[(threadIdx.x_2 + 6144)] = kernel_3[(((((floordiv(blockIdx.x, 7)*589824) + (floordiv((threadIdx.x_2 + 6144), 72)*4608)) + cse_var_1) + (floormod((floordiv(threadIdx.x_2, 3) + 8), 24)*3)) + floormod(threadIdx.x_2, 3))]
+            attr [IterVar(threadIdx.x_2, (nullptr), "ThreadIndex", "threadIdx.x")] "thread_extent" = 64;
+            kernel.shared_1[(threadIdx.x_2 + 6208)] = kernel_3[(((((floordiv(blockIdx.x, 7)*589824) + (floordiv((threadIdx.x_2 + 6208), 72)*4608)) + cse_var_1) + (floordiv(floormod((threadIdx.x_2 + 16), 72), 3)*3)) + floormod((threadIdx.x_2 + 1), 3))]
+            attr [IterVar(threadIdx.x_2, (nullptr), "ThreadIndex", "threadIdx.x")] "thread_extent" = 64;
+            kernel.shared_1[(threadIdx.x_2 + 6272)] = kernel_3[(((((floordiv(blockIdx.x, 7)*589824) + (floordiv((threadIdx.x_2 + 6272), 72)*4608)) + cse_var_1) + (floordiv(floormod((threadIdx.x_2 + 8), 72), 3)*3)) + floormod((threadIdx.x_2 + 2), 3))]
+            attr [IterVar(threadIdx.x_2, (nullptr), "ThreadIndex", "threadIdx.x")] "thread_extent" = 64;
+            kernel.shared_1[(threadIdx.x_2 + 6336)] = kernel_3[((((floordiv(blockIdx.x, 7)*589824) + cse_var_1) + threadIdx.x_2) + 405504)]
+            attr [IterVar(threadIdx.x_2, (nullptr), "ThreadIndex", "threadIdx.x")] "thread_extent" = 64;
+            kernel.shared_1[(threadIdx.x_2 + 6400)] = kernel_3[(((((floordiv(blockIdx.x, 7)*589824) + (floordiv((threadIdx.x_2 + 6400), 72)*4608)) + cse_var_1) + (floordiv(floormod((threadIdx.x_2 + 64), 72), 3)*3)) + floormod((threadIdx.x_2 + 1), 3))]
+            attr [IterVar(threadIdx.x_2, (nullptr), "ThreadIndex", "threadIdx.x")] "thread_extent" = 64;
+            kernel.shared_1[(threadIdx.x_2 + 6464)] = kernel_3[(((((floordiv(blockIdx.x, 7)*589824) + (floordiv((threadIdx.x_2 + 6464), 72)*4608)) + cse_var_1) + (floordiv(floormod((threadIdx.x_2 + 56), 72), 3)*3)) + floormod((threadIdx.x_2 + 2), 3))]
+            attr [IterVar(threadIdx.x_2, (nullptr), "ThreadIndex", "threadIdx.x")] "thread_extent" = 64;
+            kernel.shared_1[(threadIdx.x_2 + 6528)] = kernel_3[(((((floordiv(blockIdx.x, 7)*589824) + (floordiv((threadIdx.x_2 + 6528), 72)*4608)) + cse_var_1) + (floormod((floordiv(threadIdx.x_2, 3) + 16), 24)*3)) + floormod(threadIdx.x_2, 3))]
+            attr [IterVar(threadIdx.x_2, (nullptr), "ThreadIndex", "threadIdx.x")] "thread_extent" = 64;
+            kernel.shared_1[(threadIdx.x_2 + 6592)] = kernel_3[(((((floordiv(blockIdx.x, 7)*589824) + (floordiv((threadIdx.x_2 + 6592), 72)*4608)) + cse_var_1) + (floordiv(floormod((threadIdx.x_2 + 40), 72), 3)*3)) + floormod((threadIdx.x_2 + 1), 3))]
+            attr [IterVar(threadIdx.x_2, (nullptr), "ThreadIndex", "threadIdx.x")] "thread_extent" = 64;
+            kernel.shared_1[(threadIdx.x_2 + 6656)] = kernel_3[(((((floordiv(blockIdx.x, 7)*589824) + (floordiv((threadIdx.x_2 + 6656), 72)*4608)) + cse_var_1) + (floordiv(floormod((threadIdx.x_2 + 32), 72), 3)*3)) + floormod((threadIdx.x_2 + 2), 3))]
+            attr [IterVar(threadIdx.x_2, (nullptr), "ThreadIndex", "threadIdx.x")] "thread_extent" = 64;
+            kernel.shared_1[(threadIdx.x_2 + 6720)] = kernel_3[(((((floordiv(blockIdx.x, 7)*589824) + (floordiv((threadIdx.x_2 + 6720), 72)*4608)) + cse_var_1) + (floormod((floordiv(threadIdx.x_2, 3) + 8), 24)*3)) + floormod(threadIdx.x_2, 3))]
+            attr [IterVar(threadIdx.x_2, (nullptr), "ThreadIndex", "threadIdx.x")] "thread_extent" = 64;
+            kernel.shared_1[(threadIdx.x_2 + 6784)] = kernel_3[(((((floordiv(blockIdx.x, 7)*589824) + (floordiv((threadIdx.x_2 + 6784), 72)*4608)) + cse_var_1) + (floordiv(floormod((threadIdx.x_2 + 16), 72), 3)*3)) + floormod((threadIdx.x_2 + 1), 3))]
+            attr [IterVar(threadIdx.x_2, (nullptr), "ThreadIndex", "threadIdx.x")] "thread_extent" = 64;
+            kernel.shared_1[(threadIdx.x_2 + 6848)] = kernel_3[(((((floordiv(blockIdx.x, 7)*589824) + (floordiv((threadIdx.x_2 + 6848), 72)*4608)) + cse_var_1) + (floordiv(floormod((threadIdx.x_2 + 8), 72), 3)*3)) + floormod((threadIdx.x_2 + 2), 3))]
+            attr [IterVar(threadIdx.x_2, (nullptr), "ThreadIndex", "threadIdx.x")] "thread_extent" = 64;
+            kernel.shared_1[(threadIdx.x_2 + 6912)] = kernel_3[((((floordiv(blockIdx.x, 7)*589824) + cse_var_1) + threadIdx.x_2) + 442368)]
+            attr [IterVar(threadIdx.x_2, (nullptr), "ThreadIndex", "threadIdx.x")] "thread_extent" = 64;
+            kernel.shared_1[(threadIdx.x_2 + 6976)] = kernel_3[(((((floordiv(blockIdx.x, 7)*589824) + (floordiv((threadIdx.x_2 + 6976), 72)*4608)) + cse_var_1) + (floordiv(floormod((threadIdx.x_2 + 64), 72), 3)*3)) + floormod((threadIdx.x_2 + 1), 3))]
+            attr [IterVar(threadIdx.x_2, (nullptr), "ThreadIndex", "threadIdx.x")] "thread_extent" = 64;
+            kernel.shared_1[(threadIdx.x_2 + 7040)] = kernel_3[(((((floordiv(blockIdx.x, 7)*589824) + (floordiv((threadIdx.x_2 + 7040), 72)*4608)) + cse_var_1) + (floordiv(floormod((threadIdx.x_2 + 56), 72), 3)*3)) + floormod((threadIdx.x_2 + 2), 3))]
+            attr [IterVar(threadIdx.x_2, (nullptr), "ThreadIndex", "threadIdx.x")] "thread_extent" = 64;
+            kernel.shared_1[(threadIdx.x_2 + 7104)] = kernel_3[(((((floordiv(blockIdx.x, 7)*589824) + (floordiv((threadIdx.x_2 + 7104), 72)*4608)) + cse_var_1) + (floormod((floordiv(threadIdx.x_2, 3) + 16), 24)*3)) + floormod(threadIdx.x_2, 3))]
+            attr [IterVar(threadIdx.x_2, (nullptr), "ThreadIndex", "threadIdx.x")] "thread_extent" = 64;
+            kernel.shared_1[(threadIdx.x_2 + 7168)] = kernel_3[(((((floordiv(blockIdx.x, 7)*589824) + (floordiv((threadIdx.x_2 + 7168), 72)*4608)) + cse_var_1) + (floordiv(floormod((threadIdx.x_2 + 40), 72), 3)*3)) + floormod((threadIdx.x_2 + 1), 3))]
+            attr [IterVar(threadIdx.x_2, (nullptr), "ThreadIndex", "threadIdx.x")] "thread_extent" = 64;
+            kernel.shared_1[(threadIdx.x_2 + 7232)] = kernel_3[(((((floordiv(blockIdx.x, 7)*589824) + (floordiv((threadIdx.x_2 + 7232), 72)*4608)) + cse_var_1) + (floordiv(floormod((threadIdx.x_2 + 32), 72), 3)*3)) + floormod((threadIdx.x_2 + 2), 3))]
+            attr [IterVar(threadIdx.x_2, (nullptr), "ThreadIndex", "threadIdx.x")] "thread_extent" = 64;
+            kernel.shared_1[(threadIdx.x_2 + 7296)] = kernel_3[(((((floordiv(blockIdx.x, 7)*589824) + (floordiv((threadIdx.x_2 + 7296), 72)*4608)) + cse_var_1) + (floormod((floordiv(threadIdx.x_2, 3) + 8), 24)*3)) + floormod(threadIdx.x_2, 3))]
+            attr [IterVar(threadIdx.x_2, (nullptr), "ThreadIndex", "threadIdx.x")] "thread_extent" = 64;
+            kernel.shared_1[(threadIdx.x_2 + 7360)] = kernel_3[(((((floordiv(blockIdx.x, 7)*589824) + (floordiv((threadIdx.x_2 + 7360), 72)*4608)) + cse_var_1) + (floordiv(floormod((threadIdx.x_2 + 16), 72), 3)*3)) + floormod((threadIdx.x_2 + 1), 3))]
+            attr [IterVar(threadIdx.x_2, (nullptr), "ThreadIndex", "threadIdx.x")] "thread_extent" = 64;
+            kernel.shared_1[(threadIdx.x_2 + 7424)] = kernel_3[(((((floordiv(blockIdx.x, 7)*589824) + (floordiv((threadIdx.x_2 + 7424), 72)*4608)) + cse_var_1) + (floordiv(floormod((threadIdx.x_2 + 8), 72), 3)*3)) + floormod((threadIdx.x_2 + 2), 3))]
+            attr [IterVar(threadIdx.x_2, (nullptr), "ThreadIndex", "threadIdx.x")] "thread_extent" = 64;
+            kernel.shared_1[(threadIdx.x_2 + 7488)] = kernel_3[((((floordiv(blockIdx.x, 7)*589824) + cse_var_1) + threadIdx.x_2) + 479232)]
+            attr [IterVar(threadIdx.x_2, (nullptr), "ThreadIndex", "threadIdx.x")] "thread_extent" = 64;
+            kernel.shared_1[(threadIdx.x_2 + 7552)] = kernel_3[(((((floordiv(blockIdx.x, 7)*589824) + (floordiv((threadIdx.x_2 + 7552), 72)*4608)) + cse_var_1) + (floordiv(floormod((threadIdx.x_2 + 64), 72), 3)*3)) + floormod((threadIdx.x_2 + 1), 3))]
+            attr [IterVar(threadIdx.x_2, (nullptr), "ThreadIndex", "threadIdx.x")] "thread_extent" = 64;
+            kernel.shared_1[(threadIdx.x_2 + 7616)] = kernel_3[(((((floordiv(blockIdx.x, 7)*589824) + (floordiv((threadIdx.x_2 + 7616), 72)*4608)) + cse_var_1) + (floordiv(floormod((threadIdx.x_2 + 56), 72), 3)*3)) + floormod((threadIdx.x_2 + 2), 3))]
+            attr [IterVar(threadIdx.x_2, (nullptr), "ThreadIndex", "threadIdx.x")] "thread_extent" = 64;
+            kernel.shared_1[(threadIdx.x_2 + 7680)] = kernel_3[(((((floordiv(blockIdx.x, 7)*589824) + (floordiv((threadIdx.x_2 + 7680), 72)*4608)) + cse_var_1) + (floormod((floordiv(threadIdx.x_2, 3) + 16), 24)*3)) + floormod(threadIdx.x_2, 3))]
+            attr [IterVar(threadIdx.x_2, (nullptr), "ThreadIndex", "threadIdx.x")] "thread_extent" = 64;
+            kernel.shared_1[(threadIdx.x_2 + 7744)] = kernel_3[(((((floordiv(blockIdx.x, 7)*589824) + (floordiv((threadIdx.x_2 + 7744), 72)*4608)) + cse_var_1) + (floordiv(floormod((threadIdx.x_2 + 40), 72), 3)*3)) + floormod((threadIdx.x_2 + 1), 3))]
+            attr [IterVar(threadIdx.x_2, (nullptr), "ThreadIndex", "threadIdx.x")] "thread_extent" = 64;
+            kernel.shared_1[(threadIdx.x_2 + 7808)] = kernel_3[(((((floordiv(blockIdx.x, 7)*589824) + (floordiv((threadIdx.x_2 + 7808), 72)*4608)) + cse_var_1) + (floordiv(floormod((threadIdx.x_2 + 32), 72), 3)*3)) + floormod((threadIdx.x_2 + 2), 3))]
+            attr [IterVar(threadIdx.x_2, (nullptr), "ThreadIndex", "threadIdx.x")] "thread_extent" = 64;
+            kernel.shared_1[(threadIdx.x_2 + 7872)] = kernel_3[(((((floordiv(blockIdx.x, 7)*589824) + (floordiv((threadIdx.x_2 + 7872), 72)*4608)) + cse_var_1) + (floormod((floordiv(threadIdx.x_2, 3) + 8), 24)*3)) + floormod(threadIdx.x_2, 3))]
+            attr [IterVar(threadIdx.x_2, (nullptr), "ThreadIndex", "threadIdx.x")] "thread_extent" = 64;
+            kernel.shared_1[(threadIdx.x_2 + 7936)] = kernel_3[(((((floordiv(blockIdx.x, 7)*589824) + (floordiv((threadIdx.x_2 + 7936), 72)*4608)) + cse_var_1) + (floordiv(floormod((threadIdx.x_2 + 16), 72), 3)*3)) + floormod((threadIdx.x_2 + 1), 3))]
+            attr [IterVar(threadIdx.x_2, (nullptr), "ThreadIndex", "threadIdx.x")] "thread_extent" = 64;
+            kernel.shared_1[(threadIdx.x_2 + 8000)] = kernel_3[(((((floordiv(blockIdx.x, 7)*589824) + (floordiv((threadIdx.x_2 + 8000), 72)*4608)) + cse_var_1) + (floordiv(floormod((threadIdx.x_2 + 8), 72), 3)*3)) + floormod((threadIdx.x_2 + 2), 3))]
+            attr [IterVar(threadIdx.x_2, (nullptr), "ThreadIndex", "threadIdx.x")] "thread_extent" = 64;
+            kernel.shared_1[(threadIdx.x_2 + 8064)] = kernel_3[((((floordiv(blockIdx.x, 7)*589824) + cse_var_1) + threadIdx.x_2) + 516096)]
+            attr [IterVar(threadIdx.x_2, (nullptr), "ThreadIndex", "threadIdx.x")] "thread_extent" = 64;
+            kernel.shared_1[(threadIdx.x_2 + 8128)] = kernel_3[(((((floordiv(blockIdx.x, 7)*589824) + (floordiv((threadIdx.x_2 + 8128), 72)*4608)) + cse_var_1) + (floordiv(floormod((threadIdx.x_2 + 64), 72), 3)*3)) + floormod((threadIdx.x_2 + 1), 3))]
+            attr [IterVar(threadIdx.x_2, (nullptr), "ThreadIndex", "threadIdx.x")] "thread_extent" = 64;
+            kernel.shared_1[(threadIdx.x_2 + 8192)] = kernel_3[(((((floordiv(blockIdx.x, 7)*589824) + (floordiv((threadIdx.x_2 + 8192), 72)*4608)) + cse_var_1) + (floordiv(floormod((threadIdx.x_2 + 56), 72), 3)*3)) + floormod((threadIdx.x_2 + 2), 3))]
+            attr [IterVar(threadIdx.x_2, (nullptr), "ThreadIndex", "threadIdx.x")] "thread_extent" = 64;
+            kernel.shared_1[(threadIdx.x_2 + 8256)] = kernel_3[(((((floordiv(blockIdx.x, 7)*589824) + (floordiv((threadIdx.x_2 + 8256), 72)*4608)) + cse_var_1) + (floormod((floordiv(threadIdx.x_2, 3) + 16), 24)*3)) + floormod(threadIdx.x_2, 3))]
+            attr [IterVar(threadIdx.x_2, (nullptr), "ThreadIndex", "threadIdx.x")] "thread_extent" = 64;
+            kernel.shared_1[(threadIdx.x_2 + 8320)] = kernel_3[(((((floordiv(blockIdx.x, 7)*589824) + (floordiv((threadIdx.x_2 + 8320), 72)*4608)) + cse_var_1) + (floordiv(floormod((threadIdx.x_2 + 40), 72), 3)*3)) + floormod((threadIdx.x_2 + 1), 3))]
+            attr [IterVar(threadIdx.x_2, (nullptr), "ThreadIndex", "threadIdx.x")] "thread_extent" = 64;
+            kernel.shared_1[(threadIdx.x_2 + 8384)] = kernel_3[(((((floordiv(blockIdx.x, 7)*589824) + (floordiv((threadIdx.x_2 + 8384), 72)*4608)) + cse_var_1) + (floordiv(floormod((threadIdx.x_2 + 32), 72), 3)*3)) + floormod((threadIdx.x_2 + 2), 3))]
+            attr [IterVar(threadIdx.x_2, (nullptr), "ThreadIndex", "threadIdx.x")] "thread_extent" = 64;
+            kernel.shared_1[(threadIdx.x_2 + 8448)] = kernel_3[(((((floordiv(blockIdx.x, 7)*589824) + (floordiv((threadIdx.x_2 + 8448), 72)*4608)) + cse_var_1) + (floormod((floordiv(threadIdx.x_2, 3) + 8), 24)*3)) + floormod(threadIdx.x_2, 3))]
+            attr [IterVar(threadIdx.x_2, (nullptr), "ThreadIndex", "threadIdx.x")] "thread_extent" = 64;
+            kernel.shared_1[(threadIdx.x_2 + 8512)] = kernel_3[(((((floordiv(blockIdx.x, 7)*589824) + (floordiv((threadIdx.x_2 + 8512), 72)*4608)) + cse_var_1) + (floordiv(floormod((threadIdx.x_2 + 16), 72), 3)*3)) + floormod((threadIdx.x_2 + 1), 3))]
+            attr [IterVar(threadIdx.x_2, (nullptr), "ThreadIndex", "threadIdx.x")] "thread_extent" = 64;
+            kernel.shared_1[(threadIdx.x_2 + 8576)] = kernel_3[(((((floordiv(blockIdx.x, 7)*589824) + (floordiv((threadIdx.x_2 + 8576), 72)*4608)) + cse_var_1) + (floordiv(floormod((threadIdx.x_2 + 8), 72), 3)*3)) + floormod((threadIdx.x_2 + 2), 3))]
+            attr [IterVar(threadIdx.x_2, (nullptr), "ThreadIndex", "threadIdx.x")] "thread_extent" = 64;
+            kernel.shared_1[(threadIdx.x_2 + 8640)] = kernel_3[((((floordiv(blockIdx.x, 7)*589824) + cse_var_1) + threadIdx.x_2) + 552960)]
+            attr [IterVar(threadIdx.x_2, (nullptr), "ThreadIndex", "threadIdx.x")] "thread_extent" = 64;
+            kernel.shared_1[(threadIdx.x_2 + 8704)] = kernel_3[(((((floordiv(blockIdx.x, 7)*589824) + (floordiv((threadIdx.x_2 + 8704), 72)*4608)) + cse_var_1) + (floordiv(floormod((threadIdx.x_2 + 64), 72), 3)*3)) + floormod((threadIdx.x_2 + 1), 3))]
+            attr [IterVar(threadIdx.x_2, (nullptr), "ThreadIndex", "threadIdx.x")] "thread_extent" = 64;
+            kernel.shared_1[(threadIdx.x_2 + 8768)] = kernel_3[(((((floordiv(blockIdx.x, 7)*589824) + (floordiv((threadIdx.x_2 + 8768), 72)*4608)) + cse_var_1) + (floordiv(floormod((threadIdx.x_2 + 56), 72), 3)*3)) + floormod((threadIdx.x_2 + 2), 3))]
+            attr [IterVar(threadIdx.x_2, (nullptr), "ThreadIndex", "threadIdx.x")] "thread_extent" = 64;
+            kernel.shared_1[(threadIdx.x_2 + 8832)] = kernel_3[(((((floordiv(blockIdx.x, 7)*589824) + (floordiv((threadIdx.x_2 + 8832), 72)*4608)) + cse_var_1) + (floormod((floordiv(threadIdx.x_2, 3) + 16), 24)*3)) + floormod(threadIdx.x_2, 3))]
+            attr [IterVar(threadIdx.x_2, (nullptr), "ThreadIndex", "threadIdx.x")] "thread_extent" = 64;
+            kernel.shared_1[(threadIdx.x_2 + 8896)] = kernel_3[(((((floordiv(blockIdx.x, 7)*589824) + (floordiv((threadIdx.x_2 + 8896), 72)*4608)) + cse_var_1) + (floordiv(floormod((threadIdx.x_2 + 40), 72), 3)*3)) + floormod((threadIdx.x_2 + 1), 3))]
+            attr [IterVar(threadIdx.x_2, (nullptr), "ThreadIndex", "threadIdx.x")] "thread_extent" = 64;
+            kernel.shared_1[(threadIdx.x_2 + 8960)] = kernel_3[(((((floordiv(blockIdx.x, 7)*589824) + (floordiv((threadIdx.x_2 + 8960), 72)*4608)) + cse_var_1) + (floordiv(floormod((threadIdx.x_2 + 32), 72), 3)*3)) + floormod((threadIdx.x_2 + 2), 3))]
+            attr [IterVar(threadIdx.x_2, (nullptr), "ThreadIndex", "threadIdx.x")] "thread_extent" = 64;
+            kernel.shared_1[(threadIdx.x_2 + 9024)] = kernel_3[(((((floordiv(blockIdx.x, 7)*589824) + (floordiv((threadIdx.x_2 + 9024), 72)*4608)) + cse_var_1) + (floormod((floordiv(threadIdx.x_2, 3) + 8), 24)*3)) + floormod(threadIdx.x_2, 3))]
+            attr [IterVar(threadIdx.x_2, (nullptr), "ThreadIndex", "threadIdx.x")] "thread_extent" = 64;
+            kernel.shared_1[(threadIdx.x_2 + 9088)] = kernel_3[(((((floordiv(blockIdx.x, 7)*589824) + (floordiv((threadIdx.x_2 + 9088), 72)*4608)) + cse_var_1) + (floordiv(floormod((threadIdx.x_2 + 16), 72), 3)*3)) + floormod((threadIdx.x_2 + 1), 3))]
+            attr [IterVar(threadIdx.x_2, (nullptr), "ThreadIndex", "threadIdx.x")] "thread_extent" = 64;
+            kernel.shared_1[(threadIdx.x_2 + 9152)] = kernel_3[(((((floordiv(blockIdx.x, 7)*589824) + (floordiv((threadIdx.x_2 + 9152), 72)*4608)) + cse_var_1) + (floordiv(floormod((threadIdx.x_2 + 8), 72), 3)*3)) + floormod((threadIdx.x_2 + 2), 3))]
+            conv2d_nchw_1[0] = (conv2d_nchw_1[0] + (pad_temp.shared_1[0]*kernel.shared_1[(threadIdx.x*72)]))
+            conv2d_nchw_1[1] = (conv2d_nchw_1[1] + (pad_temp.shared_1[1]*kernel.shared_1[(threadIdx.x*72)]))
+            conv2d_nchw_1[2] = (conv2d_nchw_1[2] + (pad_temp.shared_1[2]*kernel.shared_1[(threadIdx.x*72)]))
+            conv2d_nchw_1[3] = (conv2d_nchw_1[3] + (pad_temp.shared_1[3]*kernel.shared_1[(threadIdx.x*72)]))
+            conv2d_nchw_1[4] = (conv2d_nchw_1[4] + (pad_temp.shared_1[4]*kernel.shared_1[(threadIdx.x*72)]))
+            conv2d_nchw_1[5] = (conv2d_nchw_1[5] + (pad_temp.shared_1[5]*kernel.shared_1[(threadIdx.x*72)]))
+            conv2d_nchw_1[6] = (conv2d_nchw_1[6] + (pad_temp.shared_1[6]*kernel.shared_1[(threadIdx.x*72)]))
+            conv2d_nchw_1[7] = (conv2d_nchw_1[7] + (pad_temp.shared_1[0]*kernel.shared_1[((threadIdx.x*72) + 4608)]))
+            conv2d_nchw_1[8] = (conv2d_nchw_1[8] + (pad_temp.shared_1[1]*kernel.shared_1[((threadIdx.x*72) + 4608)]))
+            conv2d_nchw_1[9] = (conv2d_nchw_1[9] + (pad_temp.shared_1[2]*kernel.shared_1[((threadIdx.x*72) + 4608)]))
+            conv2d_nchw_1[10] = (conv2d_nchw_1[10] + (pad_temp.shared_1[3]*kernel.shared_1[((threadIdx.x*72) + 4608)]))
+            conv2d_nchw_1[11] = (conv2d_nchw_1[11] + (pad_temp.shared_1[4]*kernel.shared_1[((threadIdx.x*72) + 4608)]))
+            conv2d_nchw_1[12] = (conv2d_nchw_1[12] + (pad_temp.shared_1[5]*kernel.shared_1[((threadIdx.x*72) + 4608)]))
+            conv2d_nchw_1[13] = (conv2d_nchw_1[13] + (pad_temp.shared_1[6]*kernel.shared_1[((threadIdx.x*72) + 4608)]))
+            conv2d_nchw_1[0] = (conv2d_nchw_1[0] + (pad_temp.shared_1[1]*kernel.shared_1[((threadIdx.x*72) + 1)]))
+            conv2d_nchw_1[1] = (conv2d_nchw_1[1] + (pad_temp.shared_1[2]*kernel.shared_1[((threadIdx.x*72) + 1)]))
+            conv2d_nchw_1[2] = (conv2d_nchw_1[2] + (pad_temp.shared_1[3]*kernel.shared_1[((threadIdx.x*72) + 1)]))
+            conv2d_nchw_1[3] = (conv2d_nchw_1[3] + (pad_temp.shared_1[4]*kernel.shared_1[((threadIdx.x*72) + 1)]))
+            conv2d_nchw_1[4] = (conv2d_nchw_1[4] + (pad_temp.shared_1[5]*kernel.shared_1[((threadIdx.x*72) + 1)]))
+            conv2d_nchw_1[5] = (conv2d_nchw_1[5] + (pad_temp.shared_1[6]*kernel.shared_1[((threadIdx.x*72) + 1)]))
+            conv2d_nchw_1[6] = (conv2d_nchw_1[6] + (pad_temp.shared_1[7]*kernel.shared_1[((threadIdx.x*72) + 1)]))
+            conv2d_nchw_1[7] = (conv2d_nchw_1[7] + (pad_temp.shared_1[1]*kernel.shared_1[((threadIdx.x*72) + 4609)]))
+            conv2d_nchw_1[8] = (conv2d_nchw_1[8] + (pad_temp.shared_1[2]*kernel.shared_1[((threadIdx.x*72) + 4609)]))
+            conv2d_nchw_1[9] = (conv2d_nchw_1[9] + (pad_temp.shared_1[3]*kernel.shared_1[((threadIdx.x*72) + 4609)]))
+            conv2d_nchw_1[10] = (conv2d_nchw_1[10] + (pad_temp.shared_1[4]*kernel.shared_1[((threadIdx.x*72) + 4609)]))
+            conv2d_nchw_1[11] = (conv2d_nchw_1[11] + (pad_temp.shared_1[5]*kernel.shared_1[((threadIdx.x*72) + 4609)]))
+            conv2d_nchw_1[12] = (conv2d_nchw_1[12] + (pad_temp.shared_1[6]*kernel.shared_1[((threadIdx.x*72) + 4609)]))
+            conv2d_nchw_1[13] = (conv2d_nchw_1[13] + (pad_temp.shared_1[7]*kernel.shared_1[((threadIdx.x*72) + 4609)]))
+            conv2d_nchw_1[0] = (conv2d_nchw_1[0] + (pad_temp.shared_1[2]*kernel.shared_1[((threadIdx.x*72) + 2)]))
+            conv2d_nchw_1[1] = (conv2d_nchw_1[1] + (pad_temp.shared_1[3]*kernel.shared_1[((threadIdx.x*72) + 2)]))
+            conv2d_nchw_1[2] = (conv2d_nchw_1[2] + (pad_temp.shared_1[4]*kernel.shared_1[((threadIdx.x*72) + 2)]))
+            conv2d_nchw_1[3] = (conv2d_nchw_1[3] + (pad_temp.shared_1[5]*kernel.shared_1[((threadIdx.x*72) + 2)]))
+            conv2d_nchw_1[4] = (conv2d_nchw_1[4] + (pad_temp.shared_1[6]*kernel.shared_1[((threadIdx.x*72) + 2)]))
+            conv2d_nchw_1[5] = (conv2d_nchw_1[5] + (pad_temp.shared_1[7]*kernel.shared_1[((threadIdx.x*72) + 2)]))
+            conv2d_nchw_1[6] = (conv2d_nchw_1[6] + (pad_temp.shared_1[8]*kernel.shared_1[((threadIdx.x*72) + 2)]))
+            conv2d_nchw_1[7] = (conv2d_nchw_1[7] + (pad_temp.shared_1[2]*kernel.shared_1[((threadIdx.x*72) + 4610)]))
+            conv2d_nchw_1[8] = (conv2d_nchw_1[8] + (pad_temp.shared_1[3]*kernel.shared_1[((threadIdx.x*72) + 4610)]))
+            conv2d_nchw_1[9] = (conv2d_nchw_1[9] + (pad_temp.shared_1[4]*kernel.shared_1[((threadIdx.x*72) + 4610)]))
+            conv2d_nchw_1[10] = (conv2d_nchw_1[10] + (pad_temp.shared_1[5]*kernel.shared_1[((threadIdx.x*72) + 4610)]))
+            conv2d_nchw_1[11] = (conv2d_nchw_1[11] + (pad_temp.shared_1[6]*kernel.shared_1[((threadIdx.x*72) + 4610)]))
+            conv2d_nchw_1[12] = (conv2d_nchw_1[12] + (pad_temp.shared_1[7]*kernel.shared_1[((threadIdx.x*72) + 4610)]))
+            conv2d_nchw_1[13] = (conv2d_nchw_1[13] + (pad_temp.shared_1[8]*kernel.shared_1[((threadIdx.x*72) + 4610)]))
+            conv2d_nchw_1[0] = (conv2d_nchw_1[0] + (pad_temp.shared_1[9]*kernel.shared_1[((threadIdx.x*72) + 3)]))
+            conv2d_nchw_1[1] = (conv2d_nchw_1[1] + (pad_temp.shared_1[10]*kernel.shared_1[((threadIdx.x*72) + 3)]))
+            conv2d_nchw_1[2] = (conv2d_nchw_1[2] + (pad_temp.shared_1[11]*kernel.shared_1[((threadIdx.x*72) + 3)]))
+            conv2d_nchw_1[3] = (conv2d_nchw_1[3] + (pad_temp.shared_1[12]*kernel.shared_1[((threadIdx.x*72) + 3)]))
+            conv2d_nchw_1[4] = (conv2d_nchw_1[4] + (pad_temp.shared_1[13]*kernel.shared_1[((threadIdx.x*72) + 3)]))
+            conv2d_nchw_1[5] = (conv2d_nchw_1[5] + (pad_temp.shared_1[14]*kernel.shared_1[((threadIdx.x*72) + 3)]))
+            conv2d_nchw_1[6] = (conv2d_nchw_1[6] + (pad_temp.shared_1[15]*kernel.shared_1[((threadIdx.x*72) + 3)]))
+            conv2d_nchw_1[7] = (conv2d_nchw_1[7] + (pad_temp.shared_1[9]*kernel.shared_1[((threadIdx.x*72) + 4611)]))
+            conv2d_nchw_1[8] = (conv2d_nchw_1[8] + (pad_temp.shared_1[10]*kernel.shared_1[((threadIdx.x*72) + 4611)]))
+            conv2d_nchw_1[9] = (conv2d_nchw_1[9] + (pad_temp.shared_1[11]*kernel.shared_1[((threadIdx.x*72) + 4611)]))
+            conv2d_nchw_1[10] = (conv2d_nchw_1[10] + (pad_temp.shared_1[12]*kernel.shared_1[((threadIdx.x*72) + 4611)]))
+            conv2d_nchw_1[11] = (conv2d_nchw_1[11] + (pad_temp.shared_1[13]*kernel.shared_1[((threadIdx.x*72) + 4611)]))
+            conv2d_nchw_1[12] = (conv2d_nchw_1[12] + (pad_temp.shared_1[14]*kernel.shared_1[((threadIdx.x*72) + 4611)]))
+            conv2d_nchw_1[13] = (conv2d_nchw_1[13] + (pad_temp.shared_1[15]*kernel.shared_1[((threadIdx.x*72) + 4611)]))
+            conv2d_nchw_1[0] = (conv2d_nchw_1[0] + (pad_temp.shared_1[10]*kernel.shared_1[((threadIdx.x*72) + 4)]))
+            conv2d_nchw_1[1] = (conv2d_nchw_1[1] + (pad_temp.shared_1[11]*kernel.shared_1[((threadIdx.x*72) + 4)]))
+            conv2d_nchw_1[2] = (conv2d_nchw_1[2] + (pad_temp.shared_1[12]*kernel.shared_1[((threadIdx.x*72) + 4)]))
+            conv2d_nchw_1[3] = (conv2d_nchw_1[3] + (pad_temp.shared_1[13]*kernel.shared_1[((threadIdx.x*72) + 4)]))
+            conv2d_nchw_1[4] = (conv2d_nchw_1[4] + (pad_temp.shared_1[14]*kernel.shared_1[((threadIdx.x*72) + 4)]))
+            conv2d_nchw_1[5] = (conv2d_nchw_1[5] + (pad_temp.shared_1[15]*kernel.shared_1[((threadIdx.x*72) + 4)]))
+            conv2d_nchw_1[6] = (conv2d_nchw_1[6] + (pad_temp.shared_1[16]*kernel.shared_1[((threadIdx.x*72) + 4)]))
+            conv2d_nchw_1[7] = (conv2d_nchw_1[7] + (pad_temp.shared_1[10]*kernel.shared_1[((threadIdx.x*72) + 4612)]))
+            conv2d_nchw_1[8] = (conv2d_nchw_1[8] + (pad_temp.shared_1[11]*kernel.shared_1[((threadIdx.x*72) + 4612)]))
+            conv2d_nchw_1[9] = (conv2d_nchw_1[9] + (pad_temp.shared_1[12]*kernel.shared_1[((threadIdx.x*72) + 4612)]))
+            conv2d_nchw_1[10] = (conv2d_nchw_1[10] + (pad_temp.shared_1[13]*kernel.shared_1[((threadIdx.x*72) + 4612)]))
+            conv2d_nchw_1[11] = (conv2d_nchw_1[11] + (pad_temp.shared_1[14]*kernel.shared_1[((threadIdx.x*72) + 4612)]))
+            conv2d_nchw_1[12] = (conv2d_nchw_1[12] + (pad_temp.shared_1[15]*kernel.shared_1[((threadIdx.x*72) + 4612)]))
+            conv2d_nchw_1[13] = (conv2d_nchw_1[13] + (pad_temp.shared_1[16]*kernel.shared_1[((threadIdx.x*72) + 4612)]))
+            conv2d_nchw_1[0] = (conv2d_nchw_1[0] + (pad_temp.shared_1[11]*kernel.shared_1[((threadIdx.x*72) + 5)]))
+            conv2d_nchw_1[1] = (conv2d_nchw_1[1] + (pad_temp.shared_1[12]*kernel.shared_1[((threadIdx.x*72) + 5)]))
+            conv2d_nchw_1[2] = (conv2d_nchw_1[2] + (pad_temp.shared_1[13]*kernel.shared_1[((threadIdx.x*72) + 5)]))
+            conv2d_nchw_1[3] = (conv2d_nchw_1[3] + (pad_temp.shared_1[14]*kernel.shared_1[((threadIdx.x*72) + 5)]))
+            conv2d_nchw_1[4] = (conv2d_nchw_1[4] + (pad_temp.shared_1[15]*kernel.shared_1[((threadIdx.x*72) + 5)]))
+            conv2d_nchw_1[5] = (conv2d_nchw_1[5] + (pad_temp.shared_1[16]*kernel.shared_1[((threadIdx.x*72) + 5)]))
+            conv2d_nchw_1[6] = (conv2d_nchw_1[6] + (pad_temp.shared_1[17]*kernel.shared_1[((threadIdx.x*72) + 5)]))
+            conv2d_nchw_1[7] = (conv2d_nchw_1[7] + (pad_temp.shared_1[11]*kernel.shared_1[((threadIdx.x*72) + 4613)]))
+            conv2d_nchw_1[8] = (conv2d_nchw_1[8] + (pad_temp.shared_1[12]*kernel.shared_1[((threadIdx.x*72) + 4613)]))
+            conv2d_nchw_1[9] = (conv2d_nchw_1[9] + (pad_temp.shared_1[13]*kernel.shared_1[((threadIdx.x*72) + 4613)]))
+            conv2d_nchw_1[10] = (conv2d_nchw_1[10] + (pad_temp.shared_1[14]*kernel.shared_1[((threadIdx.x*72) + 4613)]))
+            conv2d_nchw_1[11] = (conv2d_nchw_1[11] + (pad_temp.shared_1[15]*kernel.shared_1[((threadIdx.x*72) + 4613)]))
+            conv2d_nchw_1[12] = (conv2d_nchw_1[12] + (pad_temp.shared_1[16]*kernel.shared_1[((threadIdx.x*72) + 4613)]))
+            conv2d_nchw_1[13] = (conv2d_nchw_1[13] + (pad_temp.shared_1[17]*kernel.shared_1[((threadIdx.x*72) + 4613)]))
+            conv2d_nchw_1[0] = (conv2d_nchw_1[0] + (pad_temp.shared_1[18]*kernel.shared_1[((threadIdx.x*72) + 6)]))
+            conv2d_nchw_1[1] = (conv2d_nchw_1[1] + (pad_temp.shared_1[19]*kernel.shared_1[((threadIdx.x*72) + 6)]))
+            conv2d_nchw_1[2] = (conv2d_nchw_1[2] + (pad_temp.shared_1[20]*kernel.shared_1[((threadIdx.x*72) + 6)]))
+            conv2d_nchw_1[3] = (conv2d_nchw_1[3] + (pad_temp.shared_1[21]*kernel.shared_1[((threadIdx.x*72) + 6)]))
+            conv2d_nchw_1[4] = (conv2d_nchw_1[4] + (pad_temp.shared_1[22]*kernel.shared_1[((threadIdx.x*72) + 6)]))
+            conv2d_nchw_1[5] = (conv2d_nchw_1[5] + (pad_temp.shared_1[23]*kernel.shared_1[((threadIdx.x*72) + 6)]))
+            conv2d_nchw_1[6] = (conv2d_nchw_1[6] + (pad_temp.shared_1[24]*kernel.shared_1[((threadIdx.x*72) + 6)]))
+            conv2d_nchw_1[7] = (conv2d_nchw_1[7] + (pad_temp.shared_1[18]*kernel.shared_1[((threadIdx.x*72) + 4614)]))
+            conv2d_nchw_1[8] = (conv2d_nchw_1[8] + (pad_temp.shared_1[19]*kernel.shared_1[((threadIdx.x*72) + 4614)]))
+            conv2d_nchw_1[9] = (conv2d_nchw_1[9] + (pad_temp.shared_1[20]*kernel.shared_1[((threadIdx.x*72) + 4614)]))
+            conv2d_nchw_1[10] = (conv2d_nchw_1[10] + (pad_temp.shared_1[21]*kernel.shared_1[((threadIdx.x*72) + 4614)]))
+            conv2d_nchw_1[11] = (conv2d_nchw_1[11] + (pad_temp.shared_1[22]*kernel.shared_1[((threadIdx.x*72) + 4614)]))
+            conv2d_nchw_1[12] = (conv2d_nchw_1[12] + (pad_temp.shared_1[23]*kernel.shared_1[((threadIdx.x*72) + 4614)]))
+            conv2d_nchw_1[13] = (conv2d_nchw_1[13] + (pad_temp.shared_1[24]*kernel.shared_1[((threadIdx.x*72) + 4614)]))
+            conv2d_nchw_1[0] = (conv2d_nchw_1[0] + (pad_temp.shared_1[19]*kernel.shared_1[((threadIdx.x*72) + 7)]))
+            conv2d_nchw_1[1] = (conv2d_nchw_1[1] + (pad_temp.shared_1[20]*kernel.shared_1[((threadIdx.x*72) + 7)]))
+            conv2d_nchw_1[2] = (conv2d_nchw_1[2] + (pad_temp.shared_1[21]*kernel.shared_1[((threadIdx.x*72) + 7)]))
+            conv2d_nchw_1[3] = (conv2d_nchw_1[3] + (pad_temp.shared_1[22]*kernel.shared_1[((threadIdx.x*72) + 7)]))
+            conv2d_nchw_1[4] = (conv2d_nchw_1[4] + (pad_temp.shared_1[23]*kernel.shared_1[((threadIdx.x*72) + 7)]))
+            conv2d_nchw_1[5] = (conv2d_nchw_1[5] + (pad_temp.shared_1[24]*kernel.shared_1[((threadIdx.x*72) + 7)]))
+            conv2d_nchw_1[6] = (conv2d_nchw_1[6] + (pad_temp.shared_1[25]*kernel.shared_1[((threadIdx.x*72) + 7)]))
+            conv2d_nchw_1[7] = (conv2d_nchw_1[7] + (pad_temp.shared_1[19]*kernel.shared_1[((threadIdx.x*72) + 4615)]))
+            conv2d_nchw_1[8] = (conv2d_nchw_1[8] + (pad_temp.shared_1[20]*kernel.shared_1[((threadIdx.x*72) + 4615)]))
+            conv2d_nchw_1[9] = (conv2d_nchw_1[9] + (pad_temp.shared_1[21]*kernel.shared_1[((threadIdx.x*72) + 4615)]))
+            conv2d_nchw_1[10] = (conv2d_nchw_1[10] + (pad_temp.shared_1[22]*kernel.shared_1[((threadIdx.x*72) + 4615)]))
+            conv2d_nchw_1[11] = (conv2d_nchw_1[11] + (pad_temp.shared_1[23]*kernel.shared_1[((threadIdx.x*72) + 4615)]))
+            conv2d_nchw_1[12] = (conv2d_nchw_1[12] + (pad_temp.shared_1[24]*kernel.shared_1[((threadIdx.x*72) + 4615)]))
+            conv2d_nchw_1[13] = (conv2d_nchw_1[13] + (pad_temp.shared_1[25]*kernel.shared_1[((threadIdx.x*72) + 4615)]))
+            conv2d_nchw_1[0] = (conv2d_nchw_1[0] + (pad_temp.shared_1[20]*kernel.shared_1[((threadIdx.x*72) + 8)]))
+            conv2d_nchw_1[1] = (conv2d_nchw_1[1] + (pad_temp.shared_1[21]*kernel.shared_1[((threadIdx.x*72) + 8)]))
+            conv2d_nchw_1[2] = (conv2d_nchw_1[2] + (pad_temp.shared_1[22]*kernel.shared_1[((threadIdx.x*72) + 8)]))
+            conv2d_nchw_1[3] = (conv2d_nchw_1[3] + (pad_temp.shared_1[23]*kernel.shared_1[((threadIdx.x*72) + 8)]))
+            conv2d_nchw_1[4] = (conv2d_nchw_1[4] + (pad_temp.shared_1[24]*kernel.shared_1[((threadIdx.x*72) + 8)]))
+            conv2d_nchw_1[5] = (conv2d_nchw_1[5] + (pad_temp.shared_1[25]*kernel.shared_1[((threadIdx.x*72) + 8)]))
+            conv2d_nchw_1[6] = (conv2d_nchw_1[6] + (pad_temp.shared_1[26]*kernel.shared_1[((threadIdx.x*72) + 8)]))
+            conv2d_nchw_1[7] = (conv2d_nchw_1[7] + (pad_temp.shared_1[20]*kernel.shared_1[((threadIdx.x*72) + 4616)]))
+            conv2d_nchw_1[8] = (conv2d_nchw_1[8] + (pad_temp.shared_1[21]*kernel.shared_1[((threadIdx.x*72) + 4616)]))
+            conv2d_nchw_1[9] = (conv2d_nchw_1[9] + (pad_temp.shared_1[22]*kernel.shared_1[((threadIdx.x*72) + 4616)]))
+            conv2d_nchw_1[10] = (conv2d_nchw_1[10] + (pad_temp.shared_1[23]*kernel.shared_1[((threadIdx.x*72) + 4616)]))
+            conv2d_nchw_1[11] = (conv2d_nchw_1[11] + (pad_temp.shared_1[24]*kernel.shared_1[((threadIdx.x*72) + 4616)]))
+            conv2d_nchw_1[12] = (conv2d_nchw_1[12] + (pad_temp.shared_1[25]*kernel.shared_1[((threadIdx.x*72) + 4616)]))
+            conv2d_nchw_1[13] = (conv2d_nchw_1[13] + (pad_temp.shared_1[26]*kernel.shared_1[((threadIdx.x*72) + 4616)]))
+            conv2d_nchw_1[0] = (conv2d_nchw_1[0] + (pad_temp.shared_1[27]*kernel.shared_1[((threadIdx.x*72) + 9)]))
+            conv2d_nchw_1[1] = (conv2d_nchw_1[1] + (pad_temp.shared_1[28]*kernel.shared_1[((threadIdx.x*72) + 9)]))
+            conv2d_nchw_1[2] = (conv2d_nchw_1[2] + (pad_temp.shared_1[29]*kernel.shared_1[((threadIdx.x*72) + 9)]))
+            conv2d_nchw_1[3] = (conv2d_nchw_1[3] + (pad_temp.shared_1[30]*kernel.shared_1[((threadIdx.x*72) + 9)]))
+            conv2d_nchw_1[4] = (conv2d_nchw_1[4] + (pad_temp.shared_1[31]*kernel.shared_1[((threadIdx.x*72) + 9)]))
+            conv2d_nchw_1[5] = (conv2d_nchw_1[5] + (pad_temp.shared_1[32]*kernel.shared_1[((threadIdx.x*72) + 9)]))
+            conv2d_nchw_1[6] = (conv2d_nchw_1[6] + (pad_temp.shared_1[33]*kernel.shared_1[((threadIdx.x*72) + 9)]))
+            conv2d_nchw_1[7] = (conv2d_nchw_1[7] + (pad_temp.shared_1[27]*kernel.shared_1[((threadIdx.x*72) + 4617)]))
+            conv2d_nchw_1[8] = (conv2d_nchw_1[8] + (pad_temp.shared_1[28]*kernel.shared_1[((threadIdx.x*72) + 4617)]))
+            conv2d_nchw_1[9] = (conv2d_nchw_1[9] + (pad_temp.shared_1[29]*kernel.shared_1[((threadIdx.x*72) + 4617)]))
+            conv2d_nchw_1[10] = (conv2d_nchw_1[10] + (pad_temp.shared_1[30]*kernel.shared_1[((threadIdx.x*72) + 4617)]))
+            conv2d_nchw_1[11] = (conv2d_nchw_1[11] + (pad_temp.shared_1[31]*kernel.shared_1[((threadIdx.x*72) + 4617)]))
+            conv2d_nchw_1[12] = (conv2d_nchw_1[12] + (pad_temp.shared_1[32]*kernel.shared_1[((threadIdx.x*72) + 4617)]))
+            conv2d_nchw_1[13] = (conv2d_nchw_1[13] + (pad_temp.shared_1[33]*kernel.shared_1[((threadIdx.x*72) + 4617)]))
+            conv2d_nchw_1[0] = (conv2d_nchw_1[0] + (pad_temp.shared_1[28]*kernel.shared_1[((threadIdx.x*72) + 10)]))
+            conv2d_nchw_1[1] = (conv2d_nchw_1[1] + (pad_temp.shared_1[29]*kernel.shared_1[((threadIdx.x*72) + 10)]))
+            conv2d_nchw_1[2] = (conv2d_nchw_1[2] + (pad_temp.shared_1[30]*kernel.shared_1[((threadIdx.x*72) + 10)]))
+            conv2d_nchw_1[3] = (conv2d_nchw_1[3] + (pad_temp.shared_1[31]*kernel.shared_1[((threadIdx.x*72) + 10)]))
+            conv2d_nchw_1[4] = (conv2d_nchw_1[4] + (pad_temp.shared_1[32]*kernel.shared_1[((threadIdx.x*72) + 10)]))
+            conv2d_nchw_1[5] = (conv2d_nchw_1[5] + (pad_temp.shared_1[33]*kernel.shared_1[((threadIdx.x*72) + 10)]))
+            conv2d_nchw_1[6] = (conv2d_nchw_1[6] + (pad_temp.shared_1[34]*kernel.shared_1[((threadIdx.x*72) + 10)]))
+            conv2d_nchw_1[7] = (conv2d_nchw_1[7] + (pad_temp.shared_1[28]*kernel.shared_1[((threadIdx.x*72) + 4618)]))
+            conv2d_nchw_1[8] = (conv2d_nchw_1[8] + (pad_temp.shared_1[29]*kernel.shared_1[((threadIdx.x*72) + 4618)]))
+            conv2d_nchw_1[9] = (conv2d_nchw_1[9] + (pad_temp.shared_1[30]*kernel.shared_1[((threadIdx.x*72) + 4618)]))
+            conv2d_nchw_1[10] = (conv2d_nchw_1[10] + (pad_temp.shared_1[31]*kernel.shared_1[((threadIdx.x*72) + 4618)]))
+            conv2d_nchw_1[11] = (conv2d_nchw_1[11] + (pad_temp.shared_1[32]*kernel.shared_1[((threadIdx.x*72) + 4618)]))
+            conv2d_nchw_1[12] = (conv2d_nchw_1[12] + (pad_temp.shared_1[33]*kernel.shared_1[((threadIdx.x*72) + 4618)]))
+            conv2d_nchw_1[13] = (conv2d_nchw_1[13] + (pad_temp.shared_1[34]*kernel.shared_1[((threadIdx.x*72) + 4618)]))
+            conv2d_nchw_1[0] = (conv2d_nchw_1[0] + (pad_temp.shared_1[29]*kernel.shared_1[((threadIdx.x*72) + 11)]))
+            conv2d_nchw_1[1] = (conv2d_nchw_1[1] + (pad_temp.shared_1[30]*kernel.shared_1[((threadIdx.x*72) + 11)]))
+            conv2d_nchw_1[2] = (conv2d_nchw_1[2] + (pad_temp.shared_1[31]*kernel.shared_1[((threadIdx.x*72) + 11)]))
+            conv2d_nchw_1[3] = (conv2d_nchw_1[3] + (pad_temp.shared_1[32]*kernel.shared_1[((threadIdx.x*72) + 11)]))
+            conv2d_nchw_1[4] = (conv2d_nchw_1[4] + (pad_temp.shared_1[33]*kernel.shared_1[((threadIdx.x*72) + 11)]))
+            conv2d_nchw_1[5] = (conv2d_nchw_1[5] + (pad_temp.shared_1[34]*kernel.shared_1[((threadIdx.x*72) + 11)]))
+            conv2d_nchw_1[6] = (conv2d_nchw_1[6] + (pad_temp.shared_1[35]*kernel.shared_1[((threadIdx.x*72) + 11)]))
+            conv2d_nchw_1[7] = (conv2d_nchw_1[7] + (pad_temp.shared_1[29]*kernel.shared_1[((threadIdx.x*72) + 4619)]))
+            conv2d_nchw_1[8] = (conv2d_nchw_1[8] + (pad_temp.shared_1[30]*kernel.shared_1[((threadIdx.x*72) + 4619)]))
+            conv2d_nchw_1[9] = (conv2d_nchw_1[9] + (pad_temp.shared_1[31]*kernel.shared_1[((threadIdx.x*72) + 4619)]))
+            conv2d_nchw_1[10] = (conv2d_nchw_1[10] + (pad_temp.shared_1[32]*kernel.shared_1[((threadIdx.x*72) + 4619)]))
+            conv2d_nchw_1[11] = (conv2d_nchw_1[11] + (pad_temp.shared_1[33]*kernel.shared_1[((threadIdx.x*72) + 4619)]))
+            conv2d_nchw_1[12] = (conv2d_nchw_1[12] + (pad_temp.shared_1[34]*kernel.shared_1[((threadIdx.x*72) + 4619)]))
+            conv2d_nchw_1[13] = (conv2d_nchw_1[13] + (pad_temp.shared_1[35]*kernel.shared_1[((threadIdx.x*72) + 4619)]))
+            conv2d_nchw_1[0] = (conv2d_nchw_1[0] + (pad_temp.shared_1[36]*kernel.shared_1[((threadIdx.x*72) + 12)]))
+            conv2d_nchw_1[1] = (conv2d_nchw_1[1] + (pad_temp.shared_1[37]*kernel.shared_1[((threadIdx.x*72) + 12)]))
+            conv2d_nchw_1[2] = (conv2d_nchw_1[2] + (pad_temp.shared_1[38]*kernel.shared_1[((threadIdx.x*72) + 12)]))
+            conv2d_nchw_1[3] = (conv2d_nchw_1[3] + (pad_temp.shared_1[39]*kernel.shared_1[((threadIdx.x*72) + 12)]))
+            conv2d_nchw_1[4] = (conv2d_nchw_1[4] + (pad_temp.shared_1[40]*kernel.shared_1[((threadIdx.x*72) + 12)]))
+            conv2d_nchw_1[5] = (conv2d_nchw_1[5] + (pad_temp.shared_1[41]*kernel.shared_1[((threadIdx.x*72) + 12)]))
+            conv2d_nchw_1[6] = (conv2d_nchw_1[6] + (pad_temp.shared_1[42]*kernel.shared_1[((threadIdx.x*72) + 12)]))
+            conv2d_nchw_1[7] = (conv2d_nchw_1[7] + (pad_temp.shared_1[36]*kernel.shared_1[((threadIdx.x*72) + 4620)]))
+            conv2d_nchw_1[8] = (conv2d_nchw_1[8] + (pad_temp.shared_1[37]*kernel.shared_1[((threadIdx.x*72) + 4620)]))
+            conv2d_nchw_1[9] = (conv2d_nchw_1[9] + (pad_temp.shared_1[38]*kernel.shared_1[((threadIdx.x*72) + 4620)]))
+            conv2d_nchw_1[10] = (conv2d_nchw_1[10] + (pad_temp.shared_1[39]*kernel.shared_1[((threadIdx.x*72) + 4620)]))
+            conv2d_nchw_1[11] = (conv2d_nchw_1[11] + (pad_temp.shared_1[40]*kernel.shared_1[((threadIdx.x*72) + 4620)]))
+            conv2d_nchw_1[12] = (conv2d_nchw_1[12] + (pad_temp.shared_1[41]*kernel.shared_1[((threadIdx.x*72) + 4620)]))
+            conv2d_nchw_1[13] = (conv2d_nchw_1[13] + (pad_temp.shared_1[42]*kernel.shared_1[((threadIdx.x*72) + 4620)]))
+            conv2d_nchw_1[0] = (conv2d_nchw_1[0] + (pad_temp.shared_1[37]*kernel.shared_1[((threadIdx.x*72) + 13)]))
+            conv2d_nchw_1[1] = (conv2d_nchw_1[1] + (pad_temp.shared_1[38]*kernel.shared_1[((threadIdx.x*72) + 13)]))
+            conv2d_nchw_1[2] = (conv2d_nchw_1[2] + (pad_temp.shared_1[39]*kernel.shared_1[((threadIdx.x*72) + 13)]))
+            conv2d_nchw_1[3] = (conv2d_nchw_1[3] + (pad_temp.shared_1[40]*kernel.shared_1[((threadIdx.x*72) + 13)]))
+            conv2d_nchw_1[4] = (conv2d_nchw_1[4] + (pad_temp.shared_1[41]*kernel.shared_1[((threadIdx.x*72) + 13)]))
+            conv2d_nchw_1[5] = (conv2d_nchw_1[5] + (pad_temp.shared_1[42]*kernel.shared_1[((threadIdx.x*72) + 13)]))
+            conv2d_nchw_1[6] = (conv2d_nchw_1[6] + (pad_temp.shared_1[43]*kernel.shared_1[((threadIdx.x*72) + 13)]))
+            conv2d_nchw_1[7] = (conv2d_nchw_1[7] + (pad_temp.shared_1[37]*kernel.shared_1[((threadIdx.x*72) + 4621)]))
+            conv2d_nchw_1[8] = (conv2d_nchw_1[8] + (pad_temp.shared_1[38]*kernel.shared_1[((threadIdx.x*72) + 4621)]))
+            conv2d_nchw_1[9] = (conv2d_nchw_1[9] + (pad_temp.shared_1[39]*kernel.shared_1[((threadIdx.x*72) + 4621)]))
+            conv2d_nchw_1[10] = (conv2d_nchw_1[10] + (pad_temp.shared_1[40]*kernel.shared_1[((threadIdx.x*72) + 4621)]))
+            conv2d_nchw_1[11] = (conv2d_nchw_1[11] + (pad_temp.shared_1[41]*kernel.shared_1[((threadIdx.x*72) + 4621)]))
+            conv2d_nchw_1[12] = (conv2d_nchw_1[12] + (pad_temp.shared_1[42]*kernel.shared_1[((threadIdx.x*72) + 4621)]))
+            conv2d_nchw_1[13] = (conv2d_nchw_1[13] + (pad_temp.shared_1[43]*kernel.shared_1[((threadIdx.x*72) + 4621)]))
+            conv2d_nchw_1[0] = (conv2d_nchw_1[0] + (pad_temp.shared_1[38]*kernel.shared_1[((threadIdx.x*72) + 14)]))
+            conv2d_nchw_1[1] = (conv2d_nchw_1[1] + (pad_temp.shared_1[39]*kernel.shared_1[((threadIdx.x*72) + 14)]))
+            conv2d_nchw_1[2] = (conv2d_nchw_1[2] + (pad_temp.shared_1[40]*kernel.shared_1[((threadIdx.x*72) + 14)]))
+            conv2d_nchw_1[3] = (conv2d_nchw_1[3] + (pad_temp.shared_1[41]*kernel.shared_1[((threadIdx.x*72) + 14)]))
+            conv2d_nchw_1[4] = (conv2d_nchw_1[4] + (pad_temp.shared_1[42]*kernel.shared_1[((threadIdx.x*72) + 14)]))
+            conv2d_nchw_1[5] = (conv2d_nchw_1[5] + (pad_temp.shared_1[43]*kernel.shared_1[((threadIdx.x*72) + 14)]))
+            conv2d_nchw_1[6] = (conv2d_nchw_1[6] + (pad_temp.shared_1[44]*kernel.shared_1[((threadIdx.x*72) + 14)]))
+            conv2d_nchw_1[7] = (conv2d_nchw_1[7] + (pad_temp.shared_1[38]*kernel.shared_1[((threadIdx.x*72) + 4622)]))
+            conv2d_nchw_1[8] = (conv2d_nchw_1[8] + (pad_temp.shared_1[39]*kernel.shared_1[((threadIdx.x*72) + 4622)]))
+            conv2d_nchw_1[9] = (conv2d_nchw_1[9] + (pad_temp.shared_1[40]*kernel.shared_1[((threadIdx.x*72) + 4622)]))
+            conv2d_nchw_1[10] = (conv2d_nchw_1[10] + (pad_temp.shared_1[41]*kernel.shared_1[((threadIdx.x*72) + 4622)]))
+            conv2d_nchw_1[11] = (conv2d_nchw_1[11] + (pad_temp.shared_1[42]*kernel.shared_1[((threadIdx.x*72) + 4622)]))
+            conv2d_nchw_1[12] = (conv2d_nchw_1[12] + (pad_temp.shared_1[43]*kernel.shared_1[((threadIdx.x*72) + 4622)]))
+            conv2d_nchw_1[13] = (conv2d_nchw_1[13] + (pad_temp.shared_1[44]*kernel.shared_1[((threadIdx.x*72) + 4622)]))
+            conv2d_nchw_1[0] = (conv2d_nchw_1[0] + (pad_temp.shared_1[45]*kernel.shared_1[((threadIdx.x*72) + 15)]))
+            conv2d_nchw_1[1] = (conv2d_nchw_1[1] + (pad_temp.shared_1[46]*kernel.shared_1[((threadIdx.x*72) + 15)]))
+            conv2d_nchw_1[2] = (conv2d_nchw_1[2] + (pad_temp.shared_1[47]*kernel.shared_1[((threadIdx.x*72) + 15)]))
+            conv2d_nchw_1[3] = (conv2d_nchw_1[3] + (pad_temp.shared_1[48]*kernel.shared_1[((threadIdx.x*72) + 15)]))
+            conv2d_nchw_1[4] = (conv2d_nchw_1[4] + (pad_temp.shared_1[49]*kernel.shared_1[((threadIdx.x*72) + 15)]))
+            conv2d_nchw_1[5] = (conv2d_nchw_1[5] + (pad_temp.shared_1[50]*kernel.shared_1[((threadIdx.x*72) + 15)]))
+            conv2d_nchw_1[6] = (conv2d_nchw_1[6] + (pad_temp.shared_1[51]*kernel.shared_1[((threadIdx.x*72) + 15)]))
+            conv2d_nchw_1[7] = (conv2d_nchw_1[7] + (pad_temp.shared_1[45]*kernel.shared_1[((threadIdx.x*72) + 4623)]))
+            conv2d_nchw_1[8] = (conv2d_nchw_1[8] + (pad_temp.shared_1[46]*kernel.shared_1[((threadIdx.x*72) + 4623)]))
+            conv2d_nchw_1[9] = (conv2d_nchw_1[9] + (pad_temp.shared_1[47]*kernel.shared_1[((threadIdx.x*72) + 4623)]))
+            conv2d_nchw_1[10] = (conv2d_nchw_1[10] + (pad_temp.shared_1[48]*kernel.shared_1[((threadIdx.x*72) + 4623)]))
+            conv2d_nchw_1[11] = (conv2d_nchw_1[11] + (pad_temp.shared_1[49]*kernel.shared_1[((threadIdx.x*72) + 4623)]))
+            conv2d_nchw_1[12] = (conv2d_nchw_1[12] + (pad_temp.shared_1[50]*kernel.shared_1[((threadIdx.x*72) + 4623)]))
+            conv2d_nchw_1[13] = (conv2d_nchw_1[13] + (pad_temp.shared_1[51]*kernel.shared_1[((threadIdx.x*72) + 4623)]))
+            conv2d_nchw_1[0] = (conv2d_nchw_1[0] + (pad_temp.shared_1[46]*kernel.shared_1[((threadIdx.x*72) + 16)]))
+            conv2d_nchw_1[1] = (conv2d_nchw_1[1] + (pad_temp.shared_1[47]*kernel.shared_1[((threadIdx.x*72) + 16)]))
+            conv2d_nchw_1[2] = (conv2d_nchw_1[2] + (pad_temp.shared_1[48]*kernel.shared_1[((threadIdx.x*72) + 16)]))
+            conv2d_nchw_1[3] = (conv2d_nchw_1[3] + (pad_temp.shared_1[49]*kernel.shared_1[((threadIdx.x*72) + 16)]))
+            conv2d_nchw_1[4] = (conv2d_nchw_1[4] + (pad_temp.shared_1[50]*kernel.shared_1[((threadIdx.x*72) + 16)]))
+            conv2d_nchw_1[5] = (conv2d_nchw_1[5] + (pad_temp.shared_1[51]*kernel.shared_1[((threadIdx.x*72) + 16)]))
+            conv2d_nchw_1[6] = (conv2d_nchw_1[6] + (pad_temp.shared_1[52]*kernel.shared_1[((threadIdx.x*72) + 16)]))
+            conv2d_nchw_1[7] = (conv2d_nchw_1[7] + (pad_temp.shared_1[46]*kernel.shared_1[((threadIdx.x*72) + 4624)]))
+            conv2d_nchw_1[8] = (conv2d_nchw_1[8] + (pad_temp.shared_1[47]*kernel.shared_1[((threadIdx.x*72) + 4624)]))
+            conv2d_nchw_1[9] = (conv2d_nchw_1[9] + (pad_temp.shared_1[48]*kernel.shared_1[((threadIdx.x*72) + 4624)]))
+            conv2d_nchw_1[10] = (conv2d_nchw_1[10] + (pad_temp.shared_1[49]*kernel.shared_1[((threadIdx.x*72) + 4624)]))
+            conv2d_nchw_1[11] = (conv2d_nchw_1[11] + (pad_temp.shared_1[50]*kernel.shared_1[((threadIdx.x*72) + 4624)]))
+            conv2d_nchw_1[12] = (conv2d_nchw_1[12] + (pad_temp.shared_1[51]*kernel.shared_1[((threadIdx.x*72) + 4624)]))
+            conv2d_nchw_1[13] = (conv2d_nchw_1[13] + (pad_temp.shared_1[52]*kernel.shared_1[((threadIdx.x*72) + 4624)]))
+            conv2d_nchw_1[0] = (conv2d_nchw_1[0] + (pad_temp.shared_1[47]*kernel.shared_1[((threadIdx.x*72) + 17)]))
+            conv2d_nchw_1[1] = (conv2d_nchw_1[1] + (pad_temp.shared_1[48]*kernel.shared_1[((threadIdx.x*72) + 17)]))
+            conv2d_nchw_1[2] = (conv2d_nchw_1[2] + (pad_temp.shared_1[49]*kernel.shared_1[((threadIdx.x*72) + 17)]))
+            conv2d_nchw_1[3] = (conv2d_nchw_1[3] + (pad_temp.shared_1[50]*kernel.shared_1[((threadIdx.x*72) + 17)]))
+            conv2d_nchw_1[4] = (conv2d_nchw_1[4] + (pad_temp.shared_1[51]*kernel.shared_1[((threadIdx.x*72) + 17)]))
+            conv2d_nchw_1[5] = (conv2d_nchw_1[5] + (pad_temp.shared_1[52]*kernel.shared_1[((threadIdx.x*72) + 17)]))
+            conv2d_nchw_1[6] = (conv2d_nchw_1[6] + (pad_temp.shared_1[53]*kernel.shared_1[((threadIdx.x*72) + 17)]))
+            conv2d_nchw_1[7] = (conv2d_nchw_1[7] + (pad_temp.shared_1[47]*kernel.shared_1[((threadIdx.x*72) + 4625)]))
+            conv2d_nchw_1[8] = (conv2d_nchw_1[8] + (pad_temp.shared_1[48]*kernel.shared_1[((threadIdx.x*72) + 4625)]))
+            conv2d_nchw_1[9] = (conv2d_nchw_1[9] + (pad_temp.shared_1[49]*kernel.shared_1[((threadIdx.x*72) + 4625)]))
+            conv2d_nchw_1[10] = (conv2d_nchw_1[10] + (pad_temp.shared_1[50]*kernel.shared_1[((threadIdx.x*72) + 4625)]))
+            conv2d_nchw_1[11] = (conv2d_nchw_1[11] + (pad_temp.shared_1[51]*kernel.shared_1[((threadIdx.x*72) + 4625)]))
+            conv2d_nchw_1[12] = (conv2d_nchw_1[12] + (pad_temp.shared_1[52]*kernel.shared_1[((threadIdx.x*72) + 4625)]))
+            conv2d_nchw_1[13] = (conv2d_nchw_1[13] + (pad_temp.shared_1[53]*kernel.shared_1[((threadIdx.x*72) + 4625)]))
+            conv2d_nchw_1[0] = (conv2d_nchw_1[0] + (pad_temp.shared_1[54]*kernel.shared_1[((threadIdx.x*72) + 18)]))
+            conv2d_nchw_1[1] = (conv2d_nchw_1[1] + (pad_temp.shared_1[55]*kernel.shared_1[((threadIdx.x*72) + 18)]))
+            conv2d_nchw_1[2] = (conv2d_nchw_1[2] + (pad_temp.shared_1[56]*kernel.shared_1[((threadIdx.x*72) + 18)]))
+            conv2d_nchw_1[3] = (conv2d_nchw_1[3] + (pad_temp.shared_1[57]*kernel.shared_1[((threadIdx.x*72) + 18)]))
+            conv2d_nchw_1[4] = (conv2d_nchw_1[4] + (pad_temp.shared_1[58]*kernel.shared_1[((threadIdx.x*72) + 18)]))
+            conv2d_nchw_1[5] = (conv2d_nchw_1[5] + (pad_temp.shared_1[59]*kernel.shared_1[((threadIdx.x*72) + 18)]))
+            conv2d_nchw_1[6] = (conv2d_nchw_1[6] + (pad_temp.shared_1[60]*kernel.shared_1[((threadIdx.x*72) + 18)]))
+            conv2d_nchw_1[7] = (conv2d_nchw_1[7] + (pad_temp.shared_1[54]*kernel.shared_1[((threadIdx.x*72) + 4626)]))
+            conv2d_nchw_1[8] = (conv2d_nchw_1[8] + (pad_temp.shared_1[55]*kernel.shared_1[((threadIdx.x*72) + 4626)]))
+            conv2d_nchw_1[9] = (conv2d_nchw_1[9] + (pad_temp.shared_1[56]*kernel.shared_1[((threadIdx.x*72) + 4626)]))
+            conv2d_nchw_1[10] = (conv2d_nchw_1[10] + (pad_temp.shared_1[57]*kernel.shared_1[((threadIdx.x*72) + 4626)]))
+            conv2d_nchw_1[11] = (conv2d_nchw_1[11] + (pad_temp.shared_1[58]*kernel.shared_1[((threadIdx.x*72) + 4626)]))
+            conv2d_nchw_1[12] = (conv2d_nchw_1[12] + (pad_temp.shared_1[59]*kernel.shared_1[((threadIdx.x*72) + 4626)]))
+            conv2d_nchw_1[13] = (conv2d_nchw_1[13] + (pad_temp.shared_1[60]*kernel.shared_1[((threadIdx.x*72) + 4626)]))
+            conv2d_nchw_1[0] = (conv2d_nchw_1[0] + (pad_temp.shared_1[55]*kernel.shared_1[((threadIdx.x*72) + 19)]))
+            conv2d_nchw_1[1] = (conv2d_nchw_1[1] + (pad_temp.shared_1[56]*kernel.shared_1[((threadIdx.x*72) + 19)]))
+            conv2d_nchw_1[2] = (conv2d_nchw_1[2] + (pad_temp.shared_1[57]*kernel.shared_1[((threadIdx.x*72) + 19)]))
+            conv2d_nchw_1[3] = (conv2d_nchw_1[3] + (pad_temp.shared_1[58]*kernel.shared_1[((threadIdx.x*72) + 19)]))
+            conv2d_nchw_1[4] = (conv2d_nchw_1[4] + (pad_temp.shared_1[59]*kernel.shared_1[((threadIdx.x*72) + 19)]))
+            conv2d_nchw_1[5] = (conv2d_nchw_1[5] + (pad_temp.shared_1[60]*kernel.shared_1[((threadIdx.x*72) + 19)]))
+            conv2d_nchw_1[6] = (conv2d_nchw_1[6] + (pad_temp.shared_1[61]*kernel.shared_1[((threadIdx.x*72) + 19)]))
+            conv2d_nchw_1[7] = (conv2d_nchw_1[7] + (pad_temp.shared_1[55]*kernel.shared_1[((threadIdx.x*72) + 4627)]))
+            conv2d_nchw_1[8] = (conv2d_nchw_1[8] + (pad_temp.shared_1[56]*kernel.shared_1[((threadIdx.x*72) + 4627)]))
+            conv2d_nchw_1[9] = (conv2d_nchw_1[9] + (pad_temp.shared_1[57]*kernel.shared_1[((threadIdx.x*72) + 4627)]))
+            conv2d_nchw_1[10] = (conv2d_nchw_1[10] + (pad_temp.shared_1[58]*kernel.shared_1[((threadIdx.x*72) + 4627)]))
+            conv2d_nchw_1[11] = (conv2d_nchw_1[11] + (pad_temp.shared_1[59]*kernel.shared_1[((threadIdx.x*72) + 4627)]))
+            conv2d_nchw_1[12] = (conv2d_nchw_1[12] + (pad_temp.shared_1[60]*kernel.shared_1[((threadIdx.x*72) + 4627)]))
+            conv2d_nchw_1[13] = (conv2d_nchw_1[13] + (pad_temp.shared_1[61]*kernel.shared_1[((threadIdx.x*72) + 4627)]))
+            conv2d_nchw_1[0] = (conv2d_nchw_1[0] + (pad_temp.shared_1[56]*kernel.shared_1[((threadIdx.x*72) + 20)]))
+            conv2d_nchw_1[1] = (conv2d_nchw_1[1] + (pad_temp.shared_1[57]*kernel.shared_1[((threadIdx.x*72) + 20)]))
+            conv2d_nchw_1[2] = (conv2d_nchw_1[2] + (pad_temp.shared_1[58]*kernel.shared_1[((threadIdx.x*72) + 20)]))
+            conv2d_nchw_1[3] = (conv2d_nchw_1[3] + (pad_temp.shared_1[59]*kernel.shared_1[((threadIdx.x*72) + 20)]))
+            conv2d_nchw_1[4] = (conv2d_nchw_1[4] + (pad_temp.shared_1[60]*kernel.shared_1[((threadIdx.x*72) + 20)]))
+            conv2d_nchw_1[5] = (conv2d_nchw_1[5] + (pad_temp.shared_1[61]*kernel.shared_1[((threadIdx.x*72) + 20)]))
+            conv2d_nchw_1[6] = (conv2d_nchw_1[6] + (pad_temp.shared_1[62]*kernel.shared_1[((threadIdx.x*72) + 20)]))
+            conv2d_nchw_1[7] = (conv2d_nchw_1[7] + (pad_temp.shared_1[56]*kernel.shared_1[((threadIdx.x*72) + 4628)]))
+            conv2d_nchw_1[8] = (conv2d_nchw_1[8] + (pad_temp.shared_1[57]*kernel.shared_1[((threadIdx.x*72) + 4628)]))
+            conv2d_nchw_1[9] = (conv2d_nchw_1[9] + (pad_temp.shared_1[58]*kernel.shared_1[((threadIdx.x*72) + 4628)]))
+            conv2d_nchw_1[10] = (conv2d_nchw_1[10] + (pad_temp.shared_1[59]*kernel.shared_1[((threadIdx.x*72) + 4628)]))
+            conv2d_nchw_1[11] = (conv2d_nchw_1[11] + (pad_temp.shared_1[60]*kernel.shared_1[((threadIdx.x*72) + 4628)]))
+            conv2d_nchw_1[12] = (conv2d_nchw_1[12] + (pad_temp.shared_1[61]*kernel.shared_1[((threadIdx.x*72) + 4628)]))
+            conv2d_nchw_1[13] = (conv2d_nchw_1[13] + (pad_temp.shared_1[62]*kernel.shared_1[((threadIdx.x*72) + 4628)]))
+            conv2d_nchw_1[0] = (conv2d_nchw_1[0] + (pad_temp.shared_1[63]*kernel.shared_1[((threadIdx.x*72) + 21)]))
+            conv2d_nchw_1[1] = (conv2d_nchw_1[1] + (pad_temp.shared_1[64]*kernel.shared_1[((threadIdx.x*72) + 21)]))
+            conv2d_nchw_1[2] = (conv2d_nchw_1[2] + (pad_temp.shared_1[65]*kernel.shared_1[((threadIdx.x*72) + 21)]))
+            conv2d_nchw_1[3] = (conv2d_nchw_1[3] + (pad_temp.shared_1[66]*kernel.shared_1[((threadIdx.x*72) + 21)]))
+            conv2d_nchw_1[4] = (conv2d_nchw_1[4] + (pad_temp.shared_1[67]*kernel.shared_1[((threadIdx.x*72) + 21)]))
+            conv2d_nchw_1[5] = (conv2d_nchw_1[5] + (pad_temp.shared_1[68]*kernel.shared_1[((threadIdx.x*72) + 21)]))
+            conv2d_nchw_1[6] = (conv2d_nchw_1[6] + (pad_temp.shared_1[69]*kernel.shared_1[((threadIdx.x*72) + 21)]))
+            conv2d_nchw_1[7] = (conv2d_nchw_1[7] + (pad_temp.shared_1[63]*kernel.shared_1[((threadIdx.x*72) + 4629)]))
+            conv2d_nchw_1[8] = (conv2d_nchw_1[8] + (pad_temp.shared_1[64]*kernel.shared_1[((threadIdx.x*72) + 4629)]))
+            conv2d_nchw_1[9] = (conv2d_nchw_1[9] + (pad_temp.shared_1[65]*kernel.shared_1[((threadIdx.x*72) + 4629)]))
+            conv2d_nchw_1[10] = (conv2d_nchw_1[10] + (pad_temp.shared_1[66]*kernel.shared_1[((threadIdx.x*72) + 4629)]))
+            conv2d_nchw_1[11] = (conv2d_nchw_1[11] + (pad_temp.shared_1[67]*kernel.shared_1[((threadIdx.x*72) + 4629)]))
+            conv2d_nchw_1[12] = (conv2d_nchw_1[12] + (pad_temp.shared_1[68]*kernel.shared_1[((threadIdx.x*72) + 4629)]))
+            conv2d_nchw_1[13] = (conv2d_nchw_1[13] + (pad_temp.shared_1[69]*kernel.shared_1[((threadIdx.x*72) + 4629)]))
+            conv2d_nchw_1[0] = (conv2d_nchw_1[0] + (pad_temp.shared_1[64]*kernel.shared_1[((threadIdx.x*72) + 22)]))
+            conv2d_nchw_1[1] = (conv2d_nchw_1[1] + (pad_temp.shared_1[65]*kernel.shared_1[((threadIdx.x*72) + 22)]))
+            conv2d_nchw_1[2] = (conv2d_nchw_1[2] + (pad_temp.shared_1[66]*kernel.shared_1[((threadIdx.x*72) + 22)]))
+            conv2d_nchw_1[3] = (conv2d_nchw_1[3] + (pad_temp.shared_1[67]*kernel.shared_1[((threadIdx.x*72) + 22)]))
+            conv2d_nchw_1[4] = (conv2d_nchw_1[4] + (pad_temp.shared_1[68]*kernel.shared_1[((threadIdx.x*72) + 22)]))
+            conv2d_nchw_1[5] = (conv2d_nchw_1[5] + (pad_temp.shared_1[69]*kernel.shared_1[((threadIdx.x*72) + 22)]))
+            conv2d_nchw_1[6] = (conv2d_nchw_1[6] + (pad_temp.shared_1[70]*kernel.shared_1[((threadIdx.x*72) + 22)]))
+            conv2d_nchw_1[7] = (conv2d_nchw_1[7] + (pad_temp.shared_1[64]*kernel.shared_1[((threadIdx.x*72) + 4630)]))
+            conv2d_nchw_1[8] = (conv2d_nchw_1[8] + (pad_temp.shared_1[65]*kernel.shared_1[((threadIdx.x*72) + 4630)]))
+            conv2d_nchw_1[9] = (conv2d_nchw_1[9] + (pad_temp.shared_1[66]*kernel.shared_1[((threadIdx.x*72) + 4630)]))
+            conv2d_nchw_1[10] = (conv2d_nchw_1[10] + (pad_temp.shared_1[67]*kernel.shared_1[((threadIdx.x*72) + 4630)]))
+            conv2d_nchw_1[11] = (conv2d_nchw_1[11] + (pad_temp.shared_1[68]*kernel.shared_1[((threadIdx.x*72) + 4630)]))
+            conv2d_nchw_1[12] = (conv2d_nchw_1[12] + (pad_temp.shared_1[69]*kernel.shared_1[((threadIdx.x*72) + 4630)]))
+            conv2d_nchw_1[13] = (conv2d_nchw_1[13] + (pad_temp.shared_1[70]*kernel.shared_1[((threadIdx.x*72) + 4630)]))
+            conv2d_nchw_1[0] = (conv2d_nchw_1[0] + (pad_temp.shared_1[65]*kernel.shared_1[((threadIdx.x*72) + 23)]))
+            conv2d_nchw_1[1] = (conv2d_nchw_1[1] + (pad_temp.shared_1[66]*kernel.shared_1[((threadIdx.x*72) + 23)]))
+            conv2d_nchw_1[2] = (conv2d_nchw_1[2] + (pad_temp.shared_1[67]*kernel.shared_1[((threadIdx.x*72) + 23)]))
+            conv2d_nchw_1[3] = (conv2d_nchw_1[3] + (pad_temp.shared_1[68]*kernel.shared_1[((threadIdx.x*72) + 23)]))
+            conv2d_nchw_1[4] = (conv2d_nchw_1[4] + (pad_temp.shared_1[69]*kernel.shared_1[((threadIdx.x*72) + 23)]))
+            conv2d_nchw_1[5] = (conv2d_nchw_1[5] + (pad_temp.shared_1[70]*kernel.shared_1[((threadIdx.x*72) + 23)]))
+            conv2d_nchw_1[6] = (conv2d_nchw_1[6] + (pad_temp.shared_1[71]*kernel.shared_1[((threadIdx.x*72) + 23)]))
+            conv2d_nchw_1[7] = (conv2d_nchw_1[7] + (pad_temp.shared_1[65]*kernel.shared_1[((threadIdx.x*72) + 4631)]))
+            conv2d_nchw_1[8] = (conv2d_nchw_1[8] + (pad_temp.shared_1[66]*kernel.shared_1[((threadIdx.x*72) + 4631)]))
+            conv2d_nchw_1[9] = (conv2d_nchw_1[9] + (pad_temp.shared_1[67]*kernel.shared_1[((threadIdx.x*72) + 4631)]))
+            conv2d_nchw_1[10] = (conv2d_nchw_1[10] + (pad_temp.shared_1[68]*kernel.shared_1[((threadIdx.x*72) + 4631)]))
+            conv2d_nchw_1[11] = (conv2d_nchw_1[11] + (pad_temp.shared_1[69]*kernel.shared_1[((threadIdx.x*72) + 4631)]))
+            conv2d_nchw_1[12] = (conv2d_nchw_1[12] + (pad_temp.shared_1[70]*kernel.shared_1[((threadIdx.x*72) + 4631)]))
+            conv2d_nchw_1[13] = (conv2d_nchw_1[13] + (pad_temp.shared_1[71]*kernel.shared_1[((threadIdx.x*72) + 4631)]))
+            conv2d_nchw_1[0] = (conv2d_nchw_1[0] + (pad_temp.shared_1[72]*kernel.shared_1[((threadIdx.x*72) + 24)]))
+            conv2d_nchw_1[1] = (conv2d_nchw_1[1] + (pad_temp.shared_1[73]*kernel.shared_1[((threadIdx.x*72) + 24)]))
+            conv2d_nchw_1[2] = (conv2d_nchw_1[2] + (pad_temp.shared_1[74]*kernel.shared_1[((threadIdx.x*72) + 24)]))
+            conv2d_nchw_1[3] = (conv2d_nchw_1[3] + (pad_temp.shared_1[75]*kernel.shared_1[((threadIdx.x*72) + 24)]))
+            conv2d_nchw_1[4] = (conv2d_nchw_1[4] + (pad_temp.shared_1[76]*kernel.shared_1[((threadIdx.x*72) + 24)]))
+            conv2d_nchw_1[5] = (conv2d_nchw_1[5] + (pad_temp.shared_1[77]*kernel.shared_1[((threadIdx.x*72) + 24)]))
+            conv2d_nchw_1[6] = (conv2d_nchw_1[6] + (pad_temp.shared_1[78]*kernel.shared_1[((threadIdx.x*72) + 24)]))
+            conv2d_nchw_1[7] = (conv2d_nchw_1[7] + (pad_temp.shared_1[72]*kernel.shared_1[((threadIdx.x*72) + 4632)]))
+            conv2d_nchw_1[8] = (conv2d_nchw_1[8] + (pad_temp.shared_1[73]*kernel.shared_1[((threadIdx.x*72) + 4632)]))
+            conv2d_nchw_1[9] = (conv2d_nchw_1[9] + (pad_temp.shared_1[74]*kernel.shared_1[((threadIdx.x*72) + 4632)]))
+            conv2d_nchw_1[10] = (conv2d_nchw_1[10] + (pad_temp.shared_1[75]*kernel.shared_1[((threadIdx.x*72) + 4632)]))
+            conv2d_nchw_1[11] = (conv2d_nchw_1[11] + (pad_temp.shared_1[76]*kernel.shared_1[((threadIdx.x*72) + 4632)]))
+            conv2d_nchw_1[12] = (conv2d_nchw_1[12] + (pad_temp.shared_1[77]*kernel.shared_1[((threadIdx.x*72) + 4632)]))
+            conv2d_nchw_1[13] = (conv2d_nchw_1[13] + (pad_temp.shared_1[78]*kernel.shared_1[((threadIdx.x*72) + 4632)]))
+            conv2d_nchw_1[0] = (conv2d_nchw_1[0] + (pad_temp.shared_1[73]*kernel.shared_1[((threadIdx.x*72) + 25)]))
+            conv2d_nchw_1[1] = (conv2d_nchw_1[1] + (pad_temp.shared_1[74]*kernel.shared_1[((threadIdx.x*72) + 25)]))
+            conv2d_nchw_1[2] = (conv2d_nchw_1[2] + (pad_temp.shared_1[75]*kernel.shared_1[((threadIdx.x*72) + 25)]))
+            conv2d_nchw_1[3] = (conv2d_nchw_1[3] + (pad_temp.shared_1[76]*kernel.shared_1[((threadIdx.x*72) + 25)]))
+            conv2d_nchw_1[4] = (conv2d_nchw_1[4] + (pad_temp.shared_1[77]*kernel.shared_1[((threadIdx.x*72) + 25)]))
+            conv2d_nchw_1[5] = (conv2d_nchw_1[5] + (pad_temp.shared_1[78]*kernel.shared_1[((threadIdx.x*72) + 25)]))
+            conv2d_nchw_1[6] = (conv2d_nchw_1[6] + (pad_temp.shared_1[79]*kernel.shared_1[((threadIdx.x*72) + 25)]))
+            conv2d_nchw_1[7] = (conv2d_nchw_1[7] + (pad_temp.shared_1[73]*kernel.shared_1[((threadIdx.x*72) + 4633)]))
+            conv2d_nchw_1[8] = (conv2d_nchw_1[8] + (pad_temp.shared_1[74]*kernel.shared_1[((threadIdx.x*72) + 4633)]))
+            conv2d_nchw_1[9] = (conv2d_nchw_1[9] + (pad_temp.shared_1[75]*kernel.shared_1[((threadIdx.x*72) + 4633)]))
+            conv2d_nchw_1[10] = (conv2d_nchw_1[10] + (pad_temp.shared_1[76]*kernel.shared_1[((threadIdx.x*72) + 4633)]))
+            conv2d_nchw_1[11] = (conv2d_nchw_1[11] + (pad_temp.shared_1[77]*kernel.shared_1[((threadIdx.x*72) + 4633)]))
+            conv2d_nchw_1[12] = (conv2d_nchw_1[12] + (pad_temp.shared_1[78]*kernel.shared_1[((threadIdx.x*72) + 4633)]))
+            conv2d_nchw_1[13] = (conv2d_nchw_1[13] + (pad_temp.shared_1[79]*kernel.shared_1[((threadIdx.x*72) + 4633)]))
+            conv2d_nchw_1[0] = (conv2d_nchw_1[0] + (pad_temp.shared_1[74]*kernel.shared_1[((threadIdx.x*72) + 26)]))
+            conv2d_nchw_1[1] = (conv2d_nchw_1[1] + (pad_temp.shared_1[75]*kernel.shared_1[((threadIdx.x*72) + 26)]))
+            conv2d_nchw_1[2] = (conv2d_nchw_1[2] + (pad_temp.shared_1[76]*kernel.shared_1[((threadIdx.x*72) + 26)]))
+            conv2d_nchw_1[3] = (conv2d_nchw_1[3] + (pad_temp.shared_1[77]*kernel.shared_1[((threadIdx.x*72) + 26)]))
+            conv2d_nchw_1[4] = (conv2d_nchw_1[4] + (pad_temp.shared_1[78]*kernel.shared_1[((threadIdx.x*72) + 26)]))
+            conv2d_nchw_1[5] = (conv2d_nchw_1[5] + (pad_temp.shared_1[79]*kernel.shared_1[((threadIdx.x*72) + 26)]))
+            conv2d_nchw_1[6] = (conv2d_nchw_1[6] + (pad_temp.shared_1[80]*kernel.shared_1[((threadIdx.x*72) + 26)]))
+            conv2d_nchw_1[7] = (conv2d_nchw_1[7] + (pad_temp.shared_1[74]*kernel.shared_1[((threadIdx.x*72) + 4634)]))
+            conv2d_nchw_1[8] = (conv2d_nchw_1[8] + (pad_temp.shared_1[75]*kernel.shared_1[((threadIdx.x*72) + 4634)]))
+            conv2d_nchw_1[9] = (conv2d_nchw_1[9] + (pad_temp.shared_1[76]*kernel.shared_1[((threadIdx.x*72) + 4634)]))
+            conv2d_nchw_1[10] = (conv2d_nchw_1[10] + (pad_temp.shared_1[77]*kernel.shared_1[((threadIdx.x*72) + 4634)]))
+            conv2d_nchw_1[11] = (conv2d_nchw_1[11] + (pad_temp.shared_1[78]*kernel.shared_1[((threadIdx.x*72) + 4634)]))
+            conv2d_nchw_1[12] = (conv2d_nchw_1[12] + (pad_temp.shared_1[79]*kernel.shared_1[((threadIdx.x*72) + 4634)]))
+            conv2d_nchw_1[13] = (conv2d_nchw_1[13] + (pad_temp.shared_1[80]*kernel.shared_1[((threadIdx.x*72) + 4634)]))
+            conv2d_nchw_1[0] = (conv2d_nchw_1[0] + (pad_temp.shared_1[81]*kernel.shared_1[((threadIdx.x*72) + 27)]))
+            conv2d_nchw_1[1] = (conv2d_nchw_1[1] + (pad_temp.shared_1[82]*kernel.shared_1[((threadIdx.x*72) + 27)]))
+            conv2d_nchw_1[2] = (conv2d_nchw_1[2] + (pad_temp.shared_1[83]*kernel.shared_1[((threadIdx.x*72) + 27)]))
+            conv2d_nchw_1[3] = (conv2d_nchw_1[3] + (pad_temp.shared_1[84]*kernel.shared_1[((threadIdx.x*72) + 27)]))
+            conv2d_nchw_1[4] = (conv2d_nchw_1[4] + (pad_temp.shared_1[85]*kernel.shared_1[((threadIdx.x*72) + 27)]))
+            conv2d_nchw_1[5] = (conv2d_nchw_1[5] + (pad_temp.shared_1[86]*kernel.shared_1[((threadIdx.x*72) + 27)]))
+            conv2d_nchw_1[6] = (conv2d_nchw_1[6] + (pad_temp.shared_1[87]*kernel.shared_1[((threadIdx.x*72) + 27)]))
+            conv2d_nchw_1[7] = (conv2d_nchw_1[7] + (pad_temp.shared_1[81]*kernel.shared_1[((threadIdx.x*72) + 4635)]))
+            conv2d_nchw_1[8] = (conv2d_nchw_1[8] + (pad_temp.shared_1[82]*kernel.shared_1[((threadIdx.x*72) + 4635)]))
+            conv2d_nchw_1[9] = (conv2d_nchw_1[9] + (pad_temp.shared_1[83]*kernel.shared_1[((threadIdx.x*72) + 4635)]))
+            conv2d_nchw_1[10] = (conv2d_nchw_1[10] + (pad_temp.shared_1[84]*kernel.shared_1[((threadIdx.x*72) + 4635)]))
+            conv2d_nchw_1[11] = (conv2d_nchw_1[11] + (pad_temp.shared_1[85]*kernel.shared_1[((threadIdx.x*72) + 4635)]))
+            conv2d_nchw_1[12] = (conv2d_nchw_1[12] + (pad_temp.shared_1[86]*kernel.shared_1[((threadIdx.x*72) + 4635)]))
+            conv2d_nchw_1[13] = (conv2d_nchw_1[13] + (pad_temp.shared_1[87]*kernel.shared_1[((threadIdx.x*72) + 4635)]))
+            conv2d_nchw_1[0] = (conv2d_nchw_1[0] + (pad_temp.shared_1[82]*kernel.shared_1[((threadIdx.x*72) + 28)]))
+            conv2d_nchw_1[1] = (conv2d_nchw_1[1] + (pad_temp.shared_1[83]*kernel.shared_1[((threadIdx.x*72) + 28)]))
+            conv2d_nchw_1[2] = (conv2d_nchw_1[2] + (pad_temp.shared_1[84]*kernel.shared_1[((threadIdx.x*72) + 28)]))
+            conv2d_nchw_1[3] = (conv2d_nchw_1[3] + (pad_temp.shared_1[85]*kernel.shared_1[((threadIdx.x*72) + 28)]))
+            conv2d_nchw_1[4] = (conv2d_nchw_1[4] + (pad_temp.shared_1[86]*kernel.shared_1[((threadIdx.x*72) + 28)]))
+            conv2d_nchw_1[5] = (conv2d_nchw_1[5] + (pad_temp.shared_1[87]*kernel.shared_1[((threadIdx.x*72) + 28)]))
+            conv2d_nchw_1[6] = (conv2d_nchw_1[6] + (pad_temp.shared_1[88]*kernel.shared_1[((threadIdx.x*72) + 28)]))
+            conv2d_nchw_1[7] = (conv2d_nchw_1[7] + (pad_temp.shared_1[82]*kernel.shared_1[((threadIdx.x*72) + 4636)]))
+            conv2d_nchw_1[8] = (conv2d_nchw_1[8] + (pad_temp.shared_1[83]*kernel.shared_1[((threadIdx.x*72) + 4636)]))
+            conv2d_nchw_1[9] = (conv2d_nchw_1[9] + (pad_temp.shared_1[84]*kernel.shared_1[((threadIdx.x*72) + 4636)]))
+            conv2d_nchw_1[10] = (conv2d_nchw_1[10] + (pad_temp.shared_1[85]*kernel.shared_1[((threadIdx.x*72) + 4636)]))
+            conv2d_nchw_1[11] = (conv2d_nchw_1[11] + (pad_temp.shared_1[86]*kernel.shared_1[((threadIdx.x*72) + 4636)]))
+            conv2d_nchw_1[12] = (conv2d_nchw_1[12] + (pad_temp.shared_1[87]*kernel.shared_1[((threadIdx.x*72) + 4636)]))
+            conv2d_nchw_1[13] = (conv2d_nchw_1[13] + (pad_temp.shared_1[88]*kernel.shared_1[((threadIdx.x*72) + 4636)]))
+            conv2d_nchw_1[0] = (conv2d_nchw_1[0] + (pad_temp.shared_1[83]*kernel.shared_1[((threadIdx.x*72) + 29)]))
+            conv2d_nchw_1[1] = (conv2d_nchw_1[1] + (pad_temp.shared_1[84]*kernel.shared_1[((threadIdx.x*72) + 29)]))
+            conv2d_nchw_1[2] = (conv2d_nchw_1[2] + (pad_temp.shared_1[85]*kernel.shared_1[((threadIdx.x*72) + 29)]))
+            conv2d_nchw_1[3] = (conv2d_nchw_1[3] + (pad_temp.shared_1[86]*kernel.shared_1[((threadIdx.x*72) + 29)]))
+            conv2d_nchw_1[4] = (conv2d_nchw_1[4] + (pad_temp.shared_1[87]*kernel.shared_1[((threadIdx.x*72) + 29)]))
+            conv2d_nchw_1[5] = (conv2d_nchw_1[5] + (pad_temp.shared_1[88]*kernel.shared_1[((threadIdx.x*72) + 29)]))
+            conv2d_nchw_1[6] = (conv2d_nchw_1[6] + (pad_temp.shared_1[89]*kernel.shared_1[((threadIdx.x*72) + 29)]))
+            conv2d_nchw_1[7] = (conv2d_nchw_1[7] + (pad_temp.shared_1[83]*kernel.shared_1[((threadIdx.x*72) + 4637)]))
+            conv2d_nchw_1[8] = (conv2d_nchw_1[8] + (pad_temp.shared_1[84]*kernel.shared_1[((threadIdx.x*72) + 4637)]))
+            conv2d_nchw_1[9] = (conv2d_nchw_1[9] + (pad_temp.shared_1[85]*kernel.shared_1[((threadIdx.x*72) + 4637)]))
+            conv2d_nchw_1[10] = (conv2d_nchw_1[10] + (pad_temp.shared_1[86]*kernel.shared_1[((threadIdx.x*72) + 4637)]))
+            conv2d_nchw_1[11] = (conv2d_nchw_1[11] + (pad_temp.shared_1[87]*kernel.shared_1[((threadIdx.x*72) + 4637)]))
+            conv2d_nchw_1[12] = (conv2d_nchw_1[12] + (pad_temp.shared_1[88]*kernel.shared_1[((threadIdx.x*72) + 4637)]))
+            conv2d_nchw_1[13] = (conv2d_nchw_1[13] + (pad_temp.shared_1[89]*kernel.shared_1[((threadIdx.x*72) + 4637)]))
+            conv2d_nchw_1[0] = (conv2d_nchw_1[0] + (pad_temp.shared_1[90]*kernel.shared_1[((threadIdx.x*72) + 30)]))
+            conv2d_nchw_1[1] = (conv2d_nchw_1[1] + (pad_temp.shared_1[91]*kernel.shared_1[((threadIdx.x*72) + 30)]))
+            conv2d_nchw_1[2] = (conv2d_nchw_1[2] + (pad_temp.shared_1[92]*kernel.shared_1[((threadIdx.x*72) + 30)]))
+            conv2d_nchw_1[3] = (conv2d_nchw_1[3] + (pad_temp.shared_1[93]*kernel.shared_1[((threadIdx.x*72) + 30)]))
+            conv2d_nchw_1[4] = (conv2d_nchw_1[4] + (pad_temp.shared_1[94]*kernel.shared_1[((threadIdx.x*72) + 30)]))
+            conv2d_nchw_1[5] = (conv2d_nchw_1[5] + (pad_temp.shared_1[95]*kernel.shared_1[((threadIdx.x*72) + 30)]))
+            conv2d_nchw_1[6] = (conv2d_nchw_1[6] + (pad_temp.shared_1[96]*kernel.shared_1[((threadIdx.x*72) + 30)]))
+            conv2d_nchw_1[7] = (conv2d_nchw_1[7] + (pad_temp.shared_1[90]*kernel.shared_1[((threadIdx.x*72) + 4638)]))
+            conv2d_nchw_1[8] = (conv2d_nchw_1[8] + (pad_temp.shared_1[91]*kernel.shared_1[((threadIdx.x*72) + 4638)]))
+            conv2d_nchw_1[9] = (conv2d_nchw_1[9] + (pad_temp.shared_1[92]*kernel.shared_1[((threadIdx.x*72) + 4638)]))
+            conv2d_nchw_1[10] = (conv2d_nchw_1[10] + (pad_temp.shared_1[93]*kernel.shared_1[((threadIdx.x*72) + 4638)]))
+            conv2d_nchw_1[11] = (conv2d_nchw_1[11] + (pad_temp.shared_1[94]*kernel.shared_1[((threadIdx.x*72) + 4638)]))
+            conv2d_nchw_1[12] = (conv2d_nchw_1[12] + (pad_temp.shared_1[95]*kernel.shared_1[((threadIdx.x*72) + 4638)]))
+            conv2d_nchw_1[13] = (conv2d_nchw_1[13] + (pad_temp.shared_1[96]*kernel.shared_1[((threadIdx.x*72) + 4638)]))
+            conv2d_nchw_1[0] = (conv2d_nchw_1[0] + (pad_temp.shared_1[91]*kernel.shared_1[((threadIdx.x*72) + 31)]))
+            conv2d_nchw_1[1] = (conv2d_nchw_1[1] + (pad_temp.shared_1[92]*kernel.shared_1[((threadIdx.x*72) + 31)]))
+            conv2d_nchw_1[2] = (conv2d_nchw_1[2] + (pad_temp.shared_1[93]*kernel.shared_1[((threadIdx.x*72) + 31)]))
+            conv2d_nchw_1[3] = (conv2d_nchw_1[3] + (pad_temp.shared_1[94]*kernel.shared_1[((threadIdx.x*72) + 31)]))
+            conv2d_nchw_1[4] = (conv2d_nchw_1[4] + (pad_temp.shared_1[95]*kernel.shared_1[((threadIdx.x*72) + 31)]))
+            conv2d_nchw_1[5] = (conv2d_nchw_1[5] + (pad_temp.shared_1[96]*kernel.shared_1[((threadIdx.x*72) + 31)]))
+            conv2d_nchw_1[6] = (conv2d_nchw_1[6] + (pad_temp.shared_1[97]*kernel.shared_1[((threadIdx.x*72) + 31)]))
+            conv2d_nchw_1[7] = (conv2d_nchw_1[7] + (pad_temp.shared_1[91]*kernel.shared_1[((threadIdx.x*72) + 4639)]))
+            conv2d_nchw_1[8] = (conv2d_nchw_1[8] + (pad_temp.shared_1[92]*kernel.shared_1[((threadIdx.x*72) + 4639)]))
+            conv2d_nchw_1[9] = (conv2d_nchw_1[9] + (pad_temp.shared_1[93]*kernel.shared_1[((threadIdx.x*72) + 4639)]))
+            conv2d_nchw_1[10] = (conv2d_nchw_1[10] + (pad_temp.shared_1[94]*kernel.shared_1[((threadIdx.x*72) + 4639)]))
+            conv2d_nchw_1[11] = (conv2d_nchw_1[11] + (pad_temp.shared_1[95]*kernel.shared_1[((threadIdx.x*72) + 4639)]))
+            conv2d_nchw_1[12] = (conv2d_nchw_1[12] + (pad_temp.shared_1[96]*kernel.shared_1[((threadIdx.x*72) + 4639)]))
+            conv2d_nchw_1[13] = (conv2d_nchw_1[13] + (pad_temp.shared_1[97]*kernel.shared_1[((threadIdx.x*72) + 4639)]))
+            conv2d_nchw_1[0] = (conv2d_nchw_1[0] + (pad_temp.shared_1[92]*kernel.shared_1[((threadIdx.x*72) + 32)]))
+            conv2d_nchw_1[1] = (conv2d_nchw_1[1] + (pad_temp.shared_1[93]*kernel.shared_1[((threadIdx.x*72) + 32)]))
+            conv2d_nchw_1[2] = (conv2d_nchw_1[2] + (pad_temp.shared_1[94]*kernel.shared_1[((threadIdx.x*72) + 32)]))
+            conv2d_nchw_1[3] = (conv2d_nchw_1[3] + (pad_temp.shared_1[95]*kernel.shared_1[((threadIdx.x*72) + 32)]))
+            conv2d_nchw_1[4] = (conv2d_nchw_1[4] + (pad_temp.shared_1[96]*kernel.shared_1[((threadIdx.x*72) + 32)]))
+            conv2d_nchw_1[5] = (conv2d_nchw_1[5] + (pad_temp.shared_1[97]*kernel.shared_1[((threadIdx.x*72) + 32)]))
+            conv2d_nchw_1[6] = (conv2d_nchw_1[6] + (pad_temp.shared_1[98]*kernel.shared_1[((threadIdx.x*72) + 32)]))
+            conv2d_nchw_1[7] = (conv2d_nchw_1[7] + (pad_temp.shared_1[92]*kernel.shared_1[((threadIdx.x*72) + 4640)]))
+            conv2d_nchw_1[8] = (conv2d_nchw_1[8] + (pad_temp.shared_1[93]*kernel.shared_1[((threadIdx.x*72) + 4640)]))
+            conv2d_nchw_1[9] = (conv2d_nchw_1[9] + (pad_temp.shared_1[94]*kernel.shared_1[((threadIdx.x*72) + 4640)]))
+            conv2d_nchw_1[10] = (conv2d_nchw_1[10] + (pad_temp.shared_1[95]*kernel.shared_1[((threadIdx.x*72) + 4640)]))
+            conv2d_nchw_1[11] = (conv2d_nchw_1[11] + (pad_temp.shared_1[96]*kernel.shared_1[((threadIdx.x*72) + 4640)]))
+            conv2d_nchw_1[12] = (conv2d_nchw_1[12] + (pad_temp.shared_1[97]*kernel.shared_1[((threadIdx.x*72) + 4640)]))
+            conv2d_nchw_1[13] = (conv2d_nchw_1[13] + (pad_temp.shared_1[98]*kernel.shared_1[((threadIdx.x*72) + 4640)]))
+            conv2d_nchw_1[0] = (conv2d_nchw_1[0] + (pad_temp.shared_1[99]*kernel.shared_1[((threadIdx.x*72) + 33)]))
+            conv2d_nchw_1[1] = (conv2d_nchw_1[1] + (pad_temp.shared_1[100]*kernel.shared_1[((threadIdx.x*72) + 33)]))
+            conv2d_nchw_1[2] = (conv2d_nchw_1[2] + (pad_temp.shared_1[101]*kernel.shared_1[((threadIdx.x*72) + 33)]))
+            conv2d_nchw_1[3] = (conv2d_nchw_1[3] + (pad_temp.shared_1[102]*kernel.shared_1[((threadIdx.x*72) + 33)]))
+            conv2d_nchw_1[4] = (conv2d_nchw_1[4] + (pad_temp.shared_1[103]*kernel.shared_1[((threadIdx.x*72) + 33)]))
+            conv2d_nchw_1[5] = (conv2d_nchw_1[5] + (pad_temp.shared_1[104]*kernel.shared_1[((threadIdx.x*72) + 33)]))
+            conv2d_nchw_1[6] = (conv2d_nchw_1[6] + (pad_temp.shared_1[105]*kernel.shared_1[((threadIdx.x*72) + 33)]))
+            conv2d_nchw_1[7] = (conv2d_nchw_1[7] + (pad_temp.shared_1[99]*kernel.shared_1[((threadIdx.x*72) + 4641)]))
+            conv2d_nchw_1[8] = (conv2d_nchw_1[8] + (pad_temp.shared_1[100]*kernel.shared_1[((threadIdx.x*72) + 4641)]))
+            conv2d_nchw_1[9] = (conv2d_nchw_1[9] + (pad_temp.shared_1[101]*kernel.shared_1[((threadIdx.x*72) + 4641)]))
+            conv2d_nchw_1[10] = (conv2d_nchw_1[10] + (pad_temp.shared_1[102]*kernel.shared_1[((threadIdx.x*72) + 4641)]))
+            conv2d_nchw_1[11] = (conv2d_nchw_1[11] + (pad_temp.shared_1[103]*kernel.shared_1[((threadIdx.x*72) + 4641)]))
+            conv2d_nchw_1[12] = (conv2d_nchw_1[12] + (pad_temp.shared_1[104]*kernel.shared_1[((threadIdx.x*72) + 4641)]))
+            conv2d_nchw_1[13] = (conv2d_nchw_1[13] + (pad_temp.shared_1[105]*kernel.shared_1[((threadIdx.x*72) + 4641)]))
+            conv2d_nchw_1[0] = (conv2d_nchw_1[0] + (pad_temp.shared_1[100]*kernel.shared_1[((threadIdx.x*72) + 34)]))
+            conv2d_nchw_1[1] = (conv2d_nchw_1[1] + (pad_temp.shared_1[101]*kernel.shared_1[((threadIdx.x*72) + 34)]))
+            conv2d_nchw_1[2] = (conv2d_nchw_1[2] + (pad_temp.shared_1[102]*kernel.shared_1[((threadIdx.x*72) + 34)]))
+            conv2d_nchw_1[3] = (conv2d_nchw_1[3] + (pad_temp.shared_1[103]*kernel.shared_1[((threadIdx.x*72) + 34)]))
+            conv2d_nchw_1[4] = (conv2d_nchw_1[4] + (pad_temp.shared_1[104]*kernel.shared_1[((threadIdx.x*72) + 34)]))
+            conv2d_nchw_1[5] = (conv2d_nchw_1[5] + (pad_temp.shared_1[105]*kernel.shared_1[((threadIdx.x*72) + 34)]))
+            conv2d_nchw_1[6] = (conv2d_nchw_1[6] + (pad_temp.shared_1[106]*kernel.shared_1[((threadIdx.x*72) + 34)]))
+            conv2d_nchw_1[7] = (conv2d_nchw_1[7] + (pad_temp.shared_1[100]*kernel.shared_1[((threadIdx.x*72) + 4642)]))
+            conv2d_nchw_1[8] = (conv2d_nchw_1[8] + (pad_temp.shared_1[101]*kernel.shared_1[((threadIdx.x*72) + 4642)]))
+            conv2d_nchw_1[9] = (conv2d_nchw_1[9] + (pad_temp.shared_1[102]*kernel.shared_1[((threadIdx.x*72) + 4642)]))
+            conv2d_nchw_1[10] = (conv2d_nchw_1[10] + (pad_temp.shared_1[103]*kernel.shared_1[((threadIdx.x*72) + 4642)]))
+            conv2d_nchw_1[11] = (conv2d_nchw_1[11] + (pad_temp.shared_1[104]*kernel.shared_1[((threadIdx.x*72) + 4642)]))
+            conv2d_nchw_1[12] = (conv2d_nchw_1[12] + (pad_temp.shared_1[105]*kernel.shared_1[((threadIdx.x*72) + 4642)]))
+            conv2d_nchw_1[13] = (conv2d_nchw_1[13] + (pad_temp.shared_1[106]*kernel.shared_1[((threadIdx.x*72) + 4642)]))
+            conv2d_nchw_1[0] = (conv2d_nchw_1[0] + (pad_temp.shared_1[101]*kernel.shared_1[((threadIdx.x*72) + 35)]))
+            conv2d_nchw_1[1] = (conv2d_nchw_1[1] + (pad_temp.shared_1[102]*kernel.shared_1[((threadIdx.x*72) + 35)]))
+            conv2d_nchw_1[2] = (conv2d_nchw_1[2] + (pad_temp.shared_1[103]*kernel.shared_1[((threadIdx.x*72) + 35)]))
+            conv2d_nchw_1[3] = (conv2d_nchw_1[3] + (pad_temp.shared_1[104]*kernel.shared_1[((threadIdx.x*72) + 35)]))
+            conv2d_nchw_1[4] = (conv2d_nchw_1[4] + (pad_temp.shared_1[105]*kernel.shared_1[((threadIdx.x*72) + 35)]))
+            conv2d_nchw_1[5] = (conv2d_nchw_1[5] + (pad_temp.shared_1[106]*kernel.shared_1[((threadIdx.x*72) + 35)]))
+            conv2d_nchw_1[6] = (conv2d_nchw_1[6] + (pad_temp.shared_1[107]*kernel.shared_1[((threadIdx.x*72) + 35)]))
+            conv2d_nchw_1[7] = (conv2d_nchw_1[7] + (pad_temp.shared_1[101]*kernel.shared_1[((threadIdx.x*72) + 4643)]))
+            conv2d_nchw_1[8] = (conv2d_nchw_1[8] + (pad_temp.shared_1[102]*kernel.shared_1[((threadIdx.x*72) + 4643)]))
+            conv2d_nchw_1[9] = (conv2d_nchw_1[9] + (pad_temp.shared_1[103]*kernel.shared_1[((threadIdx.x*72) + 4643)]))
+            conv2d_nchw_1[10] = (conv2d_nchw_1[10] + (pad_temp.shared_1[104]*kernel.shared_1[((threadIdx.x*72) + 4643)]))
+            conv2d_nchw_1[11] = (conv2d_nchw_1[11] + (pad_temp.shared_1[105]*kernel.shared_1[((threadIdx.x*72) + 4643)]))
+            conv2d_nchw_1[12] = (conv2d_nchw_1[12] + (pad_temp.shared_1[106]*kernel.shared_1[((threadIdx.x*72) + 4643)]))
+            conv2d_nchw_1[13] = (conv2d_nchw_1[13] + (pad_temp.shared_1[107]*kernel.shared_1[((threadIdx.x*72) + 4643)]))
+            conv2d_nchw_1[0] = (conv2d_nchw_1[0] + (pad_temp.shared_1[108]*kernel.shared_1[((threadIdx.x*72) + 36)]))
+            conv2d_nchw_1[1] = (conv2d_nchw_1[1] + (pad_temp.shared_1[109]*kernel.shared_1[((threadIdx.x*72) + 36)]))
+            conv2d_nchw_1[2] = (conv2d_nchw_1[2] + (pad_temp.shared_1[110]*kernel.shared_1[((threadIdx.x*72) + 36)]))
+            conv2d_nchw_1[3] = (conv2d_nchw_1[3] + (pad_temp.shared_1[111]*kernel.shared_1[((threadIdx.x*72) + 36)]))
+            conv2d_nchw_1[4] = (conv2d_nchw_1[4] + (pad_temp.shared_1[112]*kernel.shared_1[((threadIdx.x*72) + 36)]))
+            conv2d_nchw_1[5] = (conv2d_nchw_1[5] + (pad_temp.shared_1[113]*kernel.shared_1[((threadIdx.x*72) + 36)]))
+            conv2d_nchw_1[6] = (conv2d_nchw_1[6] + (pad_temp.shared_1[114]*kernel.shared_1[((threadIdx.x*72) + 36)]))
+            conv2d_nchw_1[7] = (conv2d_nchw_1[7] + (pad_temp.shared_1[108]*kernel.shared_1[((threadIdx.x*72) + 4644)]))
+            conv2d_nchw_1[8] = (conv2d_nchw_1[8] + (pad_temp.shared_1[109]*kernel.shared_1[((threadIdx.x*72) + 4644)]))
+            conv2d_nchw_1[9] = (conv2d_nchw_1[9] + (pad_temp.shared_1[110]*kernel.shared_1[((threadIdx.x*72) + 4644)]))
+            conv2d_nchw_1[10] = (conv2d_nchw_1[10] + (pad_temp.shared_1[111]*kernel.shared_1[((threadIdx.x*72) + 4644)]))
+            conv2d_nchw_1[11] = (conv2d_nchw_1[11] + (pad_temp.shared_1[112]*kernel.shared_1[((threadIdx.x*72) + 4644)]))
+            conv2d_nchw_1[12] = (conv2d_nchw_1[12] + (pad_temp.shared_1[113]*kernel.shared_1[((threadIdx.x*72) + 4644)]))
+            conv2d_nchw_1[13] = (conv2d_nchw_1[13] + (pad_temp.shared_1[114]*kernel.shared_1[((threadIdx.x*72) + 4644)]))
+            conv2d_nchw_1[0] = (conv2d_nchw_1[0] + (pad_temp.shared_1[109]*kernel.shared_1[((threadIdx.x*72) + 37)]))
+            conv2d_nchw_1[1] = (conv2d_nchw_1[1] + (pad_temp.shared_1[110]*kernel.shared_1[((threadIdx.x*72) + 37)]))
+            conv2d_nchw_1[2] = (conv2d_nchw_1[2] + (pad_temp.shared_1[111]*kernel.shared_1[((threadIdx.x*72) + 37)]))
+            conv2d_nchw_1[3] = (conv2d_nchw_1[3] + (pad_temp.shared_1[112]*kernel.shared_1[((threadIdx.x*72) + 37)]))
+            conv2d_nchw_1[4] = (conv2d_nchw_1[4] + (pad_temp.shared_1[113]*kernel.shared_1[((threadIdx.x*72) + 37)]))
+            conv2d_nchw_1[5] = (conv2d_nchw_1[5] + (pad_temp.shared_1[114]*kernel.shared_1[((threadIdx.x*72) + 37)]))
+            conv2d_nchw_1[6] = (conv2d_nchw_1[6] + (pad_temp.shared_1[115]*kernel.shared_1[((threadIdx.x*72) + 37)]))
+            conv2d_nchw_1[7] = (conv2d_nchw_1[7] + (pad_temp.shared_1[109]*kernel.shared_1[((threadIdx.x*72) + 4645)]))
+            conv2d_nchw_1[8] = (conv2d_nchw_1[8] + (pad_temp.shared_1[110]*kernel.shared_1[((threadIdx.x*72) + 4645)]))
+            conv2d_nchw_1[9] = (conv2d_nchw_1[9] + (pad_temp.shared_1[111]*kernel.shared_1[((threadIdx.x*72) + 4645)]))
+            conv2d_nchw_1[10] = (conv2d_nchw_1[10] + (pad_temp.shared_1[112]*kernel.shared_1[((threadIdx.x*72) + 4645)]))
+            conv2d_nchw_1[11] = (conv2d_nchw_1[11] + (pad_temp.shared_1[113]*kernel.shared_1[((threadIdx.x*72) + 4645)]))
+            conv2d_nchw_1[12] = (conv2d_nchw_1[12] + (pad_temp.shared_1[114]*kernel.shared_1[((threadIdx.x*72) + 4645)]))
+            conv2d_nchw_1[13] = (conv2d_nchw_1[13] + (pad_temp.shared_1[115]*kernel.shared_1[((threadIdx.x*72) + 4645)]))
+            conv2d_nchw_1[0] = (conv2d_nchw_1[0] + (pad_temp.shared_1[110]*kernel.shared_1[((threadIdx.x*72) + 38)]))
+            conv2d_nchw_1[1] = (conv2d_nchw_1[1] + (pad_temp.shared_1[111]*kernel.shared_1[((threadIdx.x*72) + 38)]))
+            conv2d_nchw_1[2] = (conv2d_nchw_1[2] + (pad_temp.shared_1[112]*kernel.shared_1[((threadIdx.x*72) + 38)]))
+            conv2d_nchw_1[3] = (conv2d_nchw_1[3] + (pad_temp.shared_1[113]*kernel.shared_1[((threadIdx.x*72) + 38)]))
+            conv2d_nchw_1[4] = (conv2d_nchw_1[4] + (pad_temp.shared_1[114]*kernel.shared_1[((threadIdx.x*72) + 38)]))
+            conv2d_nchw_1[5] = (conv2d_nchw_1[5] + (pad_temp.shared_1[115]*kernel.shared_1[((threadIdx.x*72) + 38)]))
+            conv2d_nchw_1[6] = (conv2d_nchw_1[6] + (pad_temp.shared_1[116]*kernel.shared_1[((threadIdx.x*72) + 38)]))
+            conv2d_nchw_1[7] = (conv2d_nchw_1[7] + (pad_temp.shared_1[110]*kernel.shared_1[((threadIdx.x*72) + 4646)]))
+            conv2d_nchw_1[8] = (conv2d_nchw_1[8] + (pad_temp.shared_1[111]*kernel.shared_1[((threadIdx.x*72) + 4646)]))
+            conv2d_nchw_1[9] = (conv2d_nchw_1[9] + (pad_temp.shared_1[112]*kernel.shared_1[((threadIdx.x*72) + 4646)]))
+            conv2d_nchw_1[10] = (conv2d_nchw_1[10] + (pad_temp.shared_1[113]*kernel.shared_1[((threadIdx.x*72) + 4646)]))
+            conv2d_nchw_1[11] = (conv2d_nchw_1[11] + (pad_temp.shared_1[114]*kernel.shared_1[((threadIdx.x*72) + 4646)]))
+            conv2d_nchw_1[12] = (conv2d_nchw_1[12] + (pad_temp.shared_1[115]*kernel.shared_1[((threadIdx.x*72) + 4646)]))
+            conv2d_nchw_1[13] = (conv2d_nchw_1[13] + (pad_temp.shared_1[116]*kernel.shared_1[((threadIdx.x*72) + 4646)]))
+            conv2d_nchw_1[0] = (conv2d_nchw_1[0] + (pad_temp.shared_1[117]*kernel.shared_1[((threadIdx.x*72) + 39)]))
+            conv2d_nchw_1[1] = (conv2d_nchw_1[1] + (pad_temp.shared_1[118]*kernel.shared_1[((threadIdx.x*72) + 39)]))
+            conv2d_nchw_1[2] = (conv2d_nchw_1[2] + (pad_temp.shared_1[119]*kernel.shared_1[((threadIdx.x*72) + 39)]))
+            conv2d_nchw_1[3] = (conv2d_nchw_1[3] + (pad_temp.shared_1[120]*kernel.shared_1[((threadIdx.x*72) + 39)]))
+            conv2d_nchw_1[4] = (conv2d_nchw_1[4] + (pad_temp.shared_1[121]*kernel.shared_1[((threadIdx.x*72) + 39)]))
+            conv2d_nchw_1[5] = (conv2d_nchw_1[5] + (pad_temp.shared_1[122]*kernel.shared_1[((threadIdx.x*72) + 39)]))
+            conv2d_nchw_1[6] = (conv2d_nchw_1[6] + (pad_temp.shared_1[123]*kernel.shared_1[((threadIdx.x*72) + 39)]))
+            conv2d_nchw_1[7] = (conv2d_nchw_1[7] + (pad_temp.shared_1[117]*kernel.shared_1[((threadIdx.x*72) + 4647)]))
+            conv2d_nchw_1[8] = (conv2d_nchw_1[8] + (pad_temp.shared_1[118]*kernel.shared_1[((threadIdx.x*72) + 4647)]))
+            conv2d_nchw_1[9] = (conv2d_nchw_1[9] + (pad_temp.shared_1[119]*kernel.shared_1[((threadIdx.x*72) + 4647)]))
+            conv2d_nchw_1[10] = (conv2d_nchw_1[10] + (pad_temp.shared_1[120]*kernel.shared_1[((threadIdx.x*72) + 4647)]))
+            conv2d_nchw_1[11] = (conv2d_nchw_1[11] + (pad_temp.shared_1[121]*kernel.shared_1[((threadIdx.x*72) + 4647)]))
+            conv2d_nchw_1[12] = (conv2d_nchw_1[12] + (pad_temp.shared_1[122]*kernel.shared_1[((threadIdx.x*72) + 4647)]))
+            conv2d_nchw_1[13] = (conv2d_nchw_1[13] + (pad_temp.shared_1[123]*kernel.shared_1[((threadIdx.x*72) + 4647)]))
+            conv2d_nchw_1[0] = (conv2d_nchw_1[0] + (pad_temp.shared_1[118]*kernel.shared_1[((threadIdx.x*72) + 40)]))
+            conv2d_nchw_1[1] = (conv2d_nchw_1[1] + (pad_temp.shared_1[119]*kernel.shared_1[((threadIdx.x*72) + 40)]))
+            conv2d_nchw_1[2] = (conv2d_nchw_1[2] + (pad_temp.shared_1[120]*kernel.shared_1[((threadIdx.x*72) + 40)]))
+            conv2d_nchw_1[3] = (conv2d_nchw_1[3] + (pad_temp.shared_1[121]*kernel.shared_1[((threadIdx.x*72) + 40)]))
+            conv2d_nchw_1[4] = (conv2d_nchw_1[4] + (pad_temp.shared_1[122]*kernel.shared_1[((threadIdx.x*72) + 40)]))
+            conv2d_nchw_1[5] = (conv2d_nchw_1[5] + (pad_temp.shared_1[123]*kernel.shared_1[((threadIdx.x*72) + 40)]))
+            conv2d_nchw_1[6] = (conv2d_nchw_1[6] + (pad_temp.shared_1[124]*kernel.shared_1[((threadIdx.x*72) + 40)]))
+            conv2d_nchw_1[7] = (conv2d_nchw_1[7] + (pad_temp.shared_1[118]*kernel.shared_1[((threadIdx.x*72) + 4648)]))
+            conv2d_nchw_1[8] = (conv2d_nchw_1[8] + (pad_temp.shared_1[119]*kernel.shared_1[((threadIdx.x*72) + 4648)]))
+            conv2d_nchw_1[9] = (conv2d_nchw_1[9] + (pad_temp.shared_1[120]*kernel.shared_1[((threadIdx.x*72) + 4648)]))
+            conv2d_nchw_1[10] = (conv2d_nchw_1[10] + (pad_temp.shared_1[121]*kernel.shared_1[((threadIdx.x*72) + 4648)]))
+            conv2d_nchw_1[11] = (conv2d_nchw_1[11] + (pad_temp.shared_1[122]*kernel.shared_1[((threadIdx.x*72) + 4648)]))
+            conv2d_nchw_1[12] = (conv2d_nchw_1[12] + (pad_temp.shared_1[123]*kernel.shared_1[((threadIdx.x*72) + 4648)]))
+            conv2d_nchw_1[13] = (conv2d_nchw_1[13] + (pad_temp.shared_1[124]*kernel.shared_1[((threadIdx.x*72) + 4648)]))
+            conv2d_nchw_1[0] = (conv2d_nchw_1[0] + (pad_temp.shared_1[119]*kernel.shared_1[((threadIdx.x*72) + 41)]))
+            conv2d_nchw_1[1] = (conv2d_nchw_1[1] + (pad_temp.shared_1[120]*kernel.shared_1[((threadIdx.x*72) + 41)]))
+            conv2d_nchw_1[2] = (conv2d_nchw_1[2] + (pad_temp.shared_1[121]*kernel.shared_1[((threadIdx.x*72) + 41)]))
+            conv2d_nchw_1[3] = (conv2d_nchw_1[3] + (pad_temp.shared_1[122]*kernel.shared_1[((threadIdx.x*72) + 41)]))
+            conv2d_nchw_1[4] = (conv2d_nchw_1[4] + (pad_temp.shared_1[123]*kernel.shared_1[((threadIdx.x*72) + 41)]))
+            conv2d_nchw_1[5] = (conv2d_nchw_1[5] + (pad_temp.shared_1[124]*kernel.shared_1[((threadIdx.x*72) + 41)]))
+            conv2d_nchw_1[6] = (conv2d_nchw_1[6] + (pad_temp.shared_1[125]*kernel.shared_1[((threadIdx.x*72) + 41)]))
+            conv2d_nchw_1[7] = (conv2d_nchw_1[7] + (pad_temp.shared_1[119]*kernel.shared_1[((threadIdx.x*72) + 4649)]))
+            conv2d_nchw_1[8] = (conv2d_nchw_1[8] + (pad_temp.shared_1[120]*kernel.shared_1[((threadIdx.x*72) + 4649)]))
+            conv2d_nchw_1[9] = (conv2d_nchw_1[9] + (pad_temp.shared_1[121]*kernel.shared_1[((threadIdx.x*72) + 4649)]))
+            conv2d_nchw_1[10] = (conv2d_nchw_1[10] + (pad_temp.shared_1[122]*kernel.shared_1[((threadIdx.x*72) + 4649)]))
+            conv2d_nchw_1[11] = (conv2d_nchw_1[11] + (pad_temp.shared_1[123]*kernel.shared_1[((threadIdx.x*72) + 4649)]))
+            conv2d_nchw_1[12] = (conv2d_nchw_1[12] + (pad_temp.shared_1[124]*kernel.shared_1[((threadIdx.x*72) + 4649)]))
+            conv2d_nchw_1[13] = (conv2d_nchw_1[13] + (pad_temp.shared_1[125]*kernel.shared_1[((threadIdx.x*72) + 4649)]))
+            conv2d_nchw_1[0] = (conv2d_nchw_1[0] + (pad_temp.shared_1[126]*kernel.shared_1[((threadIdx.x*72) + 42)]))
+            conv2d_nchw_1[1] = (conv2d_nchw_1[1] + (pad_temp.shared_1[127]*kernel.shared_1[((threadIdx.x*72) + 42)]))
+            conv2d_nchw_1[2] = (conv2d_nchw_1[2] + (pad_temp.shared_1[128]*kernel.shared_1[((threadIdx.x*72) + 42)]))
+            conv2d_nchw_1[3] = (conv2d_nchw_1[3] + (pad_temp.shared_1[129]*kernel.shared_1[((threadIdx.x*72) + 42)]))
+            conv2d_nchw_1[4] = (conv2d_nchw_1[4] + (pad_temp.shared_1[130]*kernel.shared_1[((threadIdx.x*72) + 42)]))
+            conv2d_nchw_1[5] = (conv2d_nchw_1[5] + (pad_temp.shared_1[131]*kernel.shared_1[((threadIdx.x*72) + 42)]))
+            conv2d_nchw_1[6] = (conv2d_nchw_1[6] + (pad_temp.shared_1[132]*kernel.shared_1[((threadIdx.x*72) + 42)]))
+            conv2d_nchw_1[7] = (conv2d_nchw_1[7] + (pad_temp.shared_1[126]*kernel.shared_1[((threadIdx.x*72) + 4650)]))
+            conv2d_nchw_1[8] = (conv2d_nchw_1[8] + (pad_temp.shared_1[127]*kernel.shared_1[((threadIdx.x*72) + 4650)]))
+            conv2d_nchw_1[9] = (conv2d_nchw_1[9] + (pad_temp.shared_1[128]*kernel.shared_1[((threadIdx.x*72) + 4650)]))
+            conv2d_nchw_1[10] = (conv2d_nchw_1[10] + (pad_temp.shared_1[129]*kernel.shared_1[((threadIdx.x*72) + 4650)]))
+            conv2d_nchw_1[11] = (conv2d_nchw_1[11] + (pad_temp.shared_1[130]*kernel.shared_1[((threadIdx.x*72) + 4650)]))
+            conv2d_nchw_1[12] = (conv2d_nchw_1[12] + (pad_temp.shared_1[131]*kernel.shared_1[((threadIdx.x*72) + 4650)]))
+            conv2d_nchw_1[13] = (conv2d_nchw_1[13] + (pad_temp.shared_1[132]*kernel.shared_1[((threadIdx.x*72) + 4650)]))
+            conv2d_nchw_1[0] = (conv2d_nchw_1[0] + (pad_temp.shared_1[127]*kernel.shared_1[((threadIdx.x*72) + 43)]))
+            conv2d_nchw_1[1] = (conv2d_nchw_1[1] + (pad_temp.shared_1[128]*kernel.shared_1[((threadIdx.x*72) + 43)]))
+            conv2d_nchw_1[2] = (conv2d_nchw_1[2] + (pad_temp.shared_1[129]*kernel.shared_1[((threadIdx.x*72) + 43)]))
+            conv2d_nchw_1[3] = (conv2d_nchw_1[3] + (pad_temp.shared_1[130]*kernel.shared_1[((threadIdx.x*72) + 43)]))
+            conv2d_nchw_1[4] = (conv2d_nchw_1[4] + (pad_temp.shared_1[131]*kernel.shared_1[((threadIdx.x*72) + 43)]))
+            conv2d_nchw_1[5] = (conv2d_nchw_1[5] + (pad_temp.shared_1[132]*kernel.shared_1[((threadIdx.x*72) + 43)]))
+            conv2d_nchw_1[6] = (conv2d_nchw_1[6] + (pad_temp.shared_1[133]*kernel.shared_1[((threadIdx.x*72) + 43)]))
+            conv2d_nchw_1[7] = (conv2d_nchw_1[7] + (pad_temp.shared_1[127]*kernel.shared_1[((threadIdx.x*72) + 4651)]))
+            conv2d_nchw_1[8] = (conv2d_nchw_1[8] + (pad_temp.shared_1[128]*kernel.shared_1[((threadIdx.x*72) + 4651)]))
+            conv2d_nchw_1[9] = (conv2d_nchw_1[9] + (pad_temp.shared_1[129]*kernel.shared_1[((threadIdx.x*72) + 4651)]))
+            conv2d_nchw_1[10] = (conv2d_nchw_1[10] + (pad_temp.shared_1[130]*kernel.shared_1[((threadIdx.x*72) + 4651)]))
+            conv2d_nchw_1[11] = (conv2d_nchw_1[11] + (pad_temp.shared_1[131]*kernel.shared_1[((threadIdx.x*72) + 4651)]))
+            conv2d_nchw_1[12] = (conv2d_nchw_1[12] + (pad_temp.shared_1[132]*kernel.shared_1[((threadIdx.x*72) + 4651)]))
+            conv2d_nchw_1[13] = (conv2d_nchw_1[13] + (pad_temp.shared_1[133]*kernel.shared_1[((threadIdx.x*72) + 4651)]))
+            conv2d_nchw_1[0] = (conv2d_nchw_1[0] + (pad_temp.shared_1[128]*kernel.shared_1[((threadIdx.x*72) + 44)]))
+            conv2d_nchw_1[1] = (conv2d_nchw_1[1] + (pad_temp.shared_1[129]*kernel.shared_1[((threadIdx.x*72) + 44)]))
+            conv2d_nchw_1[2] = (conv2d_nchw_1[2] + (pad_temp.shared_1[130]*kernel.shared_1[((threadIdx.x*72) + 44)]))
+            conv2d_nchw_1[3] = (conv2d_nchw_1[3] + (pad_temp.shared_1[131]*kernel.shared_1[((threadIdx.x*72) + 44)]))
+            conv2d_nchw_1[4] = (conv2d_nchw_1[4] + (pad_temp.shared_1[132]*kernel.shared_1[((threadIdx.x*72) + 44)]))
+            conv2d_nchw_1[5] = (conv2d_nchw_1[5] + (pad_temp.shared_1[133]*kernel.shared_1[((threadIdx.x*72) + 44)]))
+            conv2d_nchw_1[6] = (conv2d_nchw_1[6] + (pad_temp.shared_1[134]*kernel.shared_1[((threadIdx.x*72) + 44)]))
+            conv2d_nchw_1[7] = (conv2d_nchw_1[7] + (pad_temp.shared_1[128]*kernel.shared_1[((threadIdx.x*72) + 4652)]))
+            conv2d_nchw_1[8] = (conv2d_nchw_1[8] + (pad_temp.shared_1[129]*kernel.shared_1[((threadIdx.x*72) + 4652)]))
+            conv2d_nchw_1[9] = (conv2d_nchw_1[9] + (pad_temp.shared_1[130]*kernel.shared_1[((threadIdx.x*72) + 4652)]))
+            conv2d_nchw_1[10] = (conv2d_nchw_1[10] + (pad_temp.shared_1[131]*kernel.shared_1[((threadIdx.x*72) + 4652)]))
+            conv2d_nchw_1[11] = (conv2d_nchw_1[11] + (pad_temp.shared_1[132]*kernel.shared_1[((threadIdx.x*72) + 4652)]))
+            conv2d_nchw_1[12] = (conv2d_nchw_1[12] + (pad_temp.shared_1[133]*kernel.shared_1[((threadIdx.x*72) + 4652)]))
+            conv2d_nchw_1[13] = (conv2d_nchw_1[13] + (pad_temp.shared_1[134]*kernel.shared_1[((threadIdx.x*72) + 4652)]))
+            conv2d_nchw_1[0] = (conv2d_nchw_1[0] + (pad_temp.shared_1[135]*kernel.shared_1[((threadIdx.x*72) + 45)]))
+            conv2d_nchw_1[1] = (conv2d_nchw_1[1] + (pad_temp.shared_1[136]*kernel.shared_1[((threadIdx.x*72) + 45)]))
+            conv2d_nchw_1[2] = (conv2d_nchw_1[2] + (pad_temp.shared_1[137]*kernel.shared_1[((threadIdx.x*72) + 45)]))
+            conv2d_nchw_1[3] = (conv2d_nchw_1[3] + (pad_temp.shared_1[138]*kernel.shared_1[((threadIdx.x*72) + 45)]))
+            conv2d_nchw_1[4] = (conv2d_nchw_1[4] + (pad_temp.shared_1[139]*kernel.shared_1[((threadIdx.x*72) + 45)]))
+            conv2d_nchw_1[5] = (conv2d_nchw_1[5] + (pad_temp.shared_1[140]*kernel.shared_1[((threadIdx.x*72) + 45)]))
+            conv2d_nchw_1[6] = (conv2d_nchw_1[6] + (pad_temp.shared_1[141]*kernel.shared_1[((threadIdx.x*72) + 45)]))
+            conv2d_nchw_1[7] = (conv2d_nchw_1[7] + (pad_temp.shared_1[135]*kernel.shared_1[((threadIdx.x*72) + 4653)]))
+            conv2d_nchw_1[8] = (conv2d_nchw_1[8] + (pad_temp.shared_1[136]*kernel.shared_1[((threadIdx.x*72) + 4653)]))
+            conv2d_nchw_1[9] = (conv2d_nchw_1[9] + (pad_temp.shared_1[137]*kernel.shared_1[((threadIdx.x*72) + 4653)]))
+            conv2d_nchw_1[10] = (conv2d_nchw_1[10] + (pad_temp.shared_1[138]*kernel.shared_1[((threadIdx.x*72) + 4653)]))
+            conv2d_nchw_1[11] = (conv2d_nchw_1[11] + (pad_temp.shared_1[139]*kernel.shared_1[((threadIdx.x*72) + 4653)]))
+            conv2d_nchw_1[12] = (conv2d_nchw_1[12] + (pad_temp.shared_1[140]*kernel.shared_1[((threadIdx.x*72) + 4653)]))
+            conv2d_nchw_1[13] = (conv2d_nchw_1[13] + (pad_temp.shared_1[141]*kernel.shared_1[((threadIdx.x*72) + 4653)]))
+            conv2d_nchw_1[0] = (conv2d_nchw_1[0] + (pad_temp.shared_1[136]*kernel.shared_1[((threadIdx.x*72) + 46)]))
+            conv2d_nchw_1[1] = (conv2d_nchw_1[1] + (pad_temp.shared_1[137]*kernel.shared_1[((threadIdx.x*72) + 46)]))
+            conv2d_nchw_1[2] = (conv2d_nchw_1[2] + (pad_temp.shared_1[138]*kernel.shared_1[((threadIdx.x*72) + 46)]))
+            conv2d_nchw_1[3] = (conv2d_nchw_1[3] + (pad_temp.shared_1[139]*kernel.shared_1[((threadIdx.x*72) + 46)]))
+            conv2d_nchw_1[4] = (conv2d_nchw_1[4] + (pad_temp.shared_1[140]*kernel.shared_1[((threadIdx.x*72) + 46)]))
+            conv2d_nchw_1[5] = (conv2d_nchw_1[5] + (pad_temp.shared_1[141]*kernel.shared_1[((threadIdx.x*72) + 46)]))
+            conv2d_nchw_1[6] = (conv2d_nchw_1[6] + (pad_temp.shared_1[142]*kernel.shared_1[((threadIdx.x*72) + 46)]))
+            conv2d_nchw_1[7] = (conv2d_nchw_1[7] + (pad_temp.shared_1[136]*kernel.shared_1[((threadIdx.x*72) + 4654)]))
+            conv2d_nchw_1[8] = (conv2d_nchw_1[8] + (pad_temp.shared_1[137]*kernel.shared_1[((threadIdx.x*72) + 4654)]))
+            conv2d_nchw_1[9] = (conv2d_nchw_1[9] + (pad_temp.shared_1[138]*kernel.shared_1[((threadIdx.x*72) + 4654)]))
+            conv2d_nchw_1[10] = (conv2d_nchw_1[10] + (pad_temp.shared_1[139]*kernel.shared_1[((threadIdx.x*72) + 4654)]))
+            conv2d_nchw_1[11] = (conv2d_nchw_1[11] + (pad_temp.shared_1[140]*kernel.shared_1[((threadIdx.x*72) + 4654)]))
+            conv2d_nchw_1[12] = (conv2d_nchw_1[12] + (pad_temp.shared_1[141]*kernel.shared_1[((threadIdx.x*72) + 4654)]))
+            conv2d_nchw_1[13] = (conv2d_nchw_1[13] + (pad_temp.shared_1[142]*kernel.shared_1[((threadIdx.x*72) + 4654)]))
+            conv2d_nchw_1[0] = (conv2d_nchw_1[0] + (pad_temp.shared_1[137]*kernel.shared_1[((threadIdx.x*72) + 47)]))
+            conv2d_nchw_1[1] = (conv2d_nchw_1[1] + (pad_temp.shared_1[138]*kernel.shared_1[((threadIdx.x*72) + 47)]))
+            conv2d_nchw_1[2] = (conv2d_nchw_1[2] + (pad_temp.shared_1[139]*kernel.shared_1[((threadIdx.x*72) + 47)]))
+            conv2d_nchw_1[3] = (conv2d_nchw_1[3] + (pad_temp.shared_1[140]*kernel.shared_1[((threadIdx.x*72) + 47)]))
+            conv2d_nchw_1[4] = (conv2d_nchw_1[4] + (pad_temp.shared_1[141]*kernel.shared_1[((threadIdx.x*72) + 47)]))
+            conv2d_nchw_1[5] = (conv2d_nchw_1[5] + (pad_temp.shared_1[142]*kernel.shared_1[((threadIdx.x*72) + 47)]))
+            conv2d_nchw_1[6] = (conv2d_nchw_1[6] + (pad_temp.shared_1[143]*kernel.shared_1[((threadIdx.x*72) + 47)]))
+            conv2d_nchw_1[7] = (conv2d_nchw_1[7] + (pad_temp.shared_1[137]*kernel.shared_1[((threadIdx.x*72) + 4655)]))
+            conv2d_nchw_1[8] = (conv2d_nchw_1[8] + (pad_temp.shared_1[138]*kernel.shared_1[((threadIdx.x*72) + 4655)]))
+            conv2d_nchw_1[9] = (conv2d_nchw_1[9] + (pad_temp.shared_1[139]*kernel.shared_1[((threadIdx.x*72) + 4655)]))
+            conv2d_nchw_1[10] = (conv2d_nchw_1[10] + (pad_temp.shared_1[140]*kernel.shared_1[((threadIdx.x*72) + 4655)]))
+            conv2d_nchw_1[11] = (conv2d_nchw_1[11] + (pad_temp.shared_1[141]*kernel.shared_1[((threadIdx.x*72) + 4655)]))
+            conv2d_nchw_1[12] = (conv2d_nchw_1[12] + (pad_temp.shared_1[142]*kernel.shared_1[((threadIdx.x*72) + 4655)]))
+            conv2d_nchw_1[13] = (conv2d_nchw_1[13] + (pad_temp.shared_1[143]*kernel.shared_1[((threadIdx.x*72) + 4655)]))
+            conv2d_nchw_1[0] = (conv2d_nchw_1[0] + (pad_temp.shared_1[144]*kernel.shared_1[((threadIdx.x*72) + 48)]))
+            conv2d_nchw_1[1] = (conv2d_nchw_1[1] + (pad_temp.shared_1[145]*kernel.shared_1[((threadIdx.x*72) + 48)]))
+            conv2d_nchw_1[2] = (conv2d_nchw_1[2] + (pad_temp.shared_1[146]*kernel.shared_1[((threadIdx.x*72) + 48)]))
+            conv2d_nchw_1[3] = (conv2d_nchw_1[3] + (pad_temp.shared_1[147]*kernel.shared_1[((threadIdx.x*72) + 48)]))
+            conv2d_nchw_1[4] = (conv2d_nchw_1[4] + (pad_temp.shared_1[148]*kernel.shared_1[((threadIdx.x*72) + 48)]))
+            conv2d_nchw_1[5] = (conv2d_nchw_1[5] + (pad_temp.shared_1[149]*kernel.shared_1[((threadIdx.x*72) + 48)]))
+            conv2d_nchw_1[6] = (conv2d_nchw_1[6] + (pad_temp.shared_1[150]*kernel.shared_1[((threadIdx.x*72) + 48)]))
+            conv2d_nchw_1[7] = (conv2d_nchw_1[7] + (pad_temp.shared_1[144]*kernel.shared_1[((threadIdx.x*72) + 4656)]))
+            conv2d_nchw_1[8] = (conv2d_nchw_1[8] + (pad_temp.shared_1[145]*kernel.shared_1[((threadIdx.x*72) + 4656)]))
+            conv2d_nchw_1[9] = (conv2d_nchw_1[9] + (pad_temp.shared_1[146]*kernel.shared_1[((threadIdx.x*72) + 4656)]))
+            conv2d_nchw_1[10] = (conv2d_nchw_1[10] + (pad_temp.shared_1[147]*kernel.shared_1[((threadIdx.x*72) + 4656)]))
+            conv2d_nchw_1[11] = (conv2d_nchw_1[11] + (pad_temp.shared_1[148]*kernel.shared_1[((threadIdx.x*72) + 4656)]))
+            conv2d_nchw_1[12] = (conv2d_nchw_1[12] + (pad_temp.shared_1[149]*kernel.shared_1[((threadIdx.x*72) + 4656)]))
+            conv2d_nchw_1[13] = (conv2d_nchw_1[13] + (pad_temp.shared_1[150]*kernel.shared_1[((threadIdx.x*72) + 4656)]))
+            conv2d_nchw_1[0] = (conv2d_nchw_1[0] + (pad_temp.shared_1[145]*kernel.shared_1[((threadIdx.x*72) + 49)]))
+            conv2d_nchw_1[1] = (conv2d_nchw_1[1] + (pad_temp.shared_1[146]*kernel.shared_1[((threadIdx.x*72) + 49)]))
+            conv2d_nchw_1[2] = (conv2d_nchw_1[2] + (pad_temp.shared_1[147]*kernel.shared_1[((threadIdx.x*72) + 49)]))
+            conv2d_nchw_1[3] = (conv2d_nchw_1[3] + (pad_temp.shared_1[148]*kernel.shared_1[((threadIdx.x*72) + 49)]))
+            conv2d_nchw_1[4] = (conv2d_nchw_1[4] + (pad_temp.shared_1[149]*kernel.shared_1[((threadIdx.x*72) + 49)]))
+            conv2d_nchw_1[5] = (conv2d_nchw_1[5] + (pad_temp.shared_1[150]*kernel.shared_1[((threadIdx.x*72) + 49)]))
+            conv2d_nchw_1[6] = (conv2d_nchw_1[6] + (pad_temp.shared_1[151]*kernel.shared_1[((threadIdx.x*72) + 49)]))
+            conv2d_nchw_1[7] = (conv2d_nchw_1[7] + (pad_temp.shared_1[145]*kernel.shared_1[((threadIdx.x*72) + 4657)]))
+            conv2d_nchw_1[8] = (conv2d_nchw_1[8] + (pad_temp.shared_1[146]*kernel.shared_1[((threadIdx.x*72) + 4657)]))
+            conv2d_nchw_1[9] = (conv2d_nchw_1[9] + (pad_temp.shared_1[147]*kernel.shared_1[((threadIdx.x*72) + 4657)]))
+            conv2d_nchw_1[10] = (conv2d_nchw_1[10] + (pad_temp.shared_1[148]*kernel.shared_1[((threadIdx.x*72) + 4657)]))
+            conv2d_nchw_1[11] = (conv2d_nchw_1[11] + (pad_temp.shared_1[149]*kernel.shared_1[((threadIdx.x*72) + 4657)]))
+            conv2d_nchw_1[12] = (conv2d_nchw_1[12] + (pad_temp.shared_1[150]*kernel.shared_1[((threadIdx.x*72) + 4657)]))
+            conv2d_nchw_1[13] = (conv2d_nchw_1[13] + (pad_temp.shared_1[151]*kernel.shared_1[((threadIdx.x*72) + 4657)]))
+            conv2d_nchw_1[0] = (conv2d_nchw_1[0] + (pad_temp.shared_1[146]*kernel.shared_1[((threadIdx.x*72) + 50)]))
+            conv2d_nchw_1[1] = (conv2d_nchw_1[1] + (pad_temp.shared_1[147]*kernel.shared_1[((threadIdx.x*72) + 50)]))
+            conv2d_nchw_1[2] = (conv2d_nchw_1[2] + (pad_temp.shared_1[148]*kernel.shared_1[((threadIdx.x*72) + 50)]))
+            conv2d_nchw_1[3] = (conv2d_nchw_1[3] + (pad_temp.shared_1[149]*kernel.shared_1[((threadIdx.x*72) + 50)]))
+            conv2d_nchw_1[4] = (conv2d_nchw_1[4] + (pad_temp.shared_1[150]*kernel.shared_1[((threadIdx.x*72) + 50)]))
+            conv2d_nchw_1[5] = (conv2d_nchw_1[5] + (pad_temp.shared_1[151]*kernel.shared_1[((threadIdx.x*72) + 50)]))
+            conv2d_nchw_1[6] = (conv2d_nchw_1[6] + (pad_temp.shared_1[152]*kernel.shared_1[((threadIdx.x*72) + 50)]))
+            conv2d_nchw_1[7] = (conv2d_nchw_1[7] + (pad_temp.shared_1[146]*kernel.shared_1[((threadIdx.x*72) + 4658)]))
+            conv2d_nchw_1[8] = (conv2d_nchw_1[8] + (pad_temp.shared_1[147]*kernel.shared_1[((threadIdx.x*72) + 4658)]))
+            conv2d_nchw_1[9] = (conv2d_nchw_1[9] + (pad_temp.shared_1[148]*kernel.shared_1[((threadIdx.x*72) + 4658)]))
+            conv2d_nchw_1[10] = (conv2d_nchw_1[10] + (pad_temp.shared_1[149]*kernel.shared_1[((threadIdx.x*72) + 4658)]))
+            conv2d_nchw_1[11] = (conv2d_nchw_1[11] + (pad_temp.shared_1[150]*kernel.shared_1[((threadIdx.x*72) + 4658)]))
+            conv2d_nchw_1[12] = (conv2d_nchw_1[12] + (pad_temp.shared_1[151]*kernel.shared_1[((threadIdx.x*72) + 4658)]))
+            conv2d_nchw_1[13] = (conv2d_nchw_1[13] + (pad_temp.shared_1[152]*kernel.shared_1[((threadIdx.x*72) + 4658)]))
+            conv2d_nchw_1[0] = (conv2d_nchw_1[0] + (pad_temp.shared_1[153]*kernel.shared_1[((threadIdx.x*72) + 51)]))
+            conv2d_nchw_1[1] = (conv2d_nchw_1[1] + (pad_temp.shared_1[154]*kernel.shared_1[((threadIdx.x*72) + 51)]))
+            conv2d_nchw_1[2] = (conv2d_nchw_1[2] + (pad_temp.shared_1[155]*kernel.shared_1[((threadIdx.x*72) + 51)]))
+            conv2d_nchw_1[3] = (conv2d_nchw_1[3] + (pad_temp.shared_1[156]*kernel.shared_1[((threadIdx.x*72) + 51)]))
+            conv2d_nchw_1[4] = (conv2d_nchw_1[4] + (pad_temp.shared_1[157]*kernel.shared_1[((threadIdx.x*72) + 51)]))
+            conv2d_nchw_1[5] = (conv2d_nchw_1[5] + (pad_temp.shared_1[158]*kernel.shared_1[((threadIdx.x*72) + 51)]))
+            conv2d_nchw_1[6] = (conv2d_nchw_1[6] + (pad_temp.shared_1[159]*kernel.shared_1[((threadIdx.x*72) + 51)]))
+            conv2d_nchw_1[7] = (conv2d_nchw_1[7] + (pad_temp.shared_1[153]*kernel.shared_1[((threadIdx.x*72) + 4659)]))
+            conv2d_nchw_1[8] = (conv2d_nchw_1[8] + (pad_temp.shared_1[154]*kernel.shared_1[((threadIdx.x*72) + 4659)]))
+            conv2d_nchw_1[9] = (conv2d_nchw_1[9] + (pad_temp.shared_1[155]*kernel.shared_1[((threadIdx.x*72) + 4659)]))
+            conv2d_nchw_1[10] = (conv2d_nchw_1[10] + (pad_temp.shared_1[156]*kernel.shared_1[((threadIdx.x*72) + 4659)]))
+            conv2d_nchw_1[11] = (conv2d_nchw_1[11] + (pad_temp.shared_1[157]*kernel.shared_1[((threadIdx.x*72) + 4659)]))
+            conv2d_nchw_1[12] = (conv2d_nchw_1[12] + (pad_temp.shared_1[158]*kernel.shared_1[((threadIdx.x*72) + 4659)]))
+            conv2d_nchw_1[13] = (conv2d_nchw_1[13] + (pad_temp.shared_1[159]*kernel.shared_1[((threadIdx.x*72) + 4659)]))
+            conv2d_nchw_1[0] = (conv2d_nchw_1[0] + (pad_temp.shared_1[154]*kernel.shared_1[((threadIdx.x*72) + 52)]))
+            conv2d_nchw_1[1] = (conv2d_nchw_1[1] + (pad_temp.shared_1[155]*kernel.shared_1[((threadIdx.x*72) + 52)]))
+            conv2d_nchw_1[2] = (conv2d_nchw_1[2] + (pad_temp.shared_1[156]*kernel.shared_1[((threadIdx.x*72) + 52)]))
+            conv2d_nchw_1[3] = (conv2d_nchw_1[3] + (pad_temp.shared_1[157]*kernel.shared_1[((threadIdx.x*72) + 52)]))
+            conv2d_nchw_1[4] = (conv2d_nchw_1[4] + (pad_temp.shared_1[158]*kernel.shared_1[((threadIdx.x*72) + 52)]))
+            conv2d_nchw_1[5] = (conv2d_nchw_1[5] + (pad_temp.shared_1[159]*kernel.shared_1[((threadIdx.x*72) + 52)]))
+            conv2d_nchw_1[6] = (conv2d_nchw_1[6] + (pad_temp.shared_1[160]*kernel.shared_1[((threadIdx.x*72) + 52)]))
+            conv2d_nchw_1[7] = (conv2d_nchw_1[7] + (pad_temp.shared_1[154]*kernel.shared_1[((threadIdx.x*72) + 4660)]))
+            conv2d_nchw_1[8] = (conv2d_nchw_1[8] + (pad_temp.shared_1[155]*kernel.shared_1[((threadIdx.x*72) + 4660)]))
+            conv2d_nchw_1[9] = (conv2d_nchw_1[9] + (pad_temp.shared_1[156]*kernel.shared_1[((threadIdx.x*72) + 4660)]))
+            conv2d_nchw_1[10] = (conv2d_nchw_1[10] + (pad_temp.shared_1[157]*kernel.shared_1[((threadIdx.x*72) + 4660)]))
+            conv2d_nchw_1[11] = (conv2d_nchw_1[11] + (pad_temp.shared_1[158]*kernel.shared_1[((threadIdx.x*72) + 4660)]))
+            conv2d_nchw_1[12] = (conv2d_nchw_1[12] + (pad_temp.shared_1[159]*kernel.shared_1[((threadIdx.x*72) + 4660)]))
+            conv2d_nchw_1[13] = (conv2d_nchw_1[13] + (pad_temp.shared_1[160]*kernel.shared_1[((threadIdx.x*72) + 4660)]))
+            conv2d_nchw_1[0] = (conv2d_nchw_1[0] + (pad_temp.shared_1[155]*kernel.shared_1[((threadIdx.x*72) + 53)]))
+            conv2d_nchw_1[1] = (conv2d_nchw_1[1] + (pad_temp.shared_1[156]*kernel.shared_1[((threadIdx.x*72) + 53)]))
+            conv2d_nchw_1[2] = (conv2d_nchw_1[2] + (pad_temp.shared_1[157]*kernel.shared_1[((threadIdx.x*72) + 53)]))
+            conv2d_nchw_1[3] = (conv2d_nchw_1[3] + (pad_temp.shared_1[158]*kernel.shared_1[((threadIdx.x*72) + 53)]))
+            conv2d_nchw_1[4] = (conv2d_nchw_1[4] + (pad_temp.shared_1[159]*kernel.shared_1[((threadIdx.x*72) + 53)]))
+            conv2d_nchw_1[5] = (conv2d_nchw_1[5] + (pad_temp.shared_1[160]*kernel.shared_1[((threadIdx.x*72) + 53)]))
+            conv2d_nchw_1[6] = (conv2d_nchw_1[6] + (pad_temp.shared_1[161]*kernel.shared_1[((threadIdx.x*72) + 53)]))
+            conv2d_nchw_1[7] = (conv2d_nchw_1[7] + (pad_temp.shared_1[155]*kernel.shared_1[((threadIdx.x*72) + 4661)]))
+            conv2d_nchw_1[8] = (conv2d_nchw_1[8] + (pad_temp.shared_1[156]*kernel.shared_1[((threadIdx.x*72) + 4661)]))
+            conv2d_nchw_1[9] = (conv2d_nchw_1[9] + (pad_temp.shared_1[157]*kernel.shared_1[((threadIdx.x*72) + 4661)]))
+            conv2d_nchw_1[10] = (conv2d_nchw_1[10] + (pad_temp.shared_1[158]*kernel.shared_1[((threadIdx.x*72) + 4661)]))
+            conv2d_nchw_1[11] = (conv2d_nchw_1[11] + (pad_temp.shared_1[159]*kernel.shared_1[((threadIdx.x*72) + 4661)]))
+            conv2d_nchw_1[12] = (conv2d_nchw_1[12] + (pad_temp.shared_1[160]*kernel.shared_1[((threadIdx.x*72) + 4661)]))
+            conv2d_nchw_1[13] = (conv2d_nchw_1[13] + (pad_temp.shared_1[161]*kernel.shared_1[((threadIdx.x*72) + 4661)]))
+            conv2d_nchw_1[0] = (conv2d_nchw_1[0] + (pad_temp.shared_1[162]*kernel.shared_1[((threadIdx.x*72) + 54)]))
+            conv2d_nchw_1[1] = (conv2d_nchw_1[1] + (pad_temp.shared_1[163]*kernel.shared_1[((threadIdx.x*72) + 54)]))
+            conv2d_nchw_1[2] = (conv2d_nchw_1[2] + (pad_temp.shared_1[164]*kernel.shared_1[((threadIdx.x*72) + 54)]))
+            conv2d_nchw_1[3] = (conv2d_nchw_1[3] + (pad_temp.shared_1[165]*kernel.shared_1[((threadIdx.x*72) + 54)]))
+            conv2d_nchw_1[4] = (conv2d_nchw_1[4] + (pad_temp.shared_1[166]*kernel.shared_1[((threadIdx.x*72) + 54)]))
+            conv2d_nchw_1[5] = (conv2d_nchw_1[5] + (pad_temp.shared_1[167]*kernel.shared_1[((threadIdx.x*72) + 54)]))
+            conv2d_nchw_1[6] = (conv2d_nchw_1[6] + (pad_temp.shared_1[168]*kernel.shared_1[((threadIdx.x*72) + 54)]))
+            conv2d_nchw_1[7] = (conv2d_nchw_1[7] + (pad_temp.shared_1[162]*kernel.shared_1[((threadIdx.x*72) + 4662)]))
+            conv2d_nchw_1[8] = (conv2d_nchw_1[8] + (pad_temp.shared_1[163]*kernel.shared_1[((threadIdx.x*72) + 4662)]))
+            conv2d_nchw_1[9] = (conv2d_nchw_1[9] + (pad_temp.shared_1[164]*kernel.shared_1[((threadIdx.x*72) + 4662)]))
+            conv2d_nchw_1[10] = (conv2d_nchw_1[10] + (pad_temp.shared_1[165]*kernel.shared_1[((threadIdx.x*72) + 4662)]))
+            conv2d_nchw_1[11] = (conv2d_nchw_1[11] + (pad_temp.shared_1[166]*kernel.shared_1[((threadIdx.x*72) + 4662)]))
+            conv2d_nchw_1[12] = (conv2d_nchw_1[12] + (pad_temp.shared_1[167]*kernel.shared_1[((threadIdx.x*72) + 4662)]))
+            conv2d_nchw_1[13] = (conv2d_nchw_1[13] + (pad_temp.shared_1[168]*kernel.shared_1[((threadIdx.x*72) + 4662)]))
+            conv2d_nchw_1[0] = (conv2d_nchw_1[0] + (pad_temp.shared_1[163]*kernel.shared_1[((threadIdx.x*72) + 55)]))
+            conv2d_nchw_1[1] = (conv2d_nchw_1[1] + (pad_temp.shared_1[164]*kernel.shared_1[((threadIdx.x*72) + 55)]))
+            conv2d_nchw_1[2] = (conv2d_nchw_1[2] + (pad_temp.shared_1[165]*kernel.shared_1[((threadIdx.x*72) + 55)]))
+            conv2d_nchw_1[3] = (conv2d_nchw_1[3] + (pad_temp.shared_1[166]*kernel.shared_1[((threadIdx.x*72) + 55)]))
+            conv2d_nchw_1[4] = (conv2d_nchw_1[4] + (pad_temp.shared_1[167]*kernel.shared_1[((threadIdx.x*72) + 55)]))
+            conv2d_nchw_1[5] = (conv2d_nchw_1[5] + (pad_temp.shared_1[168]*kernel.shared_1[((threadIdx.x*72) + 55)]))
+            conv2d_nchw_1[6] = (conv2d_nchw_1[6] + (pad_temp.shared_1[169]*kernel.shared_1[((threadIdx.x*72) + 55)]))
+            conv2d_nchw_1[7] = (conv2d_nchw_1[7] + (pad_temp.shared_1[163]*kernel.shared_1[((threadIdx.x*72) + 4663)]))
+            conv2d_nchw_1[8] = (conv2d_nchw_1[8] + (pad_temp.shared_1[164]*kernel.shared_1[((threadIdx.x*72) + 4663)]))
+            conv2d_nchw_1[9] = (conv2d_nchw_1[9] + (pad_temp.shared_1[165]*kernel.shared_1[((threadIdx.x*72) + 4663)]))
+            conv2d_nchw_1[10] = (conv2d_nchw_1[10] + (pad_temp.shared_1[166]*kernel.shared_1[((threadIdx.x*72) + 4663)]))
+            conv2d_nchw_1[11] = (conv2d_nchw_1[11] + (pad_temp.shared_1[167]*kernel.shared_1[((threadIdx.x*72) + 4663)]))
+            conv2d_nchw_1[12] = (conv2d_nchw_1[12] + (pad_temp.shared_1[168]*kernel.shared_1[((threadIdx.x*72) + 4663)]))
+            conv2d_nchw_1[13] = (conv2d_nchw_1[13] + (pad_temp.shared_1[169]*kernel.shared_1[((threadIdx.x*72) + 4663)]))
+            conv2d_nchw_1[0] = (conv2d_nchw_1[0] + (pad_temp.shared_1[164]*kernel.shared_1[((threadIdx.x*72) + 56)]))
+            conv2d_nchw_1[1] = (conv2d_nchw_1[1] + (pad_temp.shared_1[165]*kernel.shared_1[((threadIdx.x*72) + 56)]))
+            conv2d_nchw_1[2] = (conv2d_nchw_1[2] + (pad_temp.shared_1[166]*kernel.shared_1[((threadIdx.x*72) + 56)]))
+            conv2d_nchw_1[3] = (conv2d_nchw_1[3] + (pad_temp.shared_1[167]*kernel.shared_1[((threadIdx.x*72) + 56)]))
+            conv2d_nchw_1[4] = (conv2d_nchw_1[4] + (pad_temp.shared_1[168]*kernel.shared_1[((threadIdx.x*72) + 56)]))
+            conv2d_nchw_1[5] = (conv2d_nchw_1[5] + (pad_temp.shared_1[169]*kernel.shared_1[((threadIdx.x*72) + 56)]))
+            conv2d_nchw_1[6] = (conv2d_nchw_1[6] + (pad_temp.shared_1[170]*kernel.shared_1[((threadIdx.x*72) + 56)]))
+            conv2d_nchw_1[7] = (conv2d_nchw_1[7] + (pad_temp.shared_1[164]*kernel.shared_1[((threadIdx.x*72) + 4664)]))
+            conv2d_nchw_1[8] = (conv2d_nchw_1[8] + (pad_temp.shared_1[165]*kernel.shared_1[((threadIdx.x*72) + 4664)]))
+            conv2d_nchw_1[9] = (conv2d_nchw_1[9] + (pad_temp.shared_1[166]*kernel.shared_1[((threadIdx.x*72) + 4664)]))
+            conv2d_nchw_1[10] = (conv2d_nchw_1[10] + (pad_temp.shared_1[167]*kernel.shared_1[((threadIdx.x*72) + 4664)]))
+            conv2d_nchw_1[11] = (conv2d_nchw_1[11] + (pad_temp.shared_1[168]*kernel.shared_1[((threadIdx.x*72) + 4664)]))
+            conv2d_nchw_1[12] = (conv2d_nchw_1[12] + (pad_temp.shared_1[169]*kernel.shared_1[((threadIdx.x*72) + 4664)]))
+            conv2d_nchw_1[13] = (conv2d_nchw_1[13] + (pad_temp.shared_1[170]*kernel.shared_1[((threadIdx.x*72) + 4664)]))
+            conv2d_nchw_1[0] = (conv2d_nchw_1[0] + (pad_temp.shared_1[171]*kernel.shared_1[((threadIdx.x*72) + 57)]))
+            conv2d_nchw_1[1] = (conv2d_nchw_1[1] + (pad_temp.shared_1[172]*kernel.shared_1[((threadIdx.x*72) + 57)]))
+            conv2d_nchw_1[2] = (conv2d_nchw_1[2] + (pad_temp.shared_1[173]*kernel.shared_1[((threadIdx.x*72) + 57)]))
+            conv2d_nchw_1[3] = (conv2d_nchw_1[3] + (pad_temp.shared_1[174]*kernel.shared_1[((threadIdx.x*72) + 57)]))
+            conv2d_nchw_1[4] = (conv2d_nchw_1[4] + (pad_temp.shared_1[175]*kernel.shared_1[((threadIdx.x*72) + 57)]))
+            conv2d_nchw_1[5] = (conv2d_nchw_1[5] + (pad_temp.shared_1[176]*kernel.shared_1[((threadIdx.x*72) + 57)]))
+            conv2d_nchw_1[6] = (conv2d_nchw_1[6] + (pad_temp.shared_1[177]*kernel.shared_1[((threadIdx.x*72) + 57)]))
+            conv2d_nchw_1[7] = (conv2d_nchw_1[7] + (pad_temp.shared_1[171]*kernel.shared_1[((threadIdx.x*72) + 4665)]))
+            conv2d_nchw_1[8] = (conv2d_nchw_1[8] + (pad_temp.shared_1[172]*kernel.shared_1[((threadIdx.x*72) + 4665)]))
+            conv2d_nchw_1[9] = (conv2d_nchw_1[9] + (pad_temp.shared_1[173]*kernel.shared_1[((threadIdx.x*72) + 4665)]))
+            conv2d_nchw_1[10] = (conv2d_nchw_1[10] + (pad_temp.shared_1[174]*kernel.shared_1[((threadIdx.x*72) + 4665)]))
+            conv2d_nchw_1[11] = (conv2d_nchw_1[11] + (pad_temp.shared_1[175]*kernel.shared_1[((threadIdx.x*72) + 4665)]))
+            conv2d_nchw_1[12] = (conv2d_nchw_1[12] + (pad_temp.shared_1[176]*kernel.shared_1[((threadIdx.x*72) + 4665)]))
+            conv2d_nchw_1[13] = (conv2d_nchw_1[13] + (pad_temp.shared_1[177]*kernel.shared_1[((threadIdx.x*72) + 4665)]))
+            conv2d_nchw_1[0] = (conv2d_nchw_1[0] + (pad_temp.shared_1[172]*kernel.shared_1[((threadIdx.x*72) + 58)]))
+            conv2d_nchw_1[1] = (conv2d_nchw_1[1] + (pad_temp.shared_1[173]*kernel.shared_1[((threadIdx.x*72) + 58)]))
+            conv2d_nchw_1[2] = (conv2d_nchw_1[2] + (pad_temp.shared_1[174]*kernel.shared_1[((threadIdx.x*72) + 58)]))
+            conv2d_nchw_1[3] = (conv2d_nchw_1[3] + (pad_temp.shared_1[175]*kernel.shared_1[((threadIdx.x*72) + 58)]))
+            conv2d_nchw_1[4] = (conv2d_nchw_1[4] + (pad_temp.shared_1[176]*kernel.shared_1[((threadIdx.x*72) + 58)]))
+            conv2d_nchw_1[5] = (conv2d_nchw_1[5] + (pad_temp.shared_1[177]*kernel.shared_1[((threadIdx.x*72) + 58)]))
+            conv2d_nchw_1[6] = (conv2d_nchw_1[6] + (pad_temp.shared_1[178]*kernel.shared_1[((threadIdx.x*72) + 58)]))
+            conv2d_nchw_1[7] = (conv2d_nchw_1[7] + (pad_temp.shared_1[172]*kernel.shared_1[((threadIdx.x*72) + 4666)]))
+            conv2d_nchw_1[8] = (conv2d_nchw_1[8] + (pad_temp.shared_1[173]*kernel.shared_1[((threadIdx.x*72) + 4666)]))
+            conv2d_nchw_1[9] = (conv2d_nchw_1[9] + (pad_temp.shared_1[174]*kernel.shared_1[((threadIdx.x*72) + 4666)]))
+            conv2d_nchw_1[10] = (conv2d_nchw_1[10] + (pad_temp.shared_1[175]*kernel.shared_1[((threadIdx.x*72) + 4666)]))
+            conv2d_nchw_1[11] = (conv2d_nchw_1[11] + (pad_temp.shared_1[176]*kernel.shared_1[((threadIdx.x*72) + 4666)]))
+            conv2d_nchw_1[12] = (conv2d_nchw_1[12] + (pad_temp.shared_1[177]*kernel.shared_1[((threadIdx.x*72) + 4666)]))
+            conv2d_nchw_1[13] = (conv2d_nchw_1[13] + (pad_temp.shared_1[178]*kernel.shared_1[((threadIdx.x*72) + 4666)]))
+            conv2d_nchw_1[0] = (conv2d_nchw_1[0] + (pad_temp.shared_1[173]*kernel.shared_1[((threadIdx.x*72) + 59)]))
+            conv2d_nchw_1[1] = (conv2d_nchw_1[1] + (pad_temp.shared_1[174]*kernel.shared_1[((threadIdx.x*72) + 59)]))
+            conv2d_nchw_1[2] = (conv2d_nchw_1[2] + (pad_temp.shared_1[175]*kernel.shared_1[((threadIdx.x*72) + 59)]))
+            conv2d_nchw_1[3] = (conv2d_nchw_1[3] + (pad_temp.shared_1[176]*kernel.shared_1[((threadIdx.x*72) + 59)]))
+            conv2d_nchw_1[4] = (conv2d_nchw_1[4] + (pad_temp.shared_1[177]*kernel.shared_1[((threadIdx.x*72) + 59)]))
+            conv2d_nchw_1[5] = (conv2d_nchw_1[5] + (pad_temp.shared_1[178]*kernel.shared_1[((threadIdx.x*72) + 59)]))
+            conv2d_nchw_1[6] = (conv2d_nchw_1[6] + (pad_temp.shared_1[179]*kernel.shared_1[((threadIdx.x*72) + 59)]))
+            conv2d_nchw_1[7] = (conv2d_nchw_1[7] + (pad_temp.shared_1[173]*kernel.shared_1[((threadIdx.x*72) + 4667)]))
+            conv2d_nchw_1[8] = (conv2d_nchw_1[8] + (pad_temp.shared_1[174]*kernel.shared_1[((threadIdx.x*72) + 4667)]))
+            conv2d_nchw_1[9] = (conv2d_nchw_1[9] + (pad_temp.shared_1[175]*kernel.shared_1[((threadIdx.x*72) + 4667)]))
+            conv2d_nchw_1[10] = (conv2d_nchw_1[10] + (pad_temp.shared_1[176]*kernel.shared_1[((threadIdx.x*72) + 4667)]))
+            conv2d_nchw_1[11] = (conv2d_nchw_1[11] + (pad_temp.shared_1[177]*kernel.shared_1[((threadIdx.x*72) + 4667)]))
+            conv2d_nchw_1[12] = (conv2d_nchw_1[12] + (pad_temp.shared_1[178]*kernel.shared_1[((threadIdx.x*72) + 4667)]))
+            conv2d_nchw_1[13] = (conv2d_nchw_1[13] + (pad_temp.shared_1[179]*kernel.shared_1[((threadIdx.x*72) + 4667)]))
+            conv2d_nchw_1[0] = (conv2d_nchw_1[0] + (pad_temp.shared_1[180]*kernel.shared_1[((threadIdx.x*72) + 60)]))
+            conv2d_nchw_1[1] = (conv2d_nchw_1[1] + (pad_temp.shared_1[181]*kernel.shared_1[((threadIdx.x*72) + 60)]))
+            conv2d_nchw_1[2] = (conv2d_nchw_1[2] + (pad_temp.shared_1[182]*kernel.shared_1[((threadIdx.x*72) + 60)]))
+            conv2d_nchw_1[3] = (conv2d_nchw_1[3] + (pad_temp.shared_1[183]*kernel.shared_1[((threadIdx.x*72) + 60)]))
+            conv2d_nchw_1[4] = (conv2d_nchw_1[4] + (pad_temp.shared_1[184]*kernel.shared_1[((threadIdx.x*72) + 60)]))
+            conv2d_nchw_1[5] = (conv2d_nchw_1[5] + (pad_temp.shared_1[185]*kernel.shared_1[((threadIdx.x*72) + 60)]))
+            conv2d_nchw_1[6] = (conv2d_nchw_1[6] + (pad_temp.shared_1[186]*kernel.shared_1[((threadIdx.x*72) + 60)]))
+            conv2d_nchw_1[7] = (conv2d_nchw_1[7] + (pad_temp.shared_1[180]*kernel.shared_1[((threadIdx.x*72) + 4668)]))
+            conv2d_nchw_1[8] = (conv2d_nchw_1[8] + (pad_temp.shared_1[181]*kernel.shared_1[((threadIdx.x*72) + 4668)]))
+            conv2d_nchw_1[9] = (conv2d_nchw_1[9] + (pad_temp.shared_1[182]*kernel.shared_1[((threadIdx.x*72) + 4668)]))
+            conv2d_nchw_1[10] = (conv2d_nchw_1[10] + (pad_temp.shared_1[183]*kernel.shared_1[((threadIdx.x*72) + 4668)]))
+            conv2d_nchw_1[11] = (conv2d_nchw_1[11] + (pad_temp.shared_1[184]*kernel.shared_1[((threadIdx.x*72) + 4668)]))
+            conv2d_nchw_1[12] = (conv2d_nchw_1[12] + (pad_temp.shared_1[185]*kernel.shared_1[((threadIdx.x*72) + 4668)]))
+            conv2d_nchw_1[13] = (conv2d_nchw_1[13] + (pad_temp.shared_1[186]*kernel.shared_1[((threadIdx.x*72) + 4668)]))
+            conv2d_nchw_1[0] = (conv2d_nchw_1[0] + (pad_temp.shared_1[181]*kernel.shared_1[((threadIdx.x*72) + 61)]))
+            conv2d_nchw_1[1] = (conv2d_nchw_1[1] + (pad_temp.shared_1[182]*kernel.shared_1[((threadIdx.x*72) + 61)]))
+            conv2d_nchw_1[2] = (conv2d_nchw_1[2] + (pad_temp.shared_1[183]*kernel.shared_1[((threadIdx.x*72) + 61)]))
+            conv2d_nchw_1[3] = (conv2d_nchw_1[3] + (pad_temp.shared_1[184]*kernel.shared_1[((threadIdx.x*72) + 61)]))
+            conv2d_nchw_1[4] = (conv2d_nchw_1[4] + (pad_temp.shared_1[185]*kernel.shared_1[((threadIdx.x*72) + 61)]))
+            conv2d_nchw_1[5] = (conv2d_nchw_1[5] + (pad_temp.shared_1[186]*kernel.shared_1[((threadIdx.x*72) + 61)]))
+            conv2d_nchw_1[6] = (conv2d_nchw_1[6] + (pad_temp.shared_1[187]*kernel.shared_1[((threadIdx.x*72) + 61)]))
+            conv2d_nchw_1[7] = (conv2d_nchw_1[7] + (pad_temp.shared_1[181]*kernel.shared_1[((threadIdx.x*72) + 4669)]))
+            conv2d_nchw_1[8] = (conv2d_nchw_1[8] + (pad_temp.shared_1[182]*kernel.shared_1[((threadIdx.x*72) + 4669)]))
+            conv2d_nchw_1[9] = (conv2d_nchw_1[9] + (pad_temp.shared_1[183]*kernel.shared_1[((threadIdx.x*72) + 4669)]))
+            conv2d_nchw_1[10] = (conv2d_nchw_1[10] + (pad_temp.shared_1[184]*kernel.shared_1[((threadIdx.x*72) + 4669)]))
+            conv2d_nchw_1[11] = (conv2d_nchw_1[11] + (pad_temp.shared_1[185]*kernel.shared_1[((threadIdx.x*72) + 4669)]))
+            conv2d_nchw_1[12] = (conv2d_nchw_1[12] + (pad_temp.shared_1[186]*kernel.shared_1[((threadIdx.x*72) + 4669)]))
+            conv2d_nchw_1[13] = (conv2d_nchw_1[13] + (pad_temp.shared_1[187]*kernel.shared_1[((threadIdx.x*72) + 4669)]))
+            conv2d_nchw_1[0] = (conv2d_nchw_1[0] + (pad_temp.shared_1[182]*kernel.shared_1[((threadIdx.x*72) + 62)]))
+            conv2d_nchw_1[1] = (conv2d_nchw_1[1] + (pad_temp.shared_1[183]*kernel.shared_1[((threadIdx.x*72) + 62)]))
+            conv2d_nchw_1[2] = (conv2d_nchw_1[2] + (pad_temp.shared_1[184]*kernel.shared_1[((threadIdx.x*72) + 62)]))
+            conv2d_nchw_1[3] = (conv2d_nchw_1[3] + (pad_temp.shared_1[185]*kernel.shared_1[((threadIdx.x*72) + 62)]))
+            conv2d_nchw_1[4] = (conv2d_nchw_1[4] + (pad_temp.shared_1[186]*kernel.shared_1[((threadIdx.x*72) + 62)]))
+            conv2d_nchw_1[5] = (conv2d_nchw_1[5] + (pad_temp.shared_1[187]*kernel.shared_1[((threadIdx.x*72) + 62)]))
+            conv2d_nchw_1[6] = (conv2d_nchw_1[6] + (pad_temp.shared_1[188]*kernel.shared_1[((threadIdx.x*72) + 62)]))
+            conv2d_nchw_1[7] = (conv2d_nchw_1[7] + (pad_temp.shared_1[182]*kernel.shared_1[((threadIdx.x*72) + 4670)]))
+            conv2d_nchw_1[8] = (conv2d_nchw_1[8] + (pad_temp.shared_1[183]*kernel.shared_1[((threadIdx.x*72) + 4670)]))
+            conv2d_nchw_1[9] = (conv2d_nchw_1[9] + (pad_temp.shared_1[184]*kernel.shared_1[((threadIdx.x*72) + 4670)]))
+            conv2d_nchw_1[10] = (conv2d_nchw_1[10] + (pad_temp.shared_1[185]*kernel.shared_1[((threadIdx.x*72) + 4670)]))
+            conv2d_nchw_1[11] = (conv2d_nchw_1[11] + (pad_temp.shared_1[186]*kernel.shared_1[((threadIdx.x*72) + 4670)]))
+            conv2d_nchw_1[12] = (conv2d_nchw_1[12] + (pad_temp.shared_1[187]*kernel.shared_1[((threadIdx.x*72) + 4670)]))
+            conv2d_nchw_1[13] = (conv2d_nchw_1[13] + (pad_temp.shared_1[188]*kernel.shared_1[((threadIdx.x*72) + 4670)]))
+            conv2d_nchw_1[0] = (conv2d_nchw_1[0] + (pad_temp.shared_1[189]*kernel.shared_1[((threadIdx.x*72) + 63)]))
+            conv2d_nchw_1[1] = (conv2d_nchw_1[1] + (pad_temp.shared_1[190]*kernel.shared_1[((threadIdx.x*72) + 63)]))
+            conv2d_nchw_1[2] = (conv2d_nchw_1[2] + (pad_temp.shared_1[191]*kernel.shared_1[((threadIdx.x*72) + 63)]))
+            conv2d_nchw_1[3] = (conv2d_nchw_1[3] + (pad_temp.shared_1[192]*kernel.shared_1[((threadIdx.x*72) + 63)]))
+            conv2d_nchw_1[4] = (conv2d_nchw_1[4] + (pad_temp.shared_1[193]*kernel.shared_1[((threadIdx.x*72) + 63)]))
+            conv2d_nchw_1[5] = (conv2d_nchw_1[5] + (pad_temp.shared_1[194]*kernel.shared_1[((threadIdx.x*72) + 63)]))
+            conv2d_nchw_1[6] = (conv2d_nchw_1[6] + (pad_temp.shared_1[195]*kernel.shared_1[((threadIdx.x*72) + 63)]))
+            conv2d_nchw_1[7] = (conv2d_nchw_1[7] + (pad_temp.shared_1[189]*kernel.shared_1[((threadIdx.x*72) + 4671)]))
+            conv2d_nchw_1[8] = (conv2d_nchw_1[8] + (pad_temp.shared_1[190]*kernel.shared_1[((threadIdx.x*72) + 4671)]))
+            conv2d_nchw_1[9] = (conv2d_nchw_1[9] + (pad_temp.shared_1[191]*kernel.shared_1[((threadIdx.x*72) + 4671)]))
+            conv2d_nchw_1[10] = (conv2d_nchw_1[10] + (pad_temp.shared_1[192]*kernel.shared_1[((threadIdx.x*72) + 4671)]))
+            conv2d_nchw_1[11] = (conv2d_nchw_1[11] + (pad_temp.shared_1[193]*kernel.shared_1[((threadIdx.x*72) + 4671)]))
+            conv2d_nchw_1[12] = (conv2d_nchw_1[12] + (pad_temp.shared_1[194]*kernel.shared_1[((threadIdx.x*72) + 4671)]))
+            conv2d_nchw_1[13] = (conv2d_nchw_1[13] + (pad_temp.shared_1[195]*kernel.shared_1[((threadIdx.x*72) + 4671)]))
+            conv2d_nchw_1[0] = (conv2d_nchw_1[0] + (pad_temp.shared_1[190]*kernel.shared_1[((threadIdx.x*72) + 64)]))
+            conv2d_nchw_1[1] = (conv2d_nchw_1[1] + (pad_temp.shared_1[191]*kernel.shared_1[((threadIdx.x*72) + 64)]))
+            conv2d_nchw_1[2] = (conv2d_nchw_1[2] + (pad_temp.shared_1[192]*kernel.shared_1[((threadIdx.x*72) + 64)]))
+            conv2d_nchw_1[3] = (conv2d_nchw_1[3] + (pad_temp.shared_1[193]*kernel.shared_1[((threadIdx.x*72) + 64)]))
+            conv2d_nchw_1[4] = (conv2d_nchw_1[4] + (pad_temp.shared_1[194]*kernel.shared_1[((threadIdx.x*72) + 64)]))
+            conv2d_nchw_1[5] = (conv2d_nchw_1[5] + (pad_temp.shared_1[195]*kernel.shared_1[((threadIdx.x*72) + 64)]))
+            conv2d_nchw_1[6] = (conv2d_nchw_1[6] + (pad_temp.shared_1[196]*kernel.shared_1[((threadIdx.x*72) + 64)]))
+            conv2d_nchw_1[7] = (conv2d_nchw_1[7] + (pad_temp.shared_1[190]*kernel.shared_1[((threadIdx.x*72) + 4672)]))
+            conv2d_nchw_1[8] = (conv2d_nchw_1[8] + (pad_temp.shared_1[191]*kernel.shared_1[((threadIdx.x*72) + 4672)]))
+            conv2d_nchw_1[9] = (conv2d_nchw_1[9] + (pad_temp.shared_1[192]*kernel.shared_1[((threadIdx.x*72) + 4672)]))
+            conv2d_nchw_1[10] = (conv2d_nchw_1[10] + (pad_temp.shared_1[193]*kernel.shared_1[((threadIdx.x*72) + 4672)]))
+            conv2d_nchw_1[11] = (conv2d_nchw_1[11] + (pad_temp.shared_1[194]*kernel.shared_1[((threadIdx.x*72) + 4672)]))
+            conv2d_nchw_1[12] = (conv2d_nchw_1[12] + (pad_temp.shared_1[195]*kernel.shared_1[((threadIdx.x*72) + 4672)]))
+            conv2d_nchw_1[13] = (conv2d_nchw_1[13] + (pad_temp.shared_1[196]*kernel.shared_1[((threadIdx.x*72) + 4672)]))
+            conv2d_nchw_1[0] = (conv2d_nchw_1[0] + (pad_temp.shared_1[191]*kernel.shared_1[((threadIdx.x*72) + 65)]))
+            conv2d_nchw_1[1] = (conv2d_nchw_1[1] + (pad_temp.shared_1[192]*kernel.shared_1[((threadIdx.x*72) + 65)]))
+            conv2d_nchw_1[2] = (conv2d_nchw_1[2] + (pad_temp.shared_1[193]*kernel.shared_1[((threadIdx.x*72) + 65)]))
+            conv2d_nchw_1[3] = (conv2d_nchw_1[3] + (pad_temp.shared_1[194]*kernel.shared_1[((threadIdx.x*72) + 65)]))
+            conv2d_nchw_1[4] = (conv2d_nchw_1[4] + (pad_temp.shared_1[195]*kernel.shared_1[((threadIdx.x*72) + 65)]))
+            conv2d_nchw_1[5] = (conv2d_nchw_1[5] + (pad_temp.shared_1[196]*kernel.shared_1[((threadIdx.x*72) + 65)]))
+            conv2d_nchw_1[6] = (conv2d_nchw_1[6] + (pad_temp.shared_1[197]*kernel.shared_1[((threadIdx.x*72) + 65)]))
+            conv2d_nchw_1[7] = (conv2d_nchw_1[7] + (pad_temp.shared_1[191]*kernel.shared_1[((threadIdx.x*72) + 4673)]))
+            conv2d_nchw_1[8] = (conv2d_nchw_1[8] + (pad_temp.shared_1[192]*kernel.shared_1[((threadIdx.x*72) + 4673)]))
+            conv2d_nchw_1[9] = (conv2d_nchw_1[9] + (pad_temp.shared_1[193]*kernel.shared_1[((threadIdx.x*72) + 4673)]))
+            conv2d_nchw_1[10] = (conv2d_nchw_1[10] + (pad_temp.shared_1[194]*kernel.shared_1[((threadIdx.x*72) + 4673)]))
+            conv2d_nchw_1[11] = (conv2d_nchw_1[11] + (pad_temp.shared_1[195]*kernel.shared_1[((threadIdx.x*72) + 4673)]))
+            conv2d_nchw_1[12] = (conv2d_nchw_1[12] + (pad_temp.shared_1[196]*kernel.shared_1[((threadIdx.x*72) + 4673)]))
+            conv2d_nchw_1[13] = (conv2d_nchw_1[13] + (pad_temp.shared_1[197]*kernel.shared_1[((threadIdx.x*72) + 4673)]))
+            conv2d_nchw_1[0] = (conv2d_nchw_1[0] + (pad_temp.shared_1[198]*kernel.shared_1[((threadIdx.x*72) + 66)]))
+            conv2d_nchw_1[1] = (conv2d_nchw_1[1] + (pad_temp.shared_1[199]*kernel.shared_1[((threadIdx.x*72) + 66)]))
+            conv2d_nchw_1[2] = (conv2d_nchw_1[2] + (pad_temp.shared_1[200]*kernel.shared_1[((threadIdx.x*72) + 66)]))
+            conv2d_nchw_1[3] = (conv2d_nchw_1[3] + (pad_temp.shared_1[201]*kernel.shared_1[((threadIdx.x*72) + 66)]))
+            conv2d_nchw_1[4] = (conv2d_nchw_1[4] + (pad_temp.shared_1[202]*kernel.shared_1[((threadIdx.x*72) + 66)]))
+            conv2d_nchw_1[5] = (conv2d_nchw_1[5] + (pad_temp.shared_1[203]*kernel.shared_1[((threadIdx.x*72) + 66)]))
+            conv2d_nchw_1[6] = (conv2d_nchw_1[6] + (pad_temp.shared_1[204]*kernel.shared_1[((threadIdx.x*72) + 66)]))
+            conv2d_nchw_1[7] = (conv2d_nchw_1[7] + (pad_temp.shared_1[198]*kernel.shared_1[((threadIdx.x*72) + 4674)]))
+            conv2d_nchw_1[8] = (conv2d_nchw_1[8] + (pad_temp.shared_1[199]*kernel.shared_1[((threadIdx.x*72) + 4674)]))
+            conv2d_nchw_1[9] = (conv2d_nchw_1[9] + (pad_temp.shared_1[200]*kernel.shared_1[((threadIdx.x*72) + 4674)]))
+            conv2d_nchw_1[10] = (conv2d_nchw_1[10] + (pad_temp.shared_1[201]*kernel.shared_1[((threadIdx.x*72) + 4674)]))
+            conv2d_nchw_1[11] = (conv2d_nchw_1[11] + (pad_temp.shared_1[202]*kernel.shared_1[((threadIdx.x*72) + 4674)]))
+            conv2d_nchw_1[12] = (conv2d_nchw_1[12] + (pad_temp.shared_1[203]*kernel.shared_1[((threadIdx.x*72) + 4674)]))
+            conv2d_nchw_1[13] = (conv2d_nchw_1[13] + (pad_temp.shared_1[204]*kernel.shared_1[((threadIdx.x*72) + 4674)]))
+            conv2d_nchw_1[0] = (conv2d_nchw_1[0] + (pad_temp.shared_1[199]*kernel.shared_1[((threadIdx.x*72) + 67)]))
+            conv2d_nchw_1[1] = (conv2d_nchw_1[1] + (pad_temp.shared_1[200]*kernel.shared_1[((threadIdx.x*72) + 67)]))
+            conv2d_nchw_1[2] = (conv2d_nchw_1[2] + (pad_temp.shared_1[201]*kernel.shared_1[((threadIdx.x*72) + 67)]))
+            conv2d_nchw_1[3] = (conv2d_nchw_1[3] + (pad_temp.shared_1[202]*kernel.shared_1[((threadIdx.x*72) + 67)]))
+            conv2d_nchw_1[4] = (conv2d_nchw_1[4] + (pad_temp.shared_1[203]*kernel.shared_1[((threadIdx.x*72) + 67)]))
+            conv2d_nchw_1[5] = (conv2d_nchw_1[5] + (pad_temp.shared_1[204]*kernel.shared_1[((threadIdx.x*72) + 67)]))
+            conv2d_nchw_1[6] = (conv2d_nchw_1[6] + (pad_temp.shared_1[205]*kernel.shared_1[((threadIdx.x*72) + 67)]))
+            conv2d_nchw_1[7] = (conv2d_nchw_1[7] + (pad_temp.shared_1[199]*kernel.shared_1[((threadIdx.x*72) + 4675)]))
+            conv2d_nchw_1[8] = (conv2d_nchw_1[8] + (pad_temp.shared_1[200]*kernel.shared_1[((threadIdx.x*72) + 4675)]))
+            conv2d_nchw_1[9] = (conv2d_nchw_1[9] + (pad_temp.shared_1[201]*kernel.shared_1[((threadIdx.x*72) + 4675)]))
+            conv2d_nchw_1[10] = (conv2d_nchw_1[10] + (pad_temp.shared_1[202]*kernel.shared_1[((threadIdx.x*72) + 4675)]))
+            conv2d_nchw_1[11] = (conv2d_nchw_1[11] + (pad_temp.shared_1[203]*kernel.shared_1[((threadIdx.x*72) + 4675)]))
+            conv2d_nchw_1[12] = (conv2d_nchw_1[12] + (pad_temp.shared_1[204]*kernel.shared_1[((threadIdx.x*72) + 4675)]))
+            conv2d_nchw_1[13] = (conv2d_nchw_1[13] + (pad_temp.shared_1[205]*kernel.shared_1[((threadIdx.x*72) + 4675)]))
+            conv2d_nchw_1[0] = (conv2d_nchw_1[0] + (pad_temp.shared_1[200]*kernel.shared_1[((threadIdx.x*72) + 68)]))
+            conv2d_nchw_1[1] = (conv2d_nchw_1[1] + (pad_temp.shared_1[201]*kernel.shared_1[((threadIdx.x*72) + 68)]))
+            conv2d_nchw_1[2] = (conv2d_nchw_1[2] + (pad_temp.shared_1[202]*kernel.shared_1[((threadIdx.x*72) + 68)]))
+            conv2d_nchw_1[3] = (conv2d_nchw_1[3] + (pad_temp.shared_1[203]*kernel.shared_1[((threadIdx.x*72) + 68)]))
+            conv2d_nchw_1[4] = (conv2d_nchw_1[4] + (pad_temp.shared_1[204]*kernel.shared_1[((threadIdx.x*72) + 68)]))
+            conv2d_nchw_1[5] = (conv2d_nchw_1[5] + (pad_temp.shared_1[205]*kernel.shared_1[((threadIdx.x*72) + 68)]))
+            conv2d_nchw_1[6] = (conv2d_nchw_1[6] + (pad_temp.shared_1[206]*kernel.shared_1[((threadIdx.x*72) + 68)]))
+            conv2d_nchw_1[7] = (conv2d_nchw_1[7] + (pad_temp.shared_1[200]*kernel.shared_1[((threadIdx.x*72) + 4676)]))
+            conv2d_nchw_1[8] = (conv2d_nchw_1[8] + (pad_temp.shared_1[201]*kernel.shared_1[((threadIdx.x*72) + 4676)]))
+            conv2d_nchw_1[9] = (conv2d_nchw_1[9] + (pad_temp.shared_1[202]*kernel.shared_1[((threadIdx.x*72) + 4676)]))
+            conv2d_nchw_1[10] = (conv2d_nchw_1[10] + (pad_temp.shared_1[203]*kernel.shared_1[((threadIdx.x*72) + 4676)]))
+            conv2d_nchw_1[11] = (conv2d_nchw_1[11] + (pad_temp.shared_1[204]*kernel.shared_1[((threadIdx.x*72) + 4676)]))
+            conv2d_nchw_1[12] = (conv2d_nchw_1[12] + (pad_temp.shared_1[205]*kernel.shared_1[((threadIdx.x*72) + 4676)]))
+            conv2d_nchw_1[13] = (conv2d_nchw_1[13] + (pad_temp.shared_1[206]*kernel.shared_1[((threadIdx.x*72) + 4676)]))
+            conv2d_nchw_1[0] = (conv2d_nchw_1[0] + (pad_temp.shared_1[207]*kernel.shared_1[((threadIdx.x*72) + 69)]))
+            conv2d_nchw_1[1] = (conv2d_nchw_1[1] + (pad_temp.shared_1[208]*kernel.shared_1[((threadIdx.x*72) + 69)]))
+            conv2d_nchw_1[2] = (conv2d_nchw_1[2] + (pad_temp.shared_1[209]*kernel.shared_1[((threadIdx.x*72) + 69)]))
+            conv2d_nchw_1[3] = (conv2d_nchw_1[3] + (pad_temp.shared_1[210]*kernel.shared_1[((threadIdx.x*72) + 69)]))
+            conv2d_nchw_1[4] = (conv2d_nchw_1[4] + (pad_temp.shared_1[211]*kernel.shared_1[((threadIdx.x*72) + 69)]))
+            conv2d_nchw_1[5] = (conv2d_nchw_1[5] + (pad_temp.shared_1[212]*kernel.shared_1[((threadIdx.x*72) + 69)]))
+            conv2d_nchw_1[6] = (conv2d_nchw_1[6] + (pad_temp.shared_1[213]*kernel.shared_1[((threadIdx.x*72) + 69)]))
+            conv2d_nchw_1[7] = (conv2d_nchw_1[7] + (pad_temp.shared_1[207]*kernel.shared_1[((threadIdx.x*72) + 4677)]))
+            conv2d_nchw_1[8] = (conv2d_nchw_1[8] + (pad_temp.shared_1[208]*kernel.shared_1[((threadIdx.x*72) + 4677)]))
+            conv2d_nchw_1[9] = (conv2d_nchw_1[9] + (pad_temp.shared_1[209]*kernel.shared_1[((threadIdx.x*72) + 4677)]))
+            conv2d_nchw_1[10] = (conv2d_nchw_1[10] + (pad_temp.shared_1[210]*kernel.shared_1[((threadIdx.x*72) + 4677)]))
+            conv2d_nchw_1[11] = (conv2d_nchw_1[11] + (pad_temp.shared_1[211]*kernel.shared_1[((threadIdx.x*72) + 4677)]))
+            conv2d_nchw_1[12] = (conv2d_nchw_1[12] + (pad_temp.shared_1[212]*kernel.shared_1[((threadIdx.x*72) + 4677)]))
+            conv2d_nchw_1[13] = (conv2d_nchw_1[13] + (pad_temp.shared_1[213]*kernel.shared_1[((threadIdx.x*72) + 4677)]))
+            conv2d_nchw_1[0] = (conv2d_nchw_1[0] + (pad_temp.shared_1[208]*kernel.shared_1[((threadIdx.x*72) + 70)]))
+            conv2d_nchw_1[1] = (conv2d_nchw_1[1] + (pad_temp.shared_1[209]*kernel.shared_1[((threadIdx.x*72) + 70)]))
+            conv2d_nchw_1[2] = (conv2d_nchw_1[2] + (pad_temp.shared_1[210]*kernel.shared_1[((threadIdx.x*72) + 70)]))
+            conv2d_nchw_1[3] = (conv2d_nchw_1[3] + (pad_temp.shared_1[211]*kernel.shared_1[((threadIdx.x*72) + 70)]))
+            conv2d_nchw_1[4] = (conv2d_nchw_1[4] + (pad_temp.shared_1[212]*kernel.shared_1[((threadIdx.x*72) + 70)]))
+            conv2d_nchw_1[5] = (conv2d_nchw_1[5] + (pad_temp.shared_1[213]*kernel.shared_1[((threadIdx.x*72) + 70)]))
+            conv2d_nchw_1[6] = (conv2d_nchw_1[6] + (pad_temp.shared_1[214]*kernel.shared_1[((threadIdx.x*72) + 70)]))
+            conv2d_nchw_1[7] = (conv2d_nchw_1[7] + (pad_temp.shared_1[208]*kernel.shared_1[((threadIdx.x*72) + 4678)]))
+            conv2d_nchw_1[8] = (conv2d_nchw_1[8] + (pad_temp.shared_1[209]*kernel.shared_1[((threadIdx.x*72) + 4678)]))
+            conv2d_nchw_1[9] = (conv2d_nchw_1[9] + (pad_temp.shared_1[210]*kernel.shared_1[((threadIdx.x*72) + 4678)]))
+            conv2d_nchw_1[10] = (conv2d_nchw_1[10] + (pad_temp.shared_1[211]*kernel.shared_1[((threadIdx.x*72) + 4678)]))
+            conv2d_nchw_1[11] = (conv2d_nchw_1[11] + (pad_temp.shared_1[212]*kernel.shared_1[((threadIdx.x*72) + 4678)]))
+            conv2d_nchw_1[12] = (conv2d_nchw_1[12] + (pad_temp.shared_1[213]*kernel.shared_1[((threadIdx.x*72) + 4678)]))
+            conv2d_nchw_1[13] = (conv2d_nchw_1[13] + (pad_temp.shared_1[214]*kernel.shared_1[((threadIdx.x*72) + 4678)]))
+            conv2d_nchw_1[0] = (conv2d_nchw_1[0] + (pad_temp.shared_1[209]*kernel.shared_1[((threadIdx.x*72) + 71)]))
+            conv2d_nchw_1[1] = (conv2d_nchw_1[1] + (pad_temp.shared_1[210]*kernel.shared_1[((threadIdx.x*72) + 71)]))
+            conv2d_nchw_1[2] = (conv2d_nchw_1[2] + (pad_temp.shared_1[211]*kernel.shared_1[((threadIdx.x*72) + 71)]))
+            conv2d_nchw_1[3] = (conv2d_nchw_1[3] + (pad_temp.shared_1[212]*kernel.shared_1[((threadIdx.x*72) + 71)]))
+            conv2d_nchw_1[4] = (conv2d_nchw_1[4] + (pad_temp.shared_1[213]*kernel.shared_1[((threadIdx.x*72) + 71)]))
+            conv2d_nchw_1[5] = (conv2d_nchw_1[5] + (pad_temp.shared_1[214]*kernel.shared_1[((threadIdx.x*72) + 71)]))
+            conv2d_nchw_1[6] = (conv2d_nchw_1[6] + (pad_temp.shared_1[215]*kernel.shared_1[((threadIdx.x*72) + 71)]))
+            conv2d_nchw_1[7] = (conv2d_nchw_1[7] + (pad_temp.shared_1[209]*kernel.shared_1[((threadIdx.x*72) + 4679)]))
+            conv2d_nchw_1[8] = (conv2d_nchw_1[8] + (pad_temp.shared_1[210]*kernel.shared_1[((threadIdx.x*72) + 4679)]))
+            conv2d_nchw_1[9] = (conv2d_nchw_1[9] + (pad_temp.shared_1[211]*kernel.shared_1[((threadIdx.x*72) + 4679)]))
+            conv2d_nchw_1[10] = (conv2d_nchw_1[10] + (pad_temp.shared_1[212]*kernel.shared_1[((threadIdx.x*72) + 4679)]))
+            conv2d_nchw_1[11] = (conv2d_nchw_1[11] + (pad_temp.shared_1[213]*kernel.shared_1[((threadIdx.x*72) + 4679)]))
+            conv2d_nchw_1[12] = (conv2d_nchw_1[12] + (pad_temp.shared_1[214]*kernel.shared_1[((threadIdx.x*72) + 4679)]))
+            conv2d_nchw_1[13] = (conv2d_nchw_1[13] + (pad_temp.shared_1[215]*kernel.shared_1[((threadIdx.x*72) + 4679)]))
           }
         }
-        for (i1.inner: int32, 0, 2) {
-          for (i3.inner: int32, 0, 7) {
-            compute_3: Buffer(compute_2, float32, [25088], [])[(((((floordiv(blockIdx.x, 7)*6272) + (threadIdx.x*98)) + (i1.inner*49)) + (floormod(blockIdx.x, 7)*7)) + i3.inner)] = max((conv2d_nchw_1[((i1.inner*7) + i3.inner)] + bias_3: Buffer(bias_2, float32, [512], [])[(((floordiv(blockIdx.x, 7)*128) + (threadIdx.x*2)) + i1.inner)]), 0f32)
-          }
-        }
+        compute_3: Buffer(compute_2, float32, [25088], [])[(((floordiv(blockIdx.x, 7)*6272) + (threadIdx.x*49)) + (floormod(blockIdx.x, 7)*7))] = max((conv2d_nchw_1[0] + bias_3: Buffer(bias_2, float32, [512], [])[((floordiv(blockIdx.x, 7)*128) + threadIdx.x)]), 0f32)
+        compute_3[((((floordiv(blockIdx.x, 7)*6272) + (threadIdx.x*49)) + (floormod(blockIdx.x, 7)*7)) + 1)] = max((conv2d_nchw_1[1] + bias_3[((floordiv(blockIdx.x, 7)*128) + threadIdx.x)]), 0f32)
+        compute_3[((((floordiv(blockIdx.x, 7)*6272) + (threadIdx.x*49)) + (floormod(blockIdx.x, 7)*7)) + 2)] = max((conv2d_nchw_1[2] + bias_3[((floordiv(blockIdx.x, 7)*128) + threadIdx.x)]), 0f32)
+        compute_3[((((floordiv(blockIdx.x, 7)*6272) + (threadIdx.x*49)) + (floormod(blockIdx.x, 7)*7)) + 3)] = max((conv2d_nchw_1[3] + bias_3[((floordiv(blockIdx.x, 7)*128) + threadIdx.x)]), 0f32)
+        compute_3[((((floordiv(blockIdx.x, 7)*6272) + (threadIdx.x*49)) + (floormod(blockIdx.x, 7)*7)) + 4)] = max((conv2d_nchw_1[4] + bias_3[((floordiv(blockIdx.x, 7)*128) + threadIdx.x)]), 0f32)
+        compute_3[((((floordiv(blockIdx.x, 7)*6272) + (threadIdx.x*49)) + (floormod(blockIdx.x, 7)*7)) + 5)] = max((conv2d_nchw_1[5] + bias_3[((floordiv(blockIdx.x, 7)*128) + threadIdx.x)]), 0f32)
+        compute_3[((((floordiv(blockIdx.x, 7)*6272) + (threadIdx.x*49)) + (floormod(blockIdx.x, 7)*7)) + 6)] = max((conv2d_nchw_1[6] + bias_3[((floordiv(blockIdx.x, 7)*128) + threadIdx.x)]), 0f32)
+        compute_3[((((floordiv(blockIdx.x, 7)*6272) + (threadIdx.x*49)) + (floormod(blockIdx.x, 7)*7)) + 3136)] = max((conv2d_nchw_1[7] + bias_3[(((floordiv(blockIdx.x, 7)*128) + threadIdx.x) + 64)]), 0f32)
+        compute_3[((((floordiv(blockIdx.x, 7)*6272) + (threadIdx.x*49)) + (floormod(blockIdx.x, 7)*7)) + 3137)] = max((conv2d_nchw_1[8] + bias_3[(((floordiv(blockIdx.x, 7)*128) + threadIdx.x) + 64)]), 0f32)
+        compute_3[((((floordiv(blockIdx.x, 7)*6272) + (threadIdx.x*49)) + (floormod(blockIdx.x, 7)*7)) + 3138)] = max((conv2d_nchw_1[9] + bias_3[(((floordiv(blockIdx.x, 7)*128) + threadIdx.x) + 64)]), 0f32)
+        compute_3[((((floordiv(blockIdx.x, 7)*6272) + (threadIdx.x*49)) + (floormod(blockIdx.x, 7)*7)) + 3139)] = max((conv2d_nchw_1[10] + bias_3[(((floordiv(blockIdx.x, 7)*128) + threadIdx.x) + 64)]), 0f32)
+        compute_3[((((floordiv(blockIdx.x, 7)*6272) + (threadIdx.x*49)) + (floormod(blockIdx.x, 7)*7)) + 3140)] = max((conv2d_nchw_1[11] + bias_3[(((floordiv(blockIdx.x, 7)*128) + threadIdx.x) + 64)]), 0f32)
+        compute_3[((((floordiv(blockIdx.x, 7)*6272) + (threadIdx.x*49)) + (floormod(blockIdx.x, 7)*7)) + 3141)] = max((conv2d_nchw_1[12] + bias_3[(((floordiv(blockIdx.x, 7)*128) + threadIdx.x) + 64)]), 0f32)
+        compute_3[((((floordiv(blockIdx.x, 7)*6272) + (threadIdx.x*49)) + (floormod(blockIdx.x, 7)*7)) + 3142)] = max((conv2d_nchw_1[13] + bias_3[(((floordiv(blockIdx.x, 7)*128) + threadIdx.x) + 64)]), 0f32)
       }
     }
 
@@ -781,7 +1641,7 @@ We build the binary and check its correctness and performance.
 
  .. code-block:: none
 
-    Execution time of this operator: 0.360 ms
+    Execution time of this operator: 0.271 ms
 
 
 
@@ -830,36 +1690,36 @@ They can be used for debugging and learning the behavior of the auto-scheduler.
     conv2d_nchw_nn_o_o_o_i, conv2d_nchw_nn_o_o_i = s[conv2d_nchw].split(conv2d_nchw_nn_o_o_i, factor=1)
     conv2d_nchw_nn_o_o_o_o, conv2d_nchw_nn_o_o_o_i = s[conv2d_nchw].split(conv2d_nchw_nn_o_o_o_i, factor=1)
     conv2d_nchw_ff_o_i, conv2d_nchw_ff_i = s[conv2d_nchw].split(conv2d_nchw_ff, factor=1)
-    conv2d_nchw_ff_o_o_i, conv2d_nchw_ff_o_i = s[conv2d_nchw].split(conv2d_nchw_ff_o_i, factor=2)
+    conv2d_nchw_ff_o_o_i, conv2d_nchw_ff_o_i = s[conv2d_nchw].split(conv2d_nchw_ff_o_i, factor=1)
     conv2d_nchw_ff_o_o_o_i, conv2d_nchw_ff_o_o_i = s[conv2d_nchw].split(conv2d_nchw_ff_o_o_i, factor=64)
-    conv2d_nchw_ff_o_o_o_o, conv2d_nchw_ff_o_o_o_i = s[conv2d_nchw].split(conv2d_nchw_ff_o_o_o_i, factor=1)
+    conv2d_nchw_ff_o_o_o_o, conv2d_nchw_ff_o_o_o_i = s[conv2d_nchw].split(conv2d_nchw_ff_o_o_o_i, factor=2)
     conv2d_nchw_yy_o_i, conv2d_nchw_yy_i = s[conv2d_nchw].split(conv2d_nchw_yy, factor=1)
     conv2d_nchw_yy_o_o_i, conv2d_nchw_yy_o_i = s[conv2d_nchw].split(conv2d_nchw_yy_o_i, factor=1)
     conv2d_nchw_yy_o_o_o_i, conv2d_nchw_yy_o_o_i = s[conv2d_nchw].split(conv2d_nchw_yy_o_o_i, factor=1)
     conv2d_nchw_yy_o_o_o_o, conv2d_nchw_yy_o_o_o_i = s[conv2d_nchw].split(conv2d_nchw_yy_o_o_o_i, factor=1)
     conv2d_nchw_xx_o_i, conv2d_nchw_xx_i = s[conv2d_nchw].split(conv2d_nchw_xx, factor=1)
-    conv2d_nchw_xx_o_o_i, conv2d_nchw_xx_o_i = s[conv2d_nchw].split(conv2d_nchw_xx_o_i, factor=7)
+    conv2d_nchw_xx_o_o_i, conv2d_nchw_xx_o_i = s[conv2d_nchw].split(conv2d_nchw_xx_o_i, factor=1)
     conv2d_nchw_xx_o_o_o_i, conv2d_nchw_xx_o_o_i = s[conv2d_nchw].split(conv2d_nchw_xx_o_o_i, factor=1)
-    conv2d_nchw_xx_o_o_o_o, conv2d_nchw_xx_o_o_o_i = s[conv2d_nchw].split(conv2d_nchw_xx_o_o_o_i, factor=1)
-    conv2d_nchw_rc_o_i, conv2d_nchw_rc_i = s[conv2d_nchw].split(conv2d_nchw_rc, factor=2)
-    conv2d_nchw_rc_o_o, conv2d_nchw_rc_o_i = s[conv2d_nchw].split(conv2d_nchw_rc_o_i, factor=4)
-    conv2d_nchw_ry_o_i, conv2d_nchw_ry_i = s[conv2d_nchw].split(conv2d_nchw_ry, factor=1)
+    conv2d_nchw_xx_o_o_o_o, conv2d_nchw_xx_o_o_o_i = s[conv2d_nchw].split(conv2d_nchw_xx_o_o_o_i, factor=7)
+    conv2d_nchw_rc_o_i, conv2d_nchw_rc_i = s[conv2d_nchw].split(conv2d_nchw_rc, factor=1)
+    conv2d_nchw_rc_o_o, conv2d_nchw_rc_o_i = s[conv2d_nchw].split(conv2d_nchw_rc_o_i, factor=8)
+    conv2d_nchw_ry_o_i, conv2d_nchw_ry_i = s[conv2d_nchw].split(conv2d_nchw_ry, factor=3)
     conv2d_nchw_ry_o_o, conv2d_nchw_ry_o_i = s[conv2d_nchw].split(conv2d_nchw_ry_o_i, factor=1)
-    conv2d_nchw_rx_o_i, conv2d_nchw_rx_i = s[conv2d_nchw].split(conv2d_nchw_rx, factor=1)
-    conv2d_nchw_rx_o_o, conv2d_nchw_rx_o_i = s[conv2d_nchw].split(conv2d_nchw_rx_o_i, factor=3)
+    conv2d_nchw_rx_o_i, conv2d_nchw_rx_i = s[conv2d_nchw].split(conv2d_nchw_rx, factor=3)
+    conv2d_nchw_rx_o_o, conv2d_nchw_rx_o_i = s[conv2d_nchw].split(conv2d_nchw_rx_o_i, factor=1)
     s[conv2d_nchw].reorder(conv2d_nchw_nn_o_o_o_o, conv2d_nchw_ff_o_o_o_o, conv2d_nchw_yy_o_o_o_o, conv2d_nchw_xx_o_o_o_o, conv2d_nchw_nn_o_o_o_i, conv2d_nchw_ff_o_o_o_i, conv2d_nchw_yy_o_o_o_i, conv2d_nchw_xx_o_o_o_i, conv2d_nchw_nn_o_o_i, conv2d_nchw_ff_o_o_i, conv2d_nchw_yy_o_o_i, conv2d_nchw_xx_o_o_i, conv2d_nchw_rc_o_o, conv2d_nchw_ry_o_o, conv2d_nchw_rx_o_o, conv2d_nchw_rc_o_i, conv2d_nchw_ry_o_i, conv2d_nchw_rx_o_i, conv2d_nchw_nn_o_i, conv2d_nchw_ff_o_i, conv2d_nchw_yy_o_i, conv2 [...]
     compute_i0_o_i, compute_i0_i = s[compute].split(compute_i0, factor=1)
     compute_i0_o_o_i, compute_i0_o_i = s[compute].split(compute_i0_o_i, factor=1)
     compute_i0_o_o_o, compute_i0_o_o_i = s[compute].split(compute_i0_o_o_i, factor=1)
-    compute_i1_o_i, compute_i1_i = s[compute].split(compute_i1, factor=2)
+    compute_i1_o_i, compute_i1_i = s[compute].split(compute_i1, factor=1)
     compute_i1_o_o_i, compute_i1_o_i = s[compute].split(compute_i1_o_i, factor=64)
-    compute_i1_o_o_o, compute_i1_o_o_i = s[compute].split(compute_i1_o_o_i, factor=1)
+    compute_i1_o_o_o, compute_i1_o_o_i = s[compute].split(compute_i1_o_o_i, factor=2)
     compute_i2_o_i, compute_i2_i = s[compute].split(compute_i2, factor=1)
     compute_i2_o_o_i, compute_i2_o_i = s[compute].split(compute_i2_o_i, factor=1)
     compute_i2_o_o_o, compute_i2_o_o_i = s[compute].split(compute_i2_o_o_i, factor=1)
-    compute_i3_o_i, compute_i3_i = s[compute].split(compute_i3, factor=7)
+    compute_i3_o_i, compute_i3_i = s[compute].split(compute_i3, factor=1)
     compute_i3_o_o_i, compute_i3_o_i = s[compute].split(compute_i3_o_i, factor=1)
-    compute_i3_o_o_o, compute_i3_o_o_i = s[compute].split(compute_i3_o_o_i, factor=1)
+    compute_i3_o_o_o, compute_i3_o_o_i = s[compute].split(compute_i3_o_o_i, factor=7)
     s[compute].reorder(compute_i0_o_o_o, compute_i1_o_o_o, compute_i2_o_o_o, compute_i3_o_o_o, compute_i0_o_o_i, compute_i1_o_o_i, compute_i2_o_o_i, compute_i3_o_o_i, compute_i0_o_i, compute_i1_o_i, compute_i2_o_i, compute_i3_o_i, compute_i0_i, compute_i1_i, compute_i2_i, compute_i3_i)
     s[conv2d_nchw].compute_at(s[compute], compute_i3_o_i)
     kernel_shared = s.cache_read(kernel, "shared", [conv2d_nchw])
@@ -881,11 +1741,11 @@ They can be used for debugging and learning the behavior of the auto-scheduler.
     kernel_shared_ax0_ax1_fused_ax2_fused_ax3_fused_o_o, kernel_shared_ax0_ax1_fused_ax2_fused_ax3_fused_o_i = s[kernel_shared].split(kernel_shared_ax0_ax1_fused_ax2_fused_ax3_fused_o, factor=64)
     s[kernel_shared].bind(kernel_shared_ax0_ax1_fused_ax2_fused_ax3_fused_o_i, te.thread_axis("threadIdx.x"))
     pad_temp_shared_ax0_ax1_fused_ax2_fused_ax3_fused = s[pad_temp_shared].fuse(pad_temp_shared_ax0, pad_temp_shared_ax1, pad_temp_shared_ax2, pad_temp_shared_ax3)
-    pad_temp_shared_ax0_ax1_fused_ax2_fused_ax3_fused_o, pad_temp_shared_ax0_ax1_fused_ax2_fused_ax3_fused_i = s[pad_temp_shared].split(pad_temp_shared_ax0_ax1_fused_ax2_fused_ax3_fused, factor=4)
+    pad_temp_shared_ax0_ax1_fused_ax2_fused_ax3_fused_o, pad_temp_shared_ax0_ax1_fused_ax2_fused_ax3_fused_i = s[pad_temp_shared].split(pad_temp_shared_ax0_ax1_fused_ax2_fused_ax3_fused, factor=1)
     s[pad_temp_shared].vectorize(pad_temp_shared_ax0_ax1_fused_ax2_fused_ax3_fused_i)
     pad_temp_shared_ax0_ax1_fused_ax2_fused_ax3_fused_o_o, pad_temp_shared_ax0_ax1_fused_ax2_fused_ax3_fused_o_i = s[pad_temp_shared].split(pad_temp_shared_ax0_ax1_fused_ax2_fused_ax3_fused_o, factor=64)
     s[pad_temp_shared].bind(pad_temp_shared_ax0_ax1_fused_ax2_fused_ax3_fused_o_i, te.thread_axis("threadIdx.x"))
-    s[conv2d_nchw].pragma(conv2d_nchw_nn_o_o_o_o, "auto_unroll_max_step", 512)
+    s[conv2d_nchw].pragma(conv2d_nchw_nn_o_o_o_o, "auto_unroll_max_step", 1024)
     s[conv2d_nchw].pragma(conv2d_nchw_nn_o_o_o_o, "unroll_explicit", True)
 
     CUDA source code:
@@ -905,8 +1765,8 @@ They can be used for debugging and learning the behavior of the auto-scheduler.
     #endif
     extern "C" __global__ void __launch_bounds__(64) default_function_kernel0(float* __restrict__ data, float* __restrict__ kernel, float* __restrict__ compute, float* __restrict__ bias) {
       float conv2d_nchw[14];
-      __shared__ float pad_temp_shared[72];
-      __shared__ float kernel_shared[3072];
+      __shared__ float pad_temp_shared[216];
+      __shared__ float kernel_shared[9216];
       conv2d_nchw[0] = 0.000000e+00f;
       conv2d_nchw[1] = 0.000000e+00f;
       conv2d_nchw[2] = 0.000000e+00f;
@@ -922,412 +1782,1181 @@ They can be used for debugging and learning the behavior of the auto-scheduler.
       conv2d_nchw[12] = 0.000000e+00f;
       conv2d_nchw[13] = 0.000000e+00f;
       for (int rc_outer_outer = 0; rc_outer_outer < 64; ++rc_outer_outer) {
-        for (int ry_outer_outer = 0; ry_outer_outer < 3; ++ry_outer_outer) {
-          __syncthreads();
-          if (((int)threadIdx.x) < 18) {
-            pad_temp_shared[(((int)threadIdx.x) * 4)] = (((((1 <= (ry_outer_outer + (((int)blockIdx.x) % 7))) && ((ry_outer_outer + (((int)blockIdx.x) % 7)) < 8)) && (1 <= ((((int)threadIdx.x) * 4) % 9))) && (((((int)threadIdx.x) * 4) % 9) < 8)) ? data[((((((rc_outer_outer * 392) + (((((int)threadIdx.x) * 4) / 9) * 49)) + (ry_outer_outer * 7)) + ((((int)blockIdx.x) % 7) * 7)) + ((((int)threadIdx.x) * 4) % 9)) - 8)] : 0.000000e+00f);
-          }
-          if (((int)threadIdx.x) < 18) {
-            pad_temp_shared[((((int)threadIdx.x) * 4) + 1)] = (((((1 <= (ry_outer_outer + (((int)blockIdx.x) % 7))) && ((ry_outer_outer + (((int)blockIdx.x) % 7)) < 8)) && (1 <= (((((int)threadIdx.x) * 4) + 1) % 9))) && ((((((int)threadIdx.x) * 4) + 1) % 9) < 8)) ? data[((((((rc_outer_outer * 392) + ((((((int)threadIdx.x) * 4) + 1) / 9) * 49)) + (ry_outer_outer * 7)) + ((((int)blockIdx.x) % 7) * 7)) + (((((int)threadIdx.x) * 4) + 1) % 9)) - 8)] : 0.000000e+00f);
-          }
-          if (((int)threadIdx.x) < 18) {
-            pad_temp_shared[((((int)threadIdx.x) * 4) + 2)] = (((((1 <= (ry_outer_outer + (((int)blockIdx.x) % 7))) && ((ry_outer_outer + (((int)blockIdx.x) % 7)) < 8)) && (1 <= (((((int)threadIdx.x) * 4) + 2) % 9))) && ((((((int)threadIdx.x) * 4) + 2) % 9) < 8)) ? data[((((((rc_outer_outer * 392) + ((((((int)threadIdx.x) * 4) + 2) / 9) * 49)) + (ry_outer_outer * 7)) + ((((int)blockIdx.x) % 7) * 7)) + (((((int)threadIdx.x) * 4) + 2) % 9)) - 8)] : 0.000000e+00f);
-          }
-          if (((int)threadIdx.x) < 18) {
-            pad_temp_shared[((((int)threadIdx.x) * 4) + 3)] = (((((1 <= (ry_outer_outer + (((int)blockIdx.x) % 7))) && ((ry_outer_outer + (((int)blockIdx.x) % 7)) < 8)) && (1 <= (((((int)threadIdx.x) * 4) + 3) % 9))) && ((((((int)threadIdx.x) * 4) + 3) % 9) < 8)) ? data[((((((rc_outer_outer * 392) + ((((((int)threadIdx.x) * 4) + 3) / 9) * 49)) + (ry_outer_outer * 7)) + ((((int)blockIdx.x) % 7) * 7)) + (((((int)threadIdx.x) * 4) + 3) % 9)) - 8)] : 0.000000e+00f);
-          }
-          kernel_shared[((int)threadIdx.x)] = kernel[(((((((((int)blockIdx.x) / 7) * 589824) + ((((int)threadIdx.x) / 24) * 4608)) + (rc_outer_outer * 72)) + (((((int)threadIdx.x) % 24) / 3) * 9)) + (ry_outer_outer * 3)) + (((int)threadIdx.x) % 3))];
-          kernel_shared[(((int)threadIdx.x) + 64)] = kernel[(((((((((int)blockIdx.x) / 7) * 589824) + (((((int)threadIdx.x) + 64) / 24) * 4608)) + (rc_outer_outer * 72)) + ((((((int)threadIdx.x) + 16) % 24) / 3) * 9)) + (ry_outer_outer * 3)) + ((((int)threadIdx.x) + 1) % 3))];
-          kernel_shared[(((int)threadIdx.x) + 128)] = kernel[(((((((((int)blockIdx.x) / 7) * 589824) + (((((int)threadIdx.x) + 128) / 24) * 4608)) + (rc_outer_outer * 72)) + ((((((int)threadIdx.x) + 8) % 24) / 3) * 9)) + (ry_outer_outer * 3)) + ((((int)threadIdx.x) + 2) % 3))];
-          kernel_shared[(((int)threadIdx.x) + 192)] = kernel[((((((((((int)blockIdx.x) / 7) * 589824) + ((((int)threadIdx.x) / 24) * 4608)) + (rc_outer_outer * 72)) + (((((int)threadIdx.x) % 24) / 3) * 9)) + (ry_outer_outer * 3)) + (((int)threadIdx.x) % 3)) + 36864)];
-          kernel_shared[(((int)threadIdx.x) + 256)] = kernel[(((((((((int)blockIdx.x) / 7) * 589824) + (((((int)threadIdx.x) + 256) / 24) * 4608)) + (rc_outer_outer * 72)) + ((((((int)threadIdx.x) + 16) % 24) / 3) * 9)) + (ry_outer_outer * 3)) + ((((int)threadIdx.x) + 1) % 3))];
-          kernel_shared[(((int)threadIdx.x) + 320)] = kernel[(((((((((int)blockIdx.x) / 7) * 589824) + (((((int)threadIdx.x) + 320) / 24) * 4608)) + (rc_outer_outer * 72)) + ((((((int)threadIdx.x) + 8) % 24) / 3) * 9)) + (ry_outer_outer * 3)) + ((((int)threadIdx.x) + 2) % 3))];
-          kernel_shared[(((int)threadIdx.x) + 384)] = kernel[((((((((((int)blockIdx.x) / 7) * 589824) + ((((int)threadIdx.x) / 24) * 4608)) + (rc_outer_outer * 72)) + (((((int)threadIdx.x) % 24) / 3) * 9)) + (ry_outer_outer * 3)) + (((int)threadIdx.x) % 3)) + 73728)];
-          kernel_shared[(((int)threadIdx.x) + 448)] = kernel[(((((((((int)blockIdx.x) / 7) * 589824) + (((((int)threadIdx.x) + 448) / 24) * 4608)) + (rc_outer_outer * 72)) + ((((((int)threadIdx.x) + 16) % 24) / 3) * 9)) + (ry_outer_outer * 3)) + ((((int)threadIdx.x) + 1) % 3))];
-          kernel_shared[(((int)threadIdx.x) + 512)] = kernel[(((((((((int)blockIdx.x) / 7) * 589824) + (((((int)threadIdx.x) + 512) / 24) * 4608)) + (rc_outer_outer * 72)) + ((((((int)threadIdx.x) + 8) % 24) / 3) * 9)) + (ry_outer_outer * 3)) + ((((int)threadIdx.x) + 2) % 3))];
-          kernel_shared[(((int)threadIdx.x) + 576)] = kernel[((((((((((int)blockIdx.x) / 7) * 589824) + ((((int)threadIdx.x) / 24) * 4608)) + (rc_outer_outer * 72)) + (((((int)threadIdx.x) % 24) / 3) * 9)) + (ry_outer_outer * 3)) + (((int)threadIdx.x) % 3)) + 110592)];
-          kernel_shared[(((int)threadIdx.x) + 640)] = kernel[(((((((((int)blockIdx.x) / 7) * 589824) + (((((int)threadIdx.x) + 640) / 24) * 4608)) + (rc_outer_outer * 72)) + ((((((int)threadIdx.x) + 16) % 24) / 3) * 9)) + (ry_outer_outer * 3)) + ((((int)threadIdx.x) + 1) % 3))];
-          kernel_shared[(((int)threadIdx.x) + 704)] = kernel[(((((((((int)blockIdx.x) / 7) * 589824) + (((((int)threadIdx.x) + 704) / 24) * 4608)) + (rc_outer_outer * 72)) + ((((((int)threadIdx.x) + 8) % 24) / 3) * 9)) + (ry_outer_outer * 3)) + ((((int)threadIdx.x) + 2) % 3))];
-          kernel_shared[(((int)threadIdx.x) + 768)] = kernel[((((((((((int)blockIdx.x) / 7) * 589824) + ((((int)threadIdx.x) / 24) * 4608)) + (rc_outer_outer * 72)) + (((((int)threadIdx.x) % 24) / 3) * 9)) + (ry_outer_outer * 3)) + (((int)threadIdx.x) % 3)) + 147456)];
-          kernel_shared[(((int)threadIdx.x) + 832)] = kernel[(((((((((int)blockIdx.x) / 7) * 589824) + (((((int)threadIdx.x) + 832) / 24) * 4608)) + (rc_outer_outer * 72)) + ((((((int)threadIdx.x) + 16) % 24) / 3) * 9)) + (ry_outer_outer * 3)) + ((((int)threadIdx.x) + 1) % 3))];
-          kernel_shared[(((int)threadIdx.x) + 896)] = kernel[(((((((((int)blockIdx.x) / 7) * 589824) + (((((int)threadIdx.x) + 896) / 24) * 4608)) + (rc_outer_outer * 72)) + ((((((int)threadIdx.x) + 8) % 24) / 3) * 9)) + (ry_outer_outer * 3)) + ((((int)threadIdx.x) + 2) % 3))];
-          kernel_shared[(((int)threadIdx.x) + 960)] = kernel[((((((((((int)blockIdx.x) / 7) * 589824) + ((((int)threadIdx.x) / 24) * 4608)) + (rc_outer_outer * 72)) + (((((int)threadIdx.x) % 24) / 3) * 9)) + (ry_outer_outer * 3)) + (((int)threadIdx.x) % 3)) + 184320)];
-          kernel_shared[(((int)threadIdx.x) + 1024)] = kernel[(((((((((int)blockIdx.x) / 7) * 589824) + (((((int)threadIdx.x) + 1024) / 24) * 4608)) + (rc_outer_outer * 72)) + ((((((int)threadIdx.x) + 16) % 24) / 3) * 9)) + (ry_outer_outer * 3)) + ((((int)threadIdx.x) + 1) % 3))];
-          kernel_shared[(((int)threadIdx.x) + 1088)] = kernel[(((((((((int)blockIdx.x) / 7) * 589824) + (((((int)threadIdx.x) + 1088) / 24) * 4608)) + (rc_outer_outer * 72)) + ((((((int)threadIdx.x) + 8) % 24) / 3) * 9)) + (ry_outer_outer * 3)) + ((((int)threadIdx.x) + 2) % 3))];
-          kernel_shared[(((int)threadIdx.x) + 1152)] = kernel[((((((((((int)blockIdx.x) / 7) * 589824) + ((((int)threadIdx.x) / 24) * 4608)) + (rc_outer_outer * 72)) + (((((int)threadIdx.x) % 24) / 3) * 9)) + (ry_outer_outer * 3)) + (((int)threadIdx.x) % 3)) + 221184)];
-          kernel_shared[(((int)threadIdx.x) + 1216)] = kernel[(((((((((int)blockIdx.x) / 7) * 589824) + (((((int)threadIdx.x) + 1216) / 24) * 4608)) + (rc_outer_outer * 72)) + ((((((int)threadIdx.x) + 16) % 24) / 3) * 9)) + (ry_outer_outer * 3)) + ((((int)threadIdx.x) + 1) % 3))];
-          kernel_shared[(((int)threadIdx.x) + 1280)] = kernel[(((((((((int)blockIdx.x) / 7) * 589824) + (((((int)threadIdx.x) + 1280) / 24) * 4608)) + (rc_outer_outer * 72)) + ((((((int)threadIdx.x) + 8) % 24) / 3) * 9)) + (ry_outer_outer * 3)) + ((((int)threadIdx.x) + 2) % 3))];
-          kernel_shared[(((int)threadIdx.x) + 1344)] = kernel[((((((((((int)blockIdx.x) / 7) * 589824) + ((((int)threadIdx.x) / 24) * 4608)) + (rc_outer_outer * 72)) + (((((int)threadIdx.x) % 24) / 3) * 9)) + (ry_outer_outer * 3)) + (((int)threadIdx.x) % 3)) + 258048)];
-          kernel_shared[(((int)threadIdx.x) + 1408)] = kernel[(((((((((int)blockIdx.x) / 7) * 589824) + (((((int)threadIdx.x) + 1408) / 24) * 4608)) + (rc_outer_outer * 72)) + ((((((int)threadIdx.x) + 16) % 24) / 3) * 9)) + (ry_outer_outer * 3)) + ((((int)threadIdx.x) + 1) % 3))];
-          kernel_shared[(((int)threadIdx.x) + 1472)] = kernel[(((((((((int)blockIdx.x) / 7) * 589824) + (((((int)threadIdx.x) + 1472) / 24) * 4608)) + (rc_outer_outer * 72)) + ((((((int)threadIdx.x) + 8) % 24) / 3) * 9)) + (ry_outer_outer * 3)) + ((((int)threadIdx.x) + 2) % 3))];
-          kernel_shared[(((int)threadIdx.x) + 1536)] = kernel[((((((((((int)blockIdx.x) / 7) * 589824) + ((((int)threadIdx.x) / 24) * 4608)) + (rc_outer_outer * 72)) + (((((int)threadIdx.x) % 24) / 3) * 9)) + (ry_outer_outer * 3)) + (((int)threadIdx.x) % 3)) + 294912)];
-          kernel_shared[(((int)threadIdx.x) + 1600)] = kernel[(((((((((int)blockIdx.x) / 7) * 589824) + (((((int)threadIdx.x) + 1600) / 24) * 4608)) + (rc_outer_outer * 72)) + ((((((int)threadIdx.x) + 16) % 24) / 3) * 9)) + (ry_outer_outer * 3)) + ((((int)threadIdx.x) + 1) % 3))];
-          kernel_shared[(((int)threadIdx.x) + 1664)] = kernel[(((((((((int)blockIdx.x) / 7) * 589824) + (((((int)threadIdx.x) + 1664) / 24) * 4608)) + (rc_outer_outer * 72)) + ((((((int)threadIdx.x) + 8) % 24) / 3) * 9)) + (ry_outer_outer * 3)) + ((((int)threadIdx.x) + 2) % 3))];
-          kernel_shared[(((int)threadIdx.x) + 1728)] = kernel[((((((((((int)blockIdx.x) / 7) * 589824) + ((((int)threadIdx.x) / 24) * 4608)) + (rc_outer_outer * 72)) + (((((int)threadIdx.x) % 24) / 3) * 9)) + (ry_outer_outer * 3)) + (((int)threadIdx.x) % 3)) + 331776)];
-          kernel_shared[(((int)threadIdx.x) + 1792)] = kernel[(((((((((int)blockIdx.x) / 7) * 589824) + (((((int)threadIdx.x) + 1792) / 24) * 4608)) + (rc_outer_outer * 72)) + ((((((int)threadIdx.x) + 16) % 24) / 3) * 9)) + (ry_outer_outer * 3)) + ((((int)threadIdx.x) + 1) % 3))];
-          kernel_shared[(((int)threadIdx.x) + 1856)] = kernel[(((((((((int)blockIdx.x) / 7) * 589824) + (((((int)threadIdx.x) + 1856) / 24) * 4608)) + (rc_outer_outer * 72)) + ((((((int)threadIdx.x) + 8) % 24) / 3) * 9)) + (ry_outer_outer * 3)) + ((((int)threadIdx.x) + 2) % 3))];
-          kernel_shared[(((int)threadIdx.x) + 1920)] = kernel[((((((((((int)blockIdx.x) / 7) * 589824) + ((((int)threadIdx.x) / 24) * 4608)) + (rc_outer_outer * 72)) + (((((int)threadIdx.x) % 24) / 3) * 9)) + (ry_outer_outer * 3)) + (((int)threadIdx.x) % 3)) + 368640)];
-          kernel_shared[(((int)threadIdx.x) + 1984)] = kernel[(((((((((int)blockIdx.x) / 7) * 589824) + (((((int)threadIdx.x) + 1984) / 24) * 4608)) + (rc_outer_outer * 72)) + ((((((int)threadIdx.x) + 16) % 24) / 3) * 9)) + (ry_outer_outer * 3)) + ((((int)threadIdx.x) + 1) % 3))];
-          kernel_shared[(((int)threadIdx.x) + 2048)] = kernel[(((((((((int)blockIdx.x) / 7) * 589824) + (((((int)threadIdx.x) + 2048) / 24) * 4608)) + (rc_outer_outer * 72)) + ((((((int)threadIdx.x) + 8) % 24) / 3) * 9)) + (ry_outer_outer * 3)) + ((((int)threadIdx.x) + 2) % 3))];
-          kernel_shared[(((int)threadIdx.x) + 2112)] = kernel[((((((((((int)blockIdx.x) / 7) * 589824) + ((((int)threadIdx.x) / 24) * 4608)) + (rc_outer_outer * 72)) + (((((int)threadIdx.x) % 24) / 3) * 9)) + (ry_outer_outer * 3)) + (((int)threadIdx.x) % 3)) + 405504)];
-          kernel_shared[(((int)threadIdx.x) + 2176)] = kernel[(((((((((int)blockIdx.x) / 7) * 589824) + (((((int)threadIdx.x) + 2176) / 24) * 4608)) + (rc_outer_outer * 72)) + ((((((int)threadIdx.x) + 16) % 24) / 3) * 9)) + (ry_outer_outer * 3)) + ((((int)threadIdx.x) + 1) % 3))];
-          kernel_shared[(((int)threadIdx.x) + 2240)] = kernel[(((((((((int)blockIdx.x) / 7) * 589824) + (((((int)threadIdx.x) + 2240) / 24) * 4608)) + (rc_outer_outer * 72)) + ((((((int)threadIdx.x) + 8) % 24) / 3) * 9)) + (ry_outer_outer * 3)) + ((((int)threadIdx.x) + 2) % 3))];
-          kernel_shared[(((int)threadIdx.x) + 2304)] = kernel[((((((((((int)blockIdx.x) / 7) * 589824) + ((((int)threadIdx.x) / 24) * 4608)) + (rc_outer_outer * 72)) + (((((int)threadIdx.x) % 24) / 3) * 9)) + (ry_outer_outer * 3)) + (((int)threadIdx.x) % 3)) + 442368)];
-          kernel_shared[(((int)threadIdx.x) + 2368)] = kernel[(((((((((int)blockIdx.x) / 7) * 589824) + (((((int)threadIdx.x) + 2368) / 24) * 4608)) + (rc_outer_outer * 72)) + ((((((int)threadIdx.x) + 16) % 24) / 3) * 9)) + (ry_outer_outer * 3)) + ((((int)threadIdx.x) + 1) % 3))];
-          kernel_shared[(((int)threadIdx.x) + 2432)] = kernel[(((((((((int)blockIdx.x) / 7) * 589824) + (((((int)threadIdx.x) + 2432) / 24) * 4608)) + (rc_outer_outer * 72)) + ((((((int)threadIdx.x) + 8) % 24) / 3) * 9)) + (ry_outer_outer * 3)) + ((((int)threadIdx.x) + 2) % 3))];
-          kernel_shared[(((int)threadIdx.x) + 2496)] = kernel[((((((((((int)blockIdx.x) / 7) * 589824) + ((((int)threadIdx.x) / 24) * 4608)) + (rc_outer_outer * 72)) + (((((int)threadIdx.x) % 24) / 3) * 9)) + (ry_outer_outer * 3)) + (((int)threadIdx.x) % 3)) + 479232)];
-          kernel_shared[(((int)threadIdx.x) + 2560)] = kernel[(((((((((int)blockIdx.x) / 7) * 589824) + (((((int)threadIdx.x) + 2560) / 24) * 4608)) + (rc_outer_outer * 72)) + ((((((int)threadIdx.x) + 16) % 24) / 3) * 9)) + (ry_outer_outer * 3)) + ((((int)threadIdx.x) + 1) % 3))];
-          kernel_shared[(((int)threadIdx.x) + 2624)] = kernel[(((((((((int)blockIdx.x) / 7) * 589824) + (((((int)threadIdx.x) + 2624) / 24) * 4608)) + (rc_outer_outer * 72)) + ((((((int)threadIdx.x) + 8) % 24) / 3) * 9)) + (ry_outer_outer * 3)) + ((((int)threadIdx.x) + 2) % 3))];
-          kernel_shared[(((int)threadIdx.x) + 2688)] = kernel[((((((((((int)blockIdx.x) / 7) * 589824) + ((((int)threadIdx.x) / 24) * 4608)) + (rc_outer_outer * 72)) + (((((int)threadIdx.x) % 24) / 3) * 9)) + (ry_outer_outer * 3)) + (((int)threadIdx.x) % 3)) + 516096)];
-          kernel_shared[(((int)threadIdx.x) + 2752)] = kernel[(((((((((int)blockIdx.x) / 7) * 589824) + (((((int)threadIdx.x) + 2752) / 24) * 4608)) + (rc_outer_outer * 72)) + ((((((int)threadIdx.x) + 16) % 24) / 3) * 9)) + (ry_outer_outer * 3)) + ((((int)threadIdx.x) + 1) % 3))];
-          kernel_shared[(((int)threadIdx.x) + 2816)] = kernel[(((((((((int)blockIdx.x) / 7) * 589824) + (((((int)threadIdx.x) + 2816) / 24) * 4608)) + (rc_outer_outer * 72)) + ((((((int)threadIdx.x) + 8) % 24) / 3) * 9)) + (ry_outer_outer * 3)) + ((((int)threadIdx.x) + 2) % 3))];
-          kernel_shared[(((int)threadIdx.x) + 2880)] = kernel[((((((((((int)blockIdx.x) / 7) * 589824) + ((((int)threadIdx.x) / 24) * 4608)) + (rc_outer_outer * 72)) + (((((int)threadIdx.x) % 24) / 3) * 9)) + (ry_outer_outer * 3)) + (((int)threadIdx.x) % 3)) + 552960)];
-          kernel_shared[(((int)threadIdx.x) + 2944)] = kernel[(((((((((int)blockIdx.x) / 7) * 589824) + (((((int)threadIdx.x) + 2944) / 24) * 4608)) + (rc_outer_outer * 72)) + ((((((int)threadIdx.x) + 16) % 24) / 3) * 9)) + (ry_outer_outer * 3)) + ((((int)threadIdx.x) + 1) % 3))];
-          kernel_shared[(((int)threadIdx.x) + 3008)] = kernel[(((((((((int)blockIdx.x) / 7) * 589824) + (((((int)threadIdx.x) + 3008) / 24) * 4608)) + (rc_outer_outer * 72)) + ((((((int)threadIdx.x) + 8) % 24) / 3) * 9)) + (ry_outer_outer * 3)) + ((((int)threadIdx.x) + 2) % 3))];
-          __syncthreads();
-          conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[0] * kernel_shared[(((int)threadIdx.x) * 48)]));
-          conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[9] * kernel_shared[((((int)threadIdx.x) * 48) + 3)]));
-          conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[1] * kernel_shared[(((int)threadIdx.x) * 48)]));
-          conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[10] * kernel_shared[((((int)threadIdx.x) * 48) + 3)]));
-          conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[2] * kernel_shared[(((int)threadIdx.x) * 48)]));
-          conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[11] * kernel_shared[((((int)threadIdx.x) * 48) + 3)]));
-          conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[3] * kernel_shared[(((int)threadIdx.x) * 48)]));
-          conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[12] * kernel_shared[((((int)threadIdx.x) * 48) + 3)]));
-          conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[4] * kernel_shared[(((int)threadIdx.x) * 48)]));
-          conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[13] * kernel_shared[((((int)threadIdx.x) * 48) + 3)]));
-          conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[5] * kernel_shared[(((int)threadIdx.x) * 48)]));
-          conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[14] * kernel_shared[((((int)threadIdx.x) * 48) + 3)]));
-          conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[6] * kernel_shared[(((int)threadIdx.x) * 48)]));
-          conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[15] * kernel_shared[((((int)threadIdx.x) * 48) + 3)]));
-          conv2d_nchw[7] = (conv2d_nchw[7] + (pad_temp_shared[0] * kernel_shared[((((int)threadIdx.x) * 48) + 24)]));
-          conv2d_nchw[7] = (conv2d_nchw[7] + (pad_temp_shared[9] * kernel_shared[((((int)threadIdx.x) * 48) + 27)]));
-          conv2d_nchw[8] = (conv2d_nchw[8] + (pad_temp_shared[1] * kernel_shared[((((int)threadIdx.x) * 48) + 24)]));
-          conv2d_nchw[8] = (conv2d_nchw[8] + (pad_temp_shared[10] * kernel_shared[((((int)threadIdx.x) * 48) + 27)]));
-          conv2d_nchw[9] = (conv2d_nchw[9] + (pad_temp_shared[2] * kernel_shared[((((int)threadIdx.x) * 48) + 24)]));
-          conv2d_nchw[9] = (conv2d_nchw[9] + (pad_temp_shared[11] * kernel_shared[((((int)threadIdx.x) * 48) + 27)]));
-          conv2d_nchw[10] = (conv2d_nchw[10] + (pad_temp_shared[3] * kernel_shared[((((int)threadIdx.x) * 48) + 24)]));
-          conv2d_nchw[10] = (conv2d_nchw[10] + (pad_temp_shared[12] * kernel_shared[((((int)threadIdx.x) * 48) + 27)]));
-          conv2d_nchw[11] = (conv2d_nchw[11] + (pad_temp_shared[4] * kernel_shared[((((int)threadIdx.x) * 48) + 24)]));
-          conv2d_nchw[11] = (conv2d_nchw[11] + (pad_temp_shared[13] * kernel_shared[((((int)threadIdx.x) * 48) + 27)]));
-          conv2d_nchw[12] = (conv2d_nchw[12] + (pad_temp_shared[5] * kernel_shared[((((int)threadIdx.x) * 48) + 24)]));
-          conv2d_nchw[12] = (conv2d_nchw[12] + (pad_temp_shared[14] * kernel_shared[((((int)threadIdx.x) * 48) + 27)]));
-          conv2d_nchw[13] = (conv2d_nchw[13] + (pad_temp_shared[6] * kernel_shared[((((int)threadIdx.x) * 48) + 24)]));
-          conv2d_nchw[13] = (conv2d_nchw[13] + (pad_temp_shared[15] * kernel_shared[((((int)threadIdx.x) * 48) + 27)]));
-          conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[1] * kernel_shared[((((int)threadIdx.x) * 48) + 1)]));
-          conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[10] * kernel_shared[((((int)threadIdx.x) * 48) + 4)]));
-          conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[2] * kernel_shared[((((int)threadIdx.x) * 48) + 1)]));
-          conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[11] * kernel_shared[((((int)threadIdx.x) * 48) + 4)]));
-          conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[3] * kernel_shared[((((int)threadIdx.x) * 48) + 1)]));
-          conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[12] * kernel_shared[((((int)threadIdx.x) * 48) + 4)]));
-          conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[4] * kernel_shared[((((int)threadIdx.x) * 48) + 1)]));
-          conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[13] * kernel_shared[((((int)threadIdx.x) * 48) + 4)]));
-          conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[5] * kernel_shared[((((int)threadIdx.x) * 48) + 1)]));
-          conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[14] * kernel_shared[((((int)threadIdx.x) * 48) + 4)]));
-          conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[6] * kernel_shared[((((int)threadIdx.x) * 48) + 1)]));
-          conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[15] * kernel_shared[((((int)threadIdx.x) * 48) + 4)]));
-          conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[7] * kernel_shared[((((int)threadIdx.x) * 48) + 1)]));
-          conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[16] * kernel_shared[((((int)threadIdx.x) * 48) + 4)]));
-          conv2d_nchw[7] = (conv2d_nchw[7] + (pad_temp_shared[1] * kernel_shared[((((int)threadIdx.x) * 48) + 25)]));
-          conv2d_nchw[7] = (conv2d_nchw[7] + (pad_temp_shared[10] * kernel_shared[((((int)threadIdx.x) * 48) + 28)]));
-          conv2d_nchw[8] = (conv2d_nchw[8] + (pad_temp_shared[2] * kernel_shared[((((int)threadIdx.x) * 48) + 25)]));
-          conv2d_nchw[8] = (conv2d_nchw[8] + (pad_temp_shared[11] * kernel_shared[((((int)threadIdx.x) * 48) + 28)]));
-          conv2d_nchw[9] = (conv2d_nchw[9] + (pad_temp_shared[3] * kernel_shared[((((int)threadIdx.x) * 48) + 25)]));
-          conv2d_nchw[9] = (conv2d_nchw[9] + (pad_temp_shared[12] * kernel_shared[((((int)threadIdx.x) * 48) + 28)]));
-          conv2d_nchw[10] = (conv2d_nchw[10] + (pad_temp_shared[4] * kernel_shared[((((int)threadIdx.x) * 48) + 25)]));
-          conv2d_nchw[10] = (conv2d_nchw[10] + (pad_temp_shared[13] * kernel_shared[((((int)threadIdx.x) * 48) + 28)]));
-          conv2d_nchw[11] = (conv2d_nchw[11] + (pad_temp_shared[5] * kernel_shared[((((int)threadIdx.x) * 48) + 25)]));
-          conv2d_nchw[11] = (conv2d_nchw[11] + (pad_temp_shared[14] * kernel_shared[((((int)threadIdx.x) * 48) + 28)]));
-          conv2d_nchw[12] = (conv2d_nchw[12] + (pad_temp_shared[6] * kernel_shared[((((int)threadIdx.x) * 48) + 25)]));
-          conv2d_nchw[12] = (conv2d_nchw[12] + (pad_temp_shared[15] * kernel_shared[((((int)threadIdx.x) * 48) + 28)]));
-          conv2d_nchw[13] = (conv2d_nchw[13] + (pad_temp_shared[7] * kernel_shared[((((int)threadIdx.x) * 48) + 25)]));
-          conv2d_nchw[13] = (conv2d_nchw[13] + (pad_temp_shared[16] * kernel_shared[((((int)threadIdx.x) * 48) + 28)]));
-          conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[2] * kernel_shared[((((int)threadIdx.x) * 48) + 2)]));
-          conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[11] * kernel_shared[((((int)threadIdx.x) * 48) + 5)]));
-          conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[3] * kernel_shared[((((int)threadIdx.x) * 48) + 2)]));
-          conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[12] * kernel_shared[((((int)threadIdx.x) * 48) + 5)]));
-          conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[4] * kernel_shared[((((int)threadIdx.x) * 48) + 2)]));
-          conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[13] * kernel_shared[((((int)threadIdx.x) * 48) + 5)]));
-          conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[5] * kernel_shared[((((int)threadIdx.x) * 48) + 2)]));
-          conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[14] * kernel_shared[((((int)threadIdx.x) * 48) + 5)]));
-          conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[6] * kernel_shared[((((int)threadIdx.x) * 48) + 2)]));
-          conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[15] * kernel_shared[((((int)threadIdx.x) * 48) + 5)]));
-          conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[7] * kernel_shared[((((int)threadIdx.x) * 48) + 2)]));
-          conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[16] * kernel_shared[((((int)threadIdx.x) * 48) + 5)]));
-          conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[8] * kernel_shared[((((int)threadIdx.x) * 48) + 2)]));
-          conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[17] * kernel_shared[((((int)threadIdx.x) * 48) + 5)]));
-          conv2d_nchw[7] = (conv2d_nchw[7] + (pad_temp_shared[2] * kernel_shared[((((int)threadIdx.x) * 48) + 26)]));
-          conv2d_nchw[7] = (conv2d_nchw[7] + (pad_temp_shared[11] * kernel_shared[((((int)threadIdx.x) * 48) + 29)]));
-          conv2d_nchw[8] = (conv2d_nchw[8] + (pad_temp_shared[3] * kernel_shared[((((int)threadIdx.x) * 48) + 26)]));
-          conv2d_nchw[8] = (conv2d_nchw[8] + (pad_temp_shared[12] * kernel_shared[((((int)threadIdx.x) * 48) + 29)]));
-          conv2d_nchw[9] = (conv2d_nchw[9] + (pad_temp_shared[4] * kernel_shared[((((int)threadIdx.x) * 48) + 26)]));
-          conv2d_nchw[9] = (conv2d_nchw[9] + (pad_temp_shared[13] * kernel_shared[((((int)threadIdx.x) * 48) + 29)]));
-          conv2d_nchw[10] = (conv2d_nchw[10] + (pad_temp_shared[5] * kernel_shared[((((int)threadIdx.x) * 48) + 26)]));
-          conv2d_nchw[10] = (conv2d_nchw[10] + (pad_temp_shared[14] * kernel_shared[((((int)threadIdx.x) * 48) + 29)]));
-          conv2d_nchw[11] = (conv2d_nchw[11] + (pad_temp_shared[6] * kernel_shared[((((int)threadIdx.x) * 48) + 26)]));
-          conv2d_nchw[11] = (conv2d_nchw[11] + (pad_temp_shared[15] * kernel_shared[((((int)threadIdx.x) * 48) + 29)]));
-          conv2d_nchw[12] = (conv2d_nchw[12] + (pad_temp_shared[7] * kernel_shared[((((int)threadIdx.x) * 48) + 26)]));
-          conv2d_nchw[12] = (conv2d_nchw[12] + (pad_temp_shared[16] * kernel_shared[((((int)threadIdx.x) * 48) + 29)]));
-          conv2d_nchw[13] = (conv2d_nchw[13] + (pad_temp_shared[8] * kernel_shared[((((int)threadIdx.x) * 48) + 26)]));
-          conv2d_nchw[13] = (conv2d_nchw[13] + (pad_temp_shared[17] * kernel_shared[((((int)threadIdx.x) * 48) + 29)]));
-          conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[18] * kernel_shared[((((int)threadIdx.x) * 48) + 6)]));
-          conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[27] * kernel_shared[((((int)threadIdx.x) * 48) + 9)]));
-          conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[19] * kernel_shared[((((int)threadIdx.x) * 48) + 6)]));
-          conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[28] * kernel_shared[((((int)threadIdx.x) * 48) + 9)]));
-          conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[20] * kernel_shared[((((int)threadIdx.x) * 48) + 6)]));
-          conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[29] * kernel_shared[((((int)threadIdx.x) * 48) + 9)]));
-          conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[21] * kernel_shared[((((int)threadIdx.x) * 48) + 6)]));
-          conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[30] * kernel_shared[((((int)threadIdx.x) * 48) + 9)]));
-          conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[22] * kernel_shared[((((int)threadIdx.x) * 48) + 6)]));
-          conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[31] * kernel_shared[((((int)threadIdx.x) * 48) + 9)]));
-          conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[23] * kernel_shared[((((int)threadIdx.x) * 48) + 6)]));
-          conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[32] * kernel_shared[((((int)threadIdx.x) * 48) + 9)]));
-          conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[24] * kernel_shared[((((int)threadIdx.x) * 48) + 6)]));
-          conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[33] * kernel_shared[((((int)threadIdx.x) * 48) + 9)]));
-          conv2d_nchw[7] = (conv2d_nchw[7] + (pad_temp_shared[18] * kernel_shared[((((int)threadIdx.x) * 48) + 30)]));
-          conv2d_nchw[7] = (conv2d_nchw[7] + (pad_temp_shared[27] * kernel_shared[((((int)threadIdx.x) * 48) + 33)]));
-          conv2d_nchw[8] = (conv2d_nchw[8] + (pad_temp_shared[19] * kernel_shared[((((int)threadIdx.x) * 48) + 30)]));
-          conv2d_nchw[8] = (conv2d_nchw[8] + (pad_temp_shared[28] * kernel_shared[((((int)threadIdx.x) * 48) + 33)]));
-          conv2d_nchw[9] = (conv2d_nchw[9] + (pad_temp_shared[20] * kernel_shared[((((int)threadIdx.x) * 48) + 30)]));
-          conv2d_nchw[9] = (conv2d_nchw[9] + (pad_temp_shared[29] * kernel_shared[((((int)threadIdx.x) * 48) + 33)]));
-          conv2d_nchw[10] = (conv2d_nchw[10] + (pad_temp_shared[21] * kernel_shared[((((int)threadIdx.x) * 48) + 30)]));
-          conv2d_nchw[10] = (conv2d_nchw[10] + (pad_temp_shared[30] * kernel_shared[((((int)threadIdx.x) * 48) + 33)]));
-          conv2d_nchw[11] = (conv2d_nchw[11] + (pad_temp_shared[22] * kernel_shared[((((int)threadIdx.x) * 48) + 30)]));
-          conv2d_nchw[11] = (conv2d_nchw[11] + (pad_temp_shared[31] * kernel_shared[((((int)threadIdx.x) * 48) + 33)]));
-          conv2d_nchw[12] = (conv2d_nchw[12] + (pad_temp_shared[23] * kernel_shared[((((int)threadIdx.x) * 48) + 30)]));
-          conv2d_nchw[12] = (conv2d_nchw[12] + (pad_temp_shared[32] * kernel_shared[((((int)threadIdx.x) * 48) + 33)]));
-          conv2d_nchw[13] = (conv2d_nchw[13] + (pad_temp_shared[24] * kernel_shared[((((int)threadIdx.x) * 48) + 30)]));
-          conv2d_nchw[13] = (conv2d_nchw[13] + (pad_temp_shared[33] * kernel_shared[((((int)threadIdx.x) * 48) + 33)]));
-          conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[19] * kernel_shared[((((int)threadIdx.x) * 48) + 7)]));
-          conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[28] * kernel_shared[((((int)threadIdx.x) * 48) + 10)]));
-          conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[20] * kernel_shared[((((int)threadIdx.x) * 48) + 7)]));
-          conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[29] * kernel_shared[((((int)threadIdx.x) * 48) + 10)]));
-          conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[21] * kernel_shared[((((int)threadIdx.x) * 48) + 7)]));
-          conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[30] * kernel_shared[((((int)threadIdx.x) * 48) + 10)]));
-          conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[22] * kernel_shared[((((int)threadIdx.x) * 48) + 7)]));
-          conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[31] * kernel_shared[((((int)threadIdx.x) * 48) + 10)]));
-          conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[23] * kernel_shared[((((int)threadIdx.x) * 48) + 7)]));
-          conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[32] * kernel_shared[((((int)threadIdx.x) * 48) + 10)]));
-          conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[24] * kernel_shared[((((int)threadIdx.x) * 48) + 7)]));
-          conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[33] * kernel_shared[((((int)threadIdx.x) * 48) + 10)]));
-          conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[25] * kernel_shared[((((int)threadIdx.x) * 48) + 7)]));
-          conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[34] * kernel_shared[((((int)threadIdx.x) * 48) + 10)]));
-          conv2d_nchw[7] = (conv2d_nchw[7] + (pad_temp_shared[19] * kernel_shared[((((int)threadIdx.x) * 48) + 31)]));
-          conv2d_nchw[7] = (conv2d_nchw[7] + (pad_temp_shared[28] * kernel_shared[((((int)threadIdx.x) * 48) + 34)]));
-          conv2d_nchw[8] = (conv2d_nchw[8] + (pad_temp_shared[20] * kernel_shared[((((int)threadIdx.x) * 48) + 31)]));
-          conv2d_nchw[8] = (conv2d_nchw[8] + (pad_temp_shared[29] * kernel_shared[((((int)threadIdx.x) * 48) + 34)]));
-          conv2d_nchw[9] = (conv2d_nchw[9] + (pad_temp_shared[21] * kernel_shared[((((int)threadIdx.x) * 48) + 31)]));
-          conv2d_nchw[9] = (conv2d_nchw[9] + (pad_temp_shared[30] * kernel_shared[((((int)threadIdx.x) * 48) + 34)]));
-          conv2d_nchw[10] = (conv2d_nchw[10] + (pad_temp_shared[22] * kernel_shared[((((int)threadIdx.x) * 48) + 31)]));
-          conv2d_nchw[10] = (conv2d_nchw[10] + (pad_temp_shared[31] * kernel_shared[((((int)threadIdx.x) * 48) + 34)]));
-          conv2d_nchw[11] = (conv2d_nchw[11] + (pad_temp_shared[23] * kernel_shared[((((int)threadIdx.x) * 48) + 31)]));
-          conv2d_nchw[11] = (conv2d_nchw[11] + (pad_temp_shared[32] * kernel_shared[((((int)threadIdx.x) * 48) + 34)]));
-          conv2d_nchw[12] = (conv2d_nchw[12] + (pad_temp_shared[24] * kernel_shared[((((int)threadIdx.x) * 48) + 31)]));
-          conv2d_nchw[12] = (conv2d_nchw[12] + (pad_temp_shared[33] * kernel_shared[((((int)threadIdx.x) * 48) + 34)]));
-          conv2d_nchw[13] = (conv2d_nchw[13] + (pad_temp_shared[25] * kernel_shared[((((int)threadIdx.x) * 48) + 31)]));
-          conv2d_nchw[13] = (conv2d_nchw[13] + (pad_temp_shared[34] * kernel_shared[((((int)threadIdx.x) * 48) + 34)]));
-          conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[20] * kernel_shared[((((int)threadIdx.x) * 48) + 8)]));
-          conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[29] * kernel_shared[((((int)threadIdx.x) * 48) + 11)]));
-          conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[21] * kernel_shared[((((int)threadIdx.x) * 48) + 8)]));
-          conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[30] * kernel_shared[((((int)threadIdx.x) * 48) + 11)]));
-          conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[22] * kernel_shared[((((int)threadIdx.x) * 48) + 8)]));
-          conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[31] * kernel_shared[((((int)threadIdx.x) * 48) + 11)]));
-          conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[23] * kernel_shared[((((int)threadIdx.x) * 48) + 8)]));
-          conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[32] * kernel_shared[((((int)threadIdx.x) * 48) + 11)]));
-          conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[24] * kernel_shared[((((int)threadIdx.x) * 48) + 8)]));
-          conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[33] * kernel_shared[((((int)threadIdx.x) * 48) + 11)]));
-          conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[25] * kernel_shared[((((int)threadIdx.x) * 48) + 8)]));
-          conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[34] * kernel_shared[((((int)threadIdx.x) * 48) + 11)]));
-          conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[26] * kernel_shared[((((int)threadIdx.x) * 48) + 8)]));
-          conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[35] * kernel_shared[((((int)threadIdx.x) * 48) + 11)]));
-          conv2d_nchw[7] = (conv2d_nchw[7] + (pad_temp_shared[20] * kernel_shared[((((int)threadIdx.x) * 48) + 32)]));
-          conv2d_nchw[7] = (conv2d_nchw[7] + (pad_temp_shared[29] * kernel_shared[((((int)threadIdx.x) * 48) + 35)]));
-          conv2d_nchw[8] = (conv2d_nchw[8] + (pad_temp_shared[21] * kernel_shared[((((int)threadIdx.x) * 48) + 32)]));
-          conv2d_nchw[8] = (conv2d_nchw[8] + (pad_temp_shared[30] * kernel_shared[((((int)threadIdx.x) * 48) + 35)]));
-          conv2d_nchw[9] = (conv2d_nchw[9] + (pad_temp_shared[22] * kernel_shared[((((int)threadIdx.x) * 48) + 32)]));
-          conv2d_nchw[9] = (conv2d_nchw[9] + (pad_temp_shared[31] * kernel_shared[((((int)threadIdx.x) * 48) + 35)]));
-          conv2d_nchw[10] = (conv2d_nchw[10] + (pad_temp_shared[23] * kernel_shared[((((int)threadIdx.x) * 48) + 32)]));
-          conv2d_nchw[10] = (conv2d_nchw[10] + (pad_temp_shared[32] * kernel_shared[((((int)threadIdx.x) * 48) + 35)]));
-          conv2d_nchw[11] = (conv2d_nchw[11] + (pad_temp_shared[24] * kernel_shared[((((int)threadIdx.x) * 48) + 32)]));
-          conv2d_nchw[11] = (conv2d_nchw[11] + (pad_temp_shared[33] * kernel_shared[((((int)threadIdx.x) * 48) + 35)]));
-          conv2d_nchw[12] = (conv2d_nchw[12] + (pad_temp_shared[25] * kernel_shared[((((int)threadIdx.x) * 48) + 32)]));
-          conv2d_nchw[12] = (conv2d_nchw[12] + (pad_temp_shared[34] * kernel_shared[((((int)threadIdx.x) * 48) + 35)]));
-          conv2d_nchw[13] = (conv2d_nchw[13] + (pad_temp_shared[26] * kernel_shared[((((int)threadIdx.x) * 48) + 32)]));
-          conv2d_nchw[13] = (conv2d_nchw[13] + (pad_temp_shared[35] * kernel_shared[((((int)threadIdx.x) * 48) + 35)]));
-          conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[36] * kernel_shared[((((int)threadIdx.x) * 48) + 12)]));
-          conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[45] * kernel_shared[((((int)threadIdx.x) * 48) + 15)]));
-          conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[37] * kernel_shared[((((int)threadIdx.x) * 48) + 12)]));
-          conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[46] * kernel_shared[((((int)threadIdx.x) * 48) + 15)]));
-          conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[38] * kernel_shared[((((int)threadIdx.x) * 48) + 12)]));
-          conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[47] * kernel_shared[((((int)threadIdx.x) * 48) + 15)]));
-          conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[39] * kernel_shared[((((int)threadIdx.x) * 48) + 12)]));
-          conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[48] * kernel_shared[((((int)threadIdx.x) * 48) + 15)]));
-          conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[40] * kernel_shared[((((int)threadIdx.x) * 48) + 12)]));
-          conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[49] * kernel_shared[((((int)threadIdx.x) * 48) + 15)]));
-          conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[41] * kernel_shared[((((int)threadIdx.x) * 48) + 12)]));
-          conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[50] * kernel_shared[((((int)threadIdx.x) * 48) + 15)]));
-          conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[42] * kernel_shared[((((int)threadIdx.x) * 48) + 12)]));
-          conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[51] * kernel_shared[((((int)threadIdx.x) * 48) + 15)]));
-          conv2d_nchw[7] = (conv2d_nchw[7] + (pad_temp_shared[36] * kernel_shared[((((int)threadIdx.x) * 48) + 36)]));
-          conv2d_nchw[7] = (conv2d_nchw[7] + (pad_temp_shared[45] * kernel_shared[((((int)threadIdx.x) * 48) + 39)]));
-          conv2d_nchw[8] = (conv2d_nchw[8] + (pad_temp_shared[37] * kernel_shared[((((int)threadIdx.x) * 48) + 36)]));
-          conv2d_nchw[8] = (conv2d_nchw[8] + (pad_temp_shared[46] * kernel_shared[((((int)threadIdx.x) * 48) + 39)]));
-          conv2d_nchw[9] = (conv2d_nchw[9] + (pad_temp_shared[38] * kernel_shared[((((int)threadIdx.x) * 48) + 36)]));
-          conv2d_nchw[9] = (conv2d_nchw[9] + (pad_temp_shared[47] * kernel_shared[((((int)threadIdx.x) * 48) + 39)]));
-          conv2d_nchw[10] = (conv2d_nchw[10] + (pad_temp_shared[39] * kernel_shared[((((int)threadIdx.x) * 48) + 36)]));
-          conv2d_nchw[10] = (conv2d_nchw[10] + (pad_temp_shared[48] * kernel_shared[((((int)threadIdx.x) * 48) + 39)]));
-          conv2d_nchw[11] = (conv2d_nchw[11] + (pad_temp_shared[40] * kernel_shared[((((int)threadIdx.x) * 48) + 36)]));
-          conv2d_nchw[11] = (conv2d_nchw[11] + (pad_temp_shared[49] * kernel_shared[((((int)threadIdx.x) * 48) + 39)]));
-          conv2d_nchw[12] = (conv2d_nchw[12] + (pad_temp_shared[41] * kernel_shared[((((int)threadIdx.x) * 48) + 36)]));
-          conv2d_nchw[12] = (conv2d_nchw[12] + (pad_temp_shared[50] * kernel_shared[((((int)threadIdx.x) * 48) + 39)]));
-          conv2d_nchw[13] = (conv2d_nchw[13] + (pad_temp_shared[42] * kernel_shared[((((int)threadIdx.x) * 48) + 36)]));
-          conv2d_nchw[13] = (conv2d_nchw[13] + (pad_temp_shared[51] * kernel_shared[((((int)threadIdx.x) * 48) + 39)]));
-          conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[37] * kernel_shared[((((int)threadIdx.x) * 48) + 13)]));
-          conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[46] * kernel_shared[((((int)threadIdx.x) * 48) + 16)]));
-          conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[38] * kernel_shared[((((int)threadIdx.x) * 48) + 13)]));
-          conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[47] * kernel_shared[((((int)threadIdx.x) * 48) + 16)]));
-          conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[39] * kernel_shared[((((int)threadIdx.x) * 48) + 13)]));
-          conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[48] * kernel_shared[((((int)threadIdx.x) * 48) + 16)]));
-          conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[40] * kernel_shared[((((int)threadIdx.x) * 48) + 13)]));
-          conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[49] * kernel_shared[((((int)threadIdx.x) * 48) + 16)]));
-          conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[41] * kernel_shared[((((int)threadIdx.x) * 48) + 13)]));
-          conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[50] * kernel_shared[((((int)threadIdx.x) * 48) + 16)]));
-          conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[42] * kernel_shared[((((int)threadIdx.x) * 48) + 13)]));
-          conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[51] * kernel_shared[((((int)threadIdx.x) * 48) + 16)]));
-          conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[43] * kernel_shared[((((int)threadIdx.x) * 48) + 13)]));
-          conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[52] * kernel_shared[((((int)threadIdx.x) * 48) + 16)]));
-          conv2d_nchw[7] = (conv2d_nchw[7] + (pad_temp_shared[37] * kernel_shared[((((int)threadIdx.x) * 48) + 37)]));
-          conv2d_nchw[7] = (conv2d_nchw[7] + (pad_temp_shared[46] * kernel_shared[((((int)threadIdx.x) * 48) + 40)]));
-          conv2d_nchw[8] = (conv2d_nchw[8] + (pad_temp_shared[38] * kernel_shared[((((int)threadIdx.x) * 48) + 37)]));
-          conv2d_nchw[8] = (conv2d_nchw[8] + (pad_temp_shared[47] * kernel_shared[((((int)threadIdx.x) * 48) + 40)]));
-          conv2d_nchw[9] = (conv2d_nchw[9] + (pad_temp_shared[39] * kernel_shared[((((int)threadIdx.x) * 48) + 37)]));
-          conv2d_nchw[9] = (conv2d_nchw[9] + (pad_temp_shared[48] * kernel_shared[((((int)threadIdx.x) * 48) + 40)]));
-          conv2d_nchw[10] = (conv2d_nchw[10] + (pad_temp_shared[40] * kernel_shared[((((int)threadIdx.x) * 48) + 37)]));
-          conv2d_nchw[10] = (conv2d_nchw[10] + (pad_temp_shared[49] * kernel_shared[((((int)threadIdx.x) * 48) + 40)]));
-          conv2d_nchw[11] = (conv2d_nchw[11] + (pad_temp_shared[41] * kernel_shared[((((int)threadIdx.x) * 48) + 37)]));
-          conv2d_nchw[11] = (conv2d_nchw[11] + (pad_temp_shared[50] * kernel_shared[((((int)threadIdx.x) * 48) + 40)]));
-          conv2d_nchw[12] = (conv2d_nchw[12] + (pad_temp_shared[42] * kernel_shared[((((int)threadIdx.x) * 48) + 37)]));
-          conv2d_nchw[12] = (conv2d_nchw[12] + (pad_temp_shared[51] * kernel_shared[((((int)threadIdx.x) * 48) + 40)]));
-          conv2d_nchw[13] = (conv2d_nchw[13] + (pad_temp_shared[43] * kernel_shared[((((int)threadIdx.x) * 48) + 37)]));
-          conv2d_nchw[13] = (conv2d_nchw[13] + (pad_temp_shared[52] * kernel_shared[((((int)threadIdx.x) * 48) + 40)]));
-          conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[38] * kernel_shared[((((int)threadIdx.x) * 48) + 14)]));
-          conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[47] * kernel_shared[((((int)threadIdx.x) * 48) + 17)]));
-          conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[39] * kernel_shared[((((int)threadIdx.x) * 48) + 14)]));
-          conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[48] * kernel_shared[((((int)threadIdx.x) * 48) + 17)]));
-          conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[40] * kernel_shared[((((int)threadIdx.x) * 48) + 14)]));
-          conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[49] * kernel_shared[((((int)threadIdx.x) * 48) + 17)]));
-          conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[41] * kernel_shared[((((int)threadIdx.x) * 48) + 14)]));
-          conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[50] * kernel_shared[((((int)threadIdx.x) * 48) + 17)]));
-          conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[42] * kernel_shared[((((int)threadIdx.x) * 48) + 14)]));
-          conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[51] * kernel_shared[((((int)threadIdx.x) * 48) + 17)]));
-          conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[43] * kernel_shared[((((int)threadIdx.x) * 48) + 14)]));
-          conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[52] * kernel_shared[((((int)threadIdx.x) * 48) + 17)]));
-          conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[44] * kernel_shared[((((int)threadIdx.x) * 48) + 14)]));
-          conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[53] * kernel_shared[((((int)threadIdx.x) * 48) + 17)]));
-          conv2d_nchw[7] = (conv2d_nchw[7] + (pad_temp_shared[38] * kernel_shared[((((int)threadIdx.x) * 48) + 38)]));
-          conv2d_nchw[7] = (conv2d_nchw[7] + (pad_temp_shared[47] * kernel_shared[((((int)threadIdx.x) * 48) + 41)]));
-          conv2d_nchw[8] = (conv2d_nchw[8] + (pad_temp_shared[39] * kernel_shared[((((int)threadIdx.x) * 48) + 38)]));
-          conv2d_nchw[8] = (conv2d_nchw[8] + (pad_temp_shared[48] * kernel_shared[((((int)threadIdx.x) * 48) + 41)]));
-          conv2d_nchw[9] = (conv2d_nchw[9] + (pad_temp_shared[40] * kernel_shared[((((int)threadIdx.x) * 48) + 38)]));
-          conv2d_nchw[9] = (conv2d_nchw[9] + (pad_temp_shared[49] * kernel_shared[((((int)threadIdx.x) * 48) + 41)]));
-          conv2d_nchw[10] = (conv2d_nchw[10] + (pad_temp_shared[41] * kernel_shared[((((int)threadIdx.x) * 48) + 38)]));
-          conv2d_nchw[10] = (conv2d_nchw[10] + (pad_temp_shared[50] * kernel_shared[((((int)threadIdx.x) * 48) + 41)]));
-          conv2d_nchw[11] = (conv2d_nchw[11] + (pad_temp_shared[42] * kernel_shared[((((int)threadIdx.x) * 48) + 38)]));
-          conv2d_nchw[11] = (conv2d_nchw[11] + (pad_temp_shared[51] * kernel_shared[((((int)threadIdx.x) * 48) + 41)]));
-          conv2d_nchw[12] = (conv2d_nchw[12] + (pad_temp_shared[43] * kernel_shared[((((int)threadIdx.x) * 48) + 38)]));
-          conv2d_nchw[12] = (conv2d_nchw[12] + (pad_temp_shared[52] * kernel_shared[((((int)threadIdx.x) * 48) + 41)]));
-          conv2d_nchw[13] = (conv2d_nchw[13] + (pad_temp_shared[44] * kernel_shared[((((int)threadIdx.x) * 48) + 38)]));
-          conv2d_nchw[13] = (conv2d_nchw[13] + (pad_temp_shared[53] * kernel_shared[((((int)threadIdx.x) * 48) + 41)]));
-          conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[54] * kernel_shared[((((int)threadIdx.x) * 48) + 18)]));
-          conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[63] * kernel_shared[((((int)threadIdx.x) * 48) + 21)]));
-          conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[55] * kernel_shared[((((int)threadIdx.x) * 48) + 18)]));
-          conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[64] * kernel_shared[((((int)threadIdx.x) * 48) + 21)]));
-          conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[56] * kernel_shared[((((int)threadIdx.x) * 48) + 18)]));
-          conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[65] * kernel_shared[((((int)threadIdx.x) * 48) + 21)]));
-          conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[57] * kernel_shared[((((int)threadIdx.x) * 48) + 18)]));
-          conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[66] * kernel_shared[((((int)threadIdx.x) * 48) + 21)]));
-          conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[58] * kernel_shared[((((int)threadIdx.x) * 48) + 18)]));
-          conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[67] * kernel_shared[((((int)threadIdx.x) * 48) + 21)]));
-          conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[59] * kernel_shared[((((int)threadIdx.x) * 48) + 18)]));
-          conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[68] * kernel_shared[((((int)threadIdx.x) * 48) + 21)]));
-          conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[60] * kernel_shared[((((int)threadIdx.x) * 48) + 18)]));
-          conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[69] * kernel_shared[((((int)threadIdx.x) * 48) + 21)]));
-          conv2d_nchw[7] = (conv2d_nchw[7] + (pad_temp_shared[54] * kernel_shared[((((int)threadIdx.x) * 48) + 42)]));
-          conv2d_nchw[7] = (conv2d_nchw[7] + (pad_temp_shared[63] * kernel_shared[((((int)threadIdx.x) * 48) + 45)]));
-          conv2d_nchw[8] = (conv2d_nchw[8] + (pad_temp_shared[55] * kernel_shared[((((int)threadIdx.x) * 48) + 42)]));
-          conv2d_nchw[8] = (conv2d_nchw[8] + (pad_temp_shared[64] * kernel_shared[((((int)threadIdx.x) * 48) + 45)]));
-          conv2d_nchw[9] = (conv2d_nchw[9] + (pad_temp_shared[56] * kernel_shared[((((int)threadIdx.x) * 48) + 42)]));
-          conv2d_nchw[9] = (conv2d_nchw[9] + (pad_temp_shared[65] * kernel_shared[((((int)threadIdx.x) * 48) + 45)]));
-          conv2d_nchw[10] = (conv2d_nchw[10] + (pad_temp_shared[57] * kernel_shared[((((int)threadIdx.x) * 48) + 42)]));
-          conv2d_nchw[10] = (conv2d_nchw[10] + (pad_temp_shared[66] * kernel_shared[((((int)threadIdx.x) * 48) + 45)]));
-          conv2d_nchw[11] = (conv2d_nchw[11] + (pad_temp_shared[58] * kernel_shared[((((int)threadIdx.x) * 48) + 42)]));
-          conv2d_nchw[11] = (conv2d_nchw[11] + (pad_temp_shared[67] * kernel_shared[((((int)threadIdx.x) * 48) + 45)]));
-          conv2d_nchw[12] = (conv2d_nchw[12] + (pad_temp_shared[59] * kernel_shared[((((int)threadIdx.x) * 48) + 42)]));
-          conv2d_nchw[12] = (conv2d_nchw[12] + (pad_temp_shared[68] * kernel_shared[((((int)threadIdx.x) * 48) + 45)]));
-          conv2d_nchw[13] = (conv2d_nchw[13] + (pad_temp_shared[60] * kernel_shared[((((int)threadIdx.x) * 48) + 42)]));
-          conv2d_nchw[13] = (conv2d_nchw[13] + (pad_temp_shared[69] * kernel_shared[((((int)threadIdx.x) * 48) + 45)]));
-          conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[55] * kernel_shared[((((int)threadIdx.x) * 48) + 19)]));
-          conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[64] * kernel_shared[((((int)threadIdx.x) * 48) + 22)]));
-          conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[56] * kernel_shared[((((int)threadIdx.x) * 48) + 19)]));
-          conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[65] * kernel_shared[((((int)threadIdx.x) * 48) + 22)]));
-          conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[57] * kernel_shared[((((int)threadIdx.x) * 48) + 19)]));
-          conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[66] * kernel_shared[((((int)threadIdx.x) * 48) + 22)]));
-          conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[58] * kernel_shared[((((int)threadIdx.x) * 48) + 19)]));
-          conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[67] * kernel_shared[((((int)threadIdx.x) * 48) + 22)]));
-          conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[59] * kernel_shared[((((int)threadIdx.x) * 48) + 19)]));
-          conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[68] * kernel_shared[((((int)threadIdx.x) * 48) + 22)]));
-          conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[60] * kernel_shared[((((int)threadIdx.x) * 48) + 19)]));
-          conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[69] * kernel_shared[((((int)threadIdx.x) * 48) + 22)]));
-          conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[61] * kernel_shared[((((int)threadIdx.x) * 48) + 19)]));
-          conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[70] * kernel_shared[((((int)threadIdx.x) * 48) + 22)]));
-          conv2d_nchw[7] = (conv2d_nchw[7] + (pad_temp_shared[55] * kernel_shared[((((int)threadIdx.x) * 48) + 43)]));
-          conv2d_nchw[7] = (conv2d_nchw[7] + (pad_temp_shared[64] * kernel_shared[((((int)threadIdx.x) * 48) + 46)]));
-          conv2d_nchw[8] = (conv2d_nchw[8] + (pad_temp_shared[56] * kernel_shared[((((int)threadIdx.x) * 48) + 43)]));
-          conv2d_nchw[8] = (conv2d_nchw[8] + (pad_temp_shared[65] * kernel_shared[((((int)threadIdx.x) * 48) + 46)]));
-          conv2d_nchw[9] = (conv2d_nchw[9] + (pad_temp_shared[57] * kernel_shared[((((int)threadIdx.x) * 48) + 43)]));
-          conv2d_nchw[9] = (conv2d_nchw[9] + (pad_temp_shared[66] * kernel_shared[((((int)threadIdx.x) * 48) + 46)]));
-          conv2d_nchw[10] = (conv2d_nchw[10] + (pad_temp_shared[58] * kernel_shared[((((int)threadIdx.x) * 48) + 43)]));
-          conv2d_nchw[10] = (conv2d_nchw[10] + (pad_temp_shared[67] * kernel_shared[((((int)threadIdx.x) * 48) + 46)]));
-          conv2d_nchw[11] = (conv2d_nchw[11] + (pad_temp_shared[59] * kernel_shared[((((int)threadIdx.x) * 48) + 43)]));
-          conv2d_nchw[11] = (conv2d_nchw[11] + (pad_temp_shared[68] * kernel_shared[((((int)threadIdx.x) * 48) + 46)]));
-          conv2d_nchw[12] = (conv2d_nchw[12] + (pad_temp_shared[60] * kernel_shared[((((int)threadIdx.x) * 48) + 43)]));
-          conv2d_nchw[12] = (conv2d_nchw[12] + (pad_temp_shared[69] * kernel_shared[((((int)threadIdx.x) * 48) + 46)]));
-          conv2d_nchw[13] = (conv2d_nchw[13] + (pad_temp_shared[61] * kernel_shared[((((int)threadIdx.x) * 48) + 43)]));
-          conv2d_nchw[13] = (conv2d_nchw[13] + (pad_temp_shared[70] * kernel_shared[((((int)threadIdx.x) * 48) + 46)]));
-          conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[56] * kernel_shared[((((int)threadIdx.x) * 48) + 20)]));
-          conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[65] * kernel_shared[((((int)threadIdx.x) * 48) + 23)]));
-          conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[57] * kernel_shared[((((int)threadIdx.x) * 48) + 20)]));
-          conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[66] * kernel_shared[((((int)threadIdx.x) * 48) + 23)]));
-          conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[58] * kernel_shared[((((int)threadIdx.x) * 48) + 20)]));
-          conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[67] * kernel_shared[((((int)threadIdx.x) * 48) + 23)]));
-          conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[59] * kernel_shared[((((int)threadIdx.x) * 48) + 20)]));
-          conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[68] * kernel_shared[((((int)threadIdx.x) * 48) + 23)]));
-          conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[60] * kernel_shared[((((int)threadIdx.x) * 48) + 20)]));
-          conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[69] * kernel_shared[((((int)threadIdx.x) * 48) + 23)]));
-          conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[61] * kernel_shared[((((int)threadIdx.x) * 48) + 20)]));
-          conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[70] * kernel_shared[((((int)threadIdx.x) * 48) + 23)]));
-          conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[62] * kernel_shared[((((int)threadIdx.x) * 48) + 20)]));
-          conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[71] * kernel_shared[((((int)threadIdx.x) * 48) + 23)]));
-          conv2d_nchw[7] = (conv2d_nchw[7] + (pad_temp_shared[56] * kernel_shared[((((int)threadIdx.x) * 48) + 44)]));
-          conv2d_nchw[7] = (conv2d_nchw[7] + (pad_temp_shared[65] * kernel_shared[((((int)threadIdx.x) * 48) + 47)]));
-          conv2d_nchw[8] = (conv2d_nchw[8] + (pad_temp_shared[57] * kernel_shared[((((int)threadIdx.x) * 48) + 44)]));
-          conv2d_nchw[8] = (conv2d_nchw[8] + (pad_temp_shared[66] * kernel_shared[((((int)threadIdx.x) * 48) + 47)]));
-          conv2d_nchw[9] = (conv2d_nchw[9] + (pad_temp_shared[58] * kernel_shared[((((int)threadIdx.x) * 48) + 44)]));
-          conv2d_nchw[9] = (conv2d_nchw[9] + (pad_temp_shared[67] * kernel_shared[((((int)threadIdx.x) * 48) + 47)]));
-          conv2d_nchw[10] = (conv2d_nchw[10] + (pad_temp_shared[59] * kernel_shared[((((int)threadIdx.x) * 48) + 44)]));
-          conv2d_nchw[10] = (conv2d_nchw[10] + (pad_temp_shared[68] * kernel_shared[((((int)threadIdx.x) * 48) + 47)]));
-          conv2d_nchw[11] = (conv2d_nchw[11] + (pad_temp_shared[60] * kernel_shared[((((int)threadIdx.x) * 48) + 44)]));
-          conv2d_nchw[11] = (conv2d_nchw[11] + (pad_temp_shared[69] * kernel_shared[((((int)threadIdx.x) * 48) + 47)]));
-          conv2d_nchw[12] = (conv2d_nchw[12] + (pad_temp_shared[61] * kernel_shared[((((int)threadIdx.x) * 48) + 44)]));
-          conv2d_nchw[12] = (conv2d_nchw[12] + (pad_temp_shared[70] * kernel_shared[((((int)threadIdx.x) * 48) + 47)]));
-          conv2d_nchw[13] = (conv2d_nchw[13] + (pad_temp_shared[62] * kernel_shared[((((int)threadIdx.x) * 48) + 44)]));
-          conv2d_nchw[13] = (conv2d_nchw[13] + (pad_temp_shared[71] * kernel_shared[((((int)threadIdx.x) * 48) + 47)]));
-        }
-      }
-      for (int i1_inner = 0; i1_inner < 2; ++i1_inner) {
-        for (int i3_inner = 0; i3_inner < 7; ++i3_inner) {
-          compute[((((((((int)blockIdx.x) / 7) * 6272) + (((int)threadIdx.x) * 98)) + (i1_inner * 49)) + ((((int)blockIdx.x) % 7) * 7)) + i3_inner)] = max((conv2d_nchw[((i1_inner * 7) + i3_inner)] + bias[((((((int)blockIdx.x) / 7) * 128) + (((int)threadIdx.x) * 2)) + i1_inner)]), 0.000000e+00f);
+        __syncthreads();
+        pad_temp_shared[((int)threadIdx.x)] = (((((1 <= (((((int)threadIdx.x) % 27) / 9) + (((int)blockIdx.x) % 7))) && ((((((int)threadIdx.x) % 27) / 9) + (((int)blockIdx.x) % 7)) < 8)) && (1 <= (((int)threadIdx.x) % 9))) && ((((int)threadIdx.x) % 9) < 8)) ? data[((((((rc_outer_outer * 392) + ((((int)threadIdx.x) / 27) * 49)) + (((((int)threadIdx.x) % 27) / 9) * 7)) + ((((int)blockIdx.x) % 7) * 7)) + (((int)threadIdx.x) % 9)) - 8)] : 0.000000e+00f);
+        pad_temp_shared[(((int)threadIdx.x) + 64)] = (((((1 <= ((((((int)threadIdx.x) + 10) % 27) / 9) + (((int)blockIdx.x) % 7))) && (((((((int)threadIdx.x) + 10) % 27) / 9) + (((int)blockIdx.x) % 7)) < 8)) && (1 <= ((((int)threadIdx.x) + 1) % 9))) && (((((int)threadIdx.x) + 1) % 9) < 8)) ? data[((((((rc_outer_outer * 392) + (((((int)threadIdx.x) + 64) / 27) * 49)) + ((((((int)threadIdx.x) + 10) % 27) / 9) * 7)) + ((((int)blockIdx.x) % 7) * 7)) + ((((int)threadIdx.x) + 1) % 9)) - 8)] :  [...]
+        pad_temp_shared[(((int)threadIdx.x) + 128)] = (((((1 <= ((((((int)threadIdx.x) + 20) % 27) / 9) + (((int)blockIdx.x) % 7))) && (((((((int)threadIdx.x) + 20) % 27) / 9) + (((int)blockIdx.x) % 7)) < 8)) && (1 <= ((((int)threadIdx.x) + 2) % 9))) && (((((int)threadIdx.x) + 2) % 9) < 8)) ? data[((((((rc_outer_outer * 392) + (((((int)threadIdx.x) + 128) / 27) * 49)) + ((((((int)threadIdx.x) + 20) % 27) / 9) * 7)) + ((((int)blockIdx.x) % 7) * 7)) + ((((int)threadIdx.x) + 2) % 9)) - 8)]  [...]
+        if (((int)threadIdx.x) < 24) {
+          pad_temp_shared[(((int)threadIdx.x) + 192)] = (((((1 <= (((((int)threadIdx.x) + 3) / 9) + (((int)blockIdx.x) % 7))) && ((((((int)threadIdx.x) + 3) / 9) + (((int)blockIdx.x) % 7)) < 8)) && (1 <= ((((int)threadIdx.x) + 3) % 9))) && (((((int)threadIdx.x) + 3) % 9) < 8)) ? data[((((((rc_outer_outer * 392) + (((((int)threadIdx.x) + 192) / 27) * 49)) + (((((int)threadIdx.x) + 3) / 9) * 7)) + ((((int)blockIdx.x) % 7) * 7)) + ((((int)threadIdx.x) + 3) % 9)) - 8)] : 0.000000e+00f);
         }
+        kernel_shared[((int)threadIdx.x)] = kernel[((((((int)blockIdx.x) / 7) * 589824) + (rc_outer_outer * 72)) + ((int)threadIdx.x))];
+        kernel_shared[(((int)threadIdx.x) + 64)] = kernel[((((((((int)blockIdx.x) / 7) * 589824) + (((((int)threadIdx.x) + 64) / 72) * 4608)) + (rc_outer_outer * 72)) + ((((((int)threadIdx.x) + 64) % 72) / 3) * 3)) + ((((int)threadIdx.x) + 1) % 3))];
+        kernel_shared[(((int)threadIdx.x) + 128)] = kernel[((((((((int)blockIdx.x) / 7) * 589824) + (((((int)threadIdx.x) + 128) / 72) * 4608)) + (rc_outer_outer * 72)) + ((((((int)threadIdx.x) + 56) % 72) / 3) * 3)) + ((((int)threadIdx.x) + 2) % 3))];
+        kernel_shared[(((int)threadIdx.x) + 192)] = kernel[((((((((int)blockIdx.x) / 7) * 589824) + (((((int)threadIdx.x) + 192) / 72) * 4608)) + (rc_outer_outer * 72)) + ((((((int)threadIdx.x) / 3) + 16) % 24) * 3)) + (((int)threadIdx.x) % 3))];
+        kernel_shared[(((int)threadIdx.x) + 256)] = kernel[((((((((int)blockIdx.x) / 7) * 589824) + (((((int)threadIdx.x) + 256) / 72) * 4608)) + (rc_outer_outer * 72)) + ((((((int)threadIdx.x) + 40) % 72) / 3) * 3)) + ((((int)threadIdx.x) + 1) % 3))];
+        kernel_shared[(((int)threadIdx.x) + 320)] = kernel[((((((((int)blockIdx.x) / 7) * 589824) + (((((int)threadIdx.x) + 320) / 72) * 4608)) + (rc_outer_outer * 72)) + ((((((int)threadIdx.x) + 32) % 72) / 3) * 3)) + ((((int)threadIdx.x) + 2) % 3))];
+        kernel_shared[(((int)threadIdx.x) + 384)] = kernel[((((((((int)blockIdx.x) / 7) * 589824) + (((((int)threadIdx.x) + 384) / 72) * 4608)) + (rc_outer_outer * 72)) + ((((((int)threadIdx.x) / 3) + 8) % 24) * 3)) + (((int)threadIdx.x) % 3))];
+        kernel_shared[(((int)threadIdx.x) + 448)] = kernel[((((((((int)blockIdx.x) / 7) * 589824) + (((((int)threadIdx.x) + 448) / 72) * 4608)) + (rc_outer_outer * 72)) + ((((((int)threadIdx.x) + 16) % 72) / 3) * 3)) + ((((int)threadIdx.x) + 1) % 3))];
+        kernel_shared[(((int)threadIdx.x) + 512)] = kernel[((((((((int)blockIdx.x) / 7) * 589824) + (((((int)threadIdx.x) + 512) / 72) * 4608)) + (rc_outer_outer * 72)) + (((((int)threadIdx.x) + 8) / 3) * 3)) + ((((int)threadIdx.x) + 2) % 3))];
+        kernel_shared[(((int)threadIdx.x) + 576)] = kernel[(((((((int)blockIdx.x) / 7) * 589824) + (rc_outer_outer * 72)) + ((int)threadIdx.x)) + 36864)];
+        kernel_shared[(((int)threadIdx.x) + 640)] = kernel[((((((((int)blockIdx.x) / 7) * 589824) + (((((int)threadIdx.x) + 640) / 72) * 4608)) + (rc_outer_outer * 72)) + ((((((int)threadIdx.x) + 64) % 72) / 3) * 3)) + ((((int)threadIdx.x) + 1) % 3))];
+        kernel_shared[(((int)threadIdx.x) + 704)] = kernel[((((((((int)blockIdx.x) / 7) * 589824) + (((((int)threadIdx.x) + 704) / 72) * 4608)) + (rc_outer_outer * 72)) + ((((((int)threadIdx.x) + 56) % 72) / 3) * 3)) + ((((int)threadIdx.x) + 2) % 3))];
+        kernel_shared[(((int)threadIdx.x) + 768)] = kernel[((((((((int)blockIdx.x) / 7) * 589824) + (((((int)threadIdx.x) + 768) / 72) * 4608)) + (rc_outer_outer * 72)) + ((((((int)threadIdx.x) / 3) + 16) % 24) * 3)) + (((int)threadIdx.x) % 3))];
+        kernel_shared[(((int)threadIdx.x) + 832)] = kernel[((((((((int)blockIdx.x) / 7) * 589824) + (((((int)threadIdx.x) + 832) / 72) * 4608)) + (rc_outer_outer * 72)) + ((((((int)threadIdx.x) + 40) % 72) / 3) * 3)) + ((((int)threadIdx.x) + 1) % 3))];
+        kernel_shared[(((int)threadIdx.x) + 896)] = kernel[((((((((int)blockIdx.x) / 7) * 589824) + (((((int)threadIdx.x) + 896) / 72) * 4608)) + (rc_outer_outer * 72)) + ((((((int)threadIdx.x) + 32) % 72) / 3) * 3)) + ((((int)threadIdx.x) + 2) % 3))];
+        kernel_shared[(((int)threadIdx.x) + 960)] = kernel[((((((((int)blockIdx.x) / 7) * 589824) + (((((int)threadIdx.x) + 960) / 72) * 4608)) + (rc_outer_outer * 72)) + ((((((int)threadIdx.x) / 3) + 8) % 24) * 3)) + (((int)threadIdx.x) % 3))];
+        kernel_shared[(((int)threadIdx.x) + 1024)] = kernel[((((((((int)blockIdx.x) / 7) * 589824) + (((((int)threadIdx.x) + 1024) / 72) * 4608)) + (rc_outer_outer * 72)) + ((((((int)threadIdx.x) + 16) % 72) / 3) * 3)) + ((((int)threadIdx.x) + 1) % 3))];
+        kernel_shared[(((int)threadIdx.x) + 1088)] = kernel[((((((((int)blockIdx.x) / 7) * 589824) + (((((int)threadIdx.x) + 1088) / 72) * 4608)) + (rc_outer_outer * 72)) + (((((int)threadIdx.x) + 8) / 3) * 3)) + ((((int)threadIdx.x) + 2) % 3))];
+        kernel_shared[(((int)threadIdx.x) + 1152)] = kernel[(((((((int)blockIdx.x) / 7) * 589824) + (rc_outer_outer * 72)) + ((int)threadIdx.x)) + 73728)];
+        kernel_shared[(((int)threadIdx.x) + 1216)] = kernel[((((((((int)blockIdx.x) / 7) * 589824) + (((((int)threadIdx.x) + 1216) / 72) * 4608)) + (rc_outer_outer * 72)) + ((((((int)threadIdx.x) + 64) % 72) / 3) * 3)) + ((((int)threadIdx.x) + 1) % 3))];
+        kernel_shared[(((int)threadIdx.x) + 1280)] = kernel[((((((((int)blockIdx.x) / 7) * 589824) + (((((int)threadIdx.x) + 1280) / 72) * 4608)) + (rc_outer_outer * 72)) + ((((((int)threadIdx.x) + 56) % 72) / 3) * 3)) + ((((int)threadIdx.x) + 2) % 3))];
+        kernel_shared[(((int)threadIdx.x) + 1344)] = kernel[((((((((int)blockIdx.x) / 7) * 589824) + (((((int)threadIdx.x) + 1344) / 72) * 4608)) + (rc_outer_outer * 72)) + ((((((int)threadIdx.x) / 3) + 16) % 24) * 3)) + (((int)threadIdx.x) % 3))];
+        kernel_shared[(((int)threadIdx.x) + 1408)] = kernel[((((((((int)blockIdx.x) / 7) * 589824) + (((((int)threadIdx.x) + 1408) / 72) * 4608)) + (rc_outer_outer * 72)) + ((((((int)threadIdx.x) + 40) % 72) / 3) * 3)) + ((((int)threadIdx.x) + 1) % 3))];
+        kernel_shared[(((int)threadIdx.x) + 1472)] = kernel[((((((((int)blockIdx.x) / 7) * 589824) + (((((int)threadIdx.x) + 1472) / 72) * 4608)) + (rc_outer_outer * 72)) + ((((((int)threadIdx.x) + 32) % 72) / 3) * 3)) + ((((int)threadIdx.x) + 2) % 3))];
+        kernel_shared[(((int)threadIdx.x) + 1536)] = kernel[((((((((int)blockIdx.x) / 7) * 589824) + (((((int)threadIdx.x) + 1536) / 72) * 4608)) + (rc_outer_outer * 72)) + ((((((int)threadIdx.x) / 3) + 8) % 24) * 3)) + (((int)threadIdx.x) % 3))];
+        kernel_shared[(((int)threadIdx.x) + 1600)] = kernel[((((((((int)blockIdx.x) / 7) * 589824) + (((((int)threadIdx.x) + 1600) / 72) * 4608)) + (rc_outer_outer * 72)) + ((((((int)threadIdx.x) + 16) % 72) / 3) * 3)) + ((((int)threadIdx.x) + 1) % 3))];
+        kernel_shared[(((int)threadIdx.x) + 1664)] = kernel[((((((((int)blockIdx.x) / 7) * 589824) + (((((int)threadIdx.x) + 1664) / 72) * 4608)) + (rc_outer_outer * 72)) + (((((int)threadIdx.x) + 8) / 3) * 3)) + ((((int)threadIdx.x) + 2) % 3))];
+        kernel_shared[(((int)threadIdx.x) + 1728)] = kernel[(((((((int)blockIdx.x) / 7) * 589824) + (rc_outer_outer * 72)) + ((int)threadIdx.x)) + 110592)];
+        kernel_shared[(((int)threadIdx.x) + 1792)] = kernel[((((((((int)blockIdx.x) / 7) * 589824) + (((((int)threadIdx.x) + 1792) / 72) * 4608)) + (rc_outer_outer * 72)) + ((((((int)threadIdx.x) + 64) % 72) / 3) * 3)) + ((((int)threadIdx.x) + 1) % 3))];
+        kernel_shared[(((int)threadIdx.x) + 1856)] = kernel[((((((((int)blockIdx.x) / 7) * 589824) + (((((int)threadIdx.x) + 1856) / 72) * 4608)) + (rc_outer_outer * 72)) + ((((((int)threadIdx.x) + 56) % 72) / 3) * 3)) + ((((int)threadIdx.x) + 2) % 3))];
+        kernel_shared[(((int)threadIdx.x) + 1920)] = kernel[((((((((int)blockIdx.x) / 7) * 589824) + (((((int)threadIdx.x) + 1920) / 72) * 4608)) + (rc_outer_outer * 72)) + ((((((int)threadIdx.x) / 3) + 16) % 24) * 3)) + (((int)threadIdx.x) % 3))];
+        kernel_shared[(((int)threadIdx.x) + 1984)] = kernel[((((((((int)blockIdx.x) / 7) * 589824) + (((((int)threadIdx.x) + 1984) / 72) * 4608)) + (rc_outer_outer * 72)) + ((((((int)threadIdx.x) + 40) % 72) / 3) * 3)) + ((((int)threadIdx.x) + 1) % 3))];
+        kernel_shared[(((int)threadIdx.x) + 2048)] = kernel[((((((((int)blockIdx.x) / 7) * 589824) + (((((int)threadIdx.x) + 2048) / 72) * 4608)) + (rc_outer_outer * 72)) + ((((((int)threadIdx.x) + 32) % 72) / 3) * 3)) + ((((int)threadIdx.x) + 2) % 3))];
+        kernel_shared[(((int)threadIdx.x) + 2112)] = kernel[((((((((int)blockIdx.x) / 7) * 589824) + (((((int)threadIdx.x) + 2112) / 72) * 4608)) + (rc_outer_outer * 72)) + ((((((int)threadIdx.x) / 3) + 8) % 24) * 3)) + (((int)threadIdx.x) % 3))];
+        kernel_shared[(((int)threadIdx.x) + 2176)] = kernel[((((((((int)blockIdx.x) / 7) * 589824) + (((((int)threadIdx.x) + 2176) / 72) * 4608)) + (rc_outer_outer * 72)) + ((((((int)threadIdx.x) + 16) % 72) / 3) * 3)) + ((((int)threadIdx.x) + 1) % 3))];
+        kernel_shared[(((int)threadIdx.x) + 2240)] = kernel[((((((((int)blockIdx.x) / 7) * 589824) + (((((int)threadIdx.x) + 2240) / 72) * 4608)) + (rc_outer_outer * 72)) + (((((int)threadIdx.x) + 8) / 3) * 3)) + ((((int)threadIdx.x) + 2) % 3))];
+        kernel_shared[(((int)threadIdx.x) + 2304)] = kernel[(((((((int)blockIdx.x) / 7) * 589824) + (rc_outer_outer * 72)) + ((int)threadIdx.x)) + 147456)];
+        kernel_shared[(((int)threadIdx.x) + 2368)] = kernel[((((((((int)blockIdx.x) / 7) * 589824) + (((((int)threadIdx.x) + 2368) / 72) * 4608)) + (rc_outer_outer * 72)) + ((((((int)threadIdx.x) + 64) % 72) / 3) * 3)) + ((((int)threadIdx.x) + 1) % 3))];
+        kernel_shared[(((int)threadIdx.x) + 2432)] = kernel[((((((((int)blockIdx.x) / 7) * 589824) + (((((int)threadIdx.x) + 2432) / 72) * 4608)) + (rc_outer_outer * 72)) + ((((((int)threadIdx.x) + 56) % 72) / 3) * 3)) + ((((int)threadIdx.x) + 2) % 3))];
+        kernel_shared[(((int)threadIdx.x) + 2496)] = kernel[((((((((int)blockIdx.x) / 7) * 589824) + (((((int)threadIdx.x) + 2496) / 72) * 4608)) + (rc_outer_outer * 72)) + ((((((int)threadIdx.x) / 3) + 16) % 24) * 3)) + (((int)threadIdx.x) % 3))];
+        kernel_shared[(((int)threadIdx.x) + 2560)] = kernel[((((((((int)blockIdx.x) / 7) * 589824) + (((((int)threadIdx.x) + 2560) / 72) * 4608)) + (rc_outer_outer * 72)) + ((((((int)threadIdx.x) + 40) % 72) / 3) * 3)) + ((((int)threadIdx.x) + 1) % 3))];
+        kernel_shared[(((int)threadIdx.x) + 2624)] = kernel[((((((((int)blockIdx.x) / 7) * 589824) + (((((int)threadIdx.x) + 2624) / 72) * 4608)) + (rc_outer_outer * 72)) + ((((((int)threadIdx.x) + 32) % 72) / 3) * 3)) + ((((int)threadIdx.x) + 2) % 3))];
+        kernel_shared[(((int)threadIdx.x) + 2688)] = kernel[((((((((int)blockIdx.x) / 7) * 589824) + (((((int)threadIdx.x) + 2688) / 72) * 4608)) + (rc_outer_outer * 72)) + ((((((int)threadIdx.x) / 3) + 8) % 24) * 3)) + (((int)threadIdx.x) % 3))];
+        kernel_shared[(((int)threadIdx.x) + 2752)] = kernel[((((((((int)blockIdx.x) / 7) * 589824) + (((((int)threadIdx.x) + 2752) / 72) * 4608)) + (rc_outer_outer * 72)) + ((((((int)threadIdx.x) + 16) % 72) / 3) * 3)) + ((((int)threadIdx.x) + 1) % 3))];
+        kernel_shared[(((int)threadIdx.x) + 2816)] = kernel[((((((((int)blockIdx.x) / 7) * 589824) + (((((int)threadIdx.x) + 2816) / 72) * 4608)) + (rc_outer_outer * 72)) + (((((int)threadIdx.x) + 8) / 3) * 3)) + ((((int)threadIdx.x) + 2) % 3))];
+        kernel_shared[(((int)threadIdx.x) + 2880)] = kernel[(((((((int)blockIdx.x) / 7) * 589824) + (rc_outer_outer * 72)) + ((int)threadIdx.x)) + 184320)];
+        kernel_shared[(((int)threadIdx.x) + 2944)] = kernel[((((((((int)blockIdx.x) / 7) * 589824) + (((((int)threadIdx.x) + 2944) / 72) * 4608)) + (rc_outer_outer * 72)) + ((((((int)threadIdx.x) + 64) % 72) / 3) * 3)) + ((((int)threadIdx.x) + 1) % 3))];
+        kernel_shared[(((int)threadIdx.x) + 3008)] = kernel[((((((((int)blockIdx.x) / 7) * 589824) + (((((int)threadIdx.x) + 3008) / 72) * 4608)) + (rc_outer_outer * 72)) + ((((((int)threadIdx.x) + 56) % 72) / 3) * 3)) + ((((int)threadIdx.x) + 2) % 3))];
+        kernel_shared[(((int)threadIdx.x) + 3072)] = kernel[((((((((int)blockIdx.x) / 7) * 589824) + (((((int)threadIdx.x) + 3072) / 72) * 4608)) + (rc_outer_outer * 72)) + ((((((int)threadIdx.x) / 3) + 16) % 24) * 3)) + (((int)threadIdx.x) % 3))];
+        kernel_shared[(((int)threadIdx.x) + 3136)] = kernel[((((((((int)blockIdx.x) / 7) * 589824) + (((((int)threadIdx.x) + 3136) / 72) * 4608)) + (rc_outer_outer * 72)) + ((((((int)threadIdx.x) + 40) % 72) / 3) * 3)) + ((((int)threadIdx.x) + 1) % 3))];
+        kernel_shared[(((int)threadIdx.x) + 3200)] = kernel[((((((((int)blockIdx.x) / 7) * 589824) + (((((int)threadIdx.x) + 3200) / 72) * 4608)) + (rc_outer_outer * 72)) + ((((((int)threadIdx.x) + 32) % 72) / 3) * 3)) + ((((int)threadIdx.x) + 2) % 3))];
+        kernel_shared[(((int)threadIdx.x) + 3264)] = kernel[((((((((int)blockIdx.x) / 7) * 589824) + (((((int)threadIdx.x) + 3264) / 72) * 4608)) + (rc_outer_outer * 72)) + ((((((int)threadIdx.x) / 3) + 8) % 24) * 3)) + (((int)threadIdx.x) % 3))];
+        kernel_shared[(((int)threadIdx.x) + 3328)] = kernel[((((((((int)blockIdx.x) / 7) * 589824) + (((((int)threadIdx.x) + 3328) / 72) * 4608)) + (rc_outer_outer * 72)) + ((((((int)threadIdx.x) + 16) % 72) / 3) * 3)) + ((((int)threadIdx.x) + 1) % 3))];
+        kernel_shared[(((int)threadIdx.x) + 3392)] = kernel[((((((((int)blockIdx.x) / 7) * 589824) + (((((int)threadIdx.x) + 3392) / 72) * 4608)) + (rc_outer_outer * 72)) + (((((int)threadIdx.x) + 8) / 3) * 3)) + ((((int)threadIdx.x) + 2) % 3))];
+        kernel_shared[(((int)threadIdx.x) + 3456)] = kernel[(((((((int)blockIdx.x) / 7) * 589824) + (rc_outer_outer * 72)) + ((int)threadIdx.x)) + 221184)];
+        kernel_shared[(((int)threadIdx.x) + 3520)] = kernel[((((((((int)blockIdx.x) / 7) * 589824) + (((((int)threadIdx.x) + 3520) / 72) * 4608)) + (rc_outer_outer * 72)) + ((((((int)threadIdx.x) + 64) % 72) / 3) * 3)) + ((((int)threadIdx.x) + 1) % 3))];
+        kernel_shared[(((int)threadIdx.x) + 3584)] = kernel[((((((((int)blockIdx.x) / 7) * 589824) + (((((int)threadIdx.x) + 3584) / 72) * 4608)) + (rc_outer_outer * 72)) + ((((((int)threadIdx.x) + 56) % 72) / 3) * 3)) + ((((int)threadIdx.x) + 2) % 3))];
+        kernel_shared[(((int)threadIdx.x) + 3648)] = kernel[((((((((int)blockIdx.x) / 7) * 589824) + (((((int)threadIdx.x) + 3648) / 72) * 4608)) + (rc_outer_outer * 72)) + ((((((int)threadIdx.x) / 3) + 16) % 24) * 3)) + (((int)threadIdx.x) % 3))];
+        kernel_shared[(((int)threadIdx.x) + 3712)] = kernel[((((((((int)blockIdx.x) / 7) * 589824) + (((((int)threadIdx.x) + 3712) / 72) * 4608)) + (rc_outer_outer * 72)) + ((((((int)threadIdx.x) + 40) % 72) / 3) * 3)) + ((((int)threadIdx.x) + 1) % 3))];
+        kernel_shared[(((int)threadIdx.x) + 3776)] = kernel[((((((((int)blockIdx.x) / 7) * 589824) + (((((int)threadIdx.x) + 3776) / 72) * 4608)) + (rc_outer_outer * 72)) + ((((((int)threadIdx.x) + 32) % 72) / 3) * 3)) + ((((int)threadIdx.x) + 2) % 3))];
+        kernel_shared[(((int)threadIdx.x) + 3840)] = kernel[((((((((int)blockIdx.x) / 7) * 589824) + (((((int)threadIdx.x) + 3840) / 72) * 4608)) + (rc_outer_outer * 72)) + ((((((int)threadIdx.x) / 3) + 8) % 24) * 3)) + (((int)threadIdx.x) % 3))];
+        kernel_shared[(((int)threadIdx.x) + 3904)] = kernel[((((((((int)blockIdx.x) / 7) * 589824) + (((((int)threadIdx.x) + 3904) / 72) * 4608)) + (rc_outer_outer * 72)) + ((((((int)threadIdx.x) + 16) % 72) / 3) * 3)) + ((((int)threadIdx.x) + 1) % 3))];
+        kernel_shared[(((int)threadIdx.x) + 3968)] = kernel[((((((((int)blockIdx.x) / 7) * 589824) + (((((int)threadIdx.x) + 3968) / 72) * 4608)) + (rc_outer_outer * 72)) + (((((int)threadIdx.x) + 8) / 3) * 3)) + ((((int)threadIdx.x) + 2) % 3))];
+        kernel_shared[(((int)threadIdx.x) + 4032)] = kernel[(((((((int)blockIdx.x) / 7) * 589824) + (rc_outer_outer * 72)) + ((int)threadIdx.x)) + 258048)];
+        kernel_shared[(((int)threadIdx.x) + 4096)] = kernel[((((((((int)blockIdx.x) / 7) * 589824) + (((((int)threadIdx.x) + 4096) / 72) * 4608)) + (rc_outer_outer * 72)) + ((((((int)threadIdx.x) + 64) % 72) / 3) * 3)) + ((((int)threadIdx.x) + 1) % 3))];
+        kernel_shared[(((int)threadIdx.x) + 4160)] = kernel[((((((((int)blockIdx.x) / 7) * 589824) + (((((int)threadIdx.x) + 4160) / 72) * 4608)) + (rc_outer_outer * 72)) + ((((((int)threadIdx.x) + 56) % 72) / 3) * 3)) + ((((int)threadIdx.x) + 2) % 3))];
+        kernel_shared[(((int)threadIdx.x) + 4224)] = kernel[((((((((int)blockIdx.x) / 7) * 589824) + (((((int)threadIdx.x) + 4224) / 72) * 4608)) + (rc_outer_outer * 72)) + ((((((int)threadIdx.x) / 3) + 16) % 24) * 3)) + (((int)threadIdx.x) % 3))];
+        kernel_shared[(((int)threadIdx.x) + 4288)] = kernel[((((((((int)blockIdx.x) / 7) * 589824) + (((((int)threadIdx.x) + 4288) / 72) * 4608)) + (rc_outer_outer * 72)) + ((((((int)threadIdx.x) + 40) % 72) / 3) * 3)) + ((((int)threadIdx.x) + 1) % 3))];
+        kernel_shared[(((int)threadIdx.x) + 4352)] = kernel[((((((((int)blockIdx.x) / 7) * 589824) + (((((int)threadIdx.x) + 4352) / 72) * 4608)) + (rc_outer_outer * 72)) + ((((((int)threadIdx.x) + 32) % 72) / 3) * 3)) + ((((int)threadIdx.x) + 2) % 3))];
+        kernel_shared[(((int)threadIdx.x) + 4416)] = kernel[((((((((int)blockIdx.x) / 7) * 589824) + (((((int)threadIdx.x) + 4416) / 72) * 4608)) + (rc_outer_outer * 72)) + ((((((int)threadIdx.x) / 3) + 8) % 24) * 3)) + (((int)threadIdx.x) % 3))];
+        kernel_shared[(((int)threadIdx.x) + 4480)] = kernel[((((((((int)blockIdx.x) / 7) * 589824) + (((((int)threadIdx.x) + 4480) / 72) * 4608)) + (rc_outer_outer * 72)) + ((((((int)threadIdx.x) + 16) % 72) / 3) * 3)) + ((((int)threadIdx.x) + 1) % 3))];
+        kernel_shared[(((int)threadIdx.x) + 4544)] = kernel[((((((((int)blockIdx.x) / 7) * 589824) + (((((int)threadIdx.x) + 4544) / 72) * 4608)) + (rc_outer_outer * 72)) + (((((int)threadIdx.x) + 8) / 3) * 3)) + ((((int)threadIdx.x) + 2) % 3))];
+        kernel_shared[(((int)threadIdx.x) + 4608)] = kernel[(((((((int)blockIdx.x) / 7) * 589824) + (rc_outer_outer * 72)) + ((int)threadIdx.x)) + 294912)];
+        kernel_shared[(((int)threadIdx.x) + 4672)] = kernel[((((((((int)blockIdx.x) / 7) * 589824) + (((((int)threadIdx.x) + 4672) / 72) * 4608)) + (rc_outer_outer * 72)) + ((((((int)threadIdx.x) + 64) % 72) / 3) * 3)) + ((((int)threadIdx.x) + 1) % 3))];
+        kernel_shared[(((int)threadIdx.x) + 4736)] = kernel[((((((((int)blockIdx.x) / 7) * 589824) + (((((int)threadIdx.x) + 4736) / 72) * 4608)) + (rc_outer_outer * 72)) + ((((((int)threadIdx.x) + 56) % 72) / 3) * 3)) + ((((int)threadIdx.x) + 2) % 3))];
+        kernel_shared[(((int)threadIdx.x) + 4800)] = kernel[((((((((int)blockIdx.x) / 7) * 589824) + (((((int)threadIdx.x) + 4800) / 72) * 4608)) + (rc_outer_outer * 72)) + ((((((int)threadIdx.x) / 3) + 16) % 24) * 3)) + (((int)threadIdx.x) % 3))];
+        kernel_shared[(((int)threadIdx.x) + 4864)] = kernel[((((((((int)blockIdx.x) / 7) * 589824) + (((((int)threadIdx.x) + 4864) / 72) * 4608)) + (rc_outer_outer * 72)) + ((((((int)threadIdx.x) + 40) % 72) / 3) * 3)) + ((((int)threadIdx.x) + 1) % 3))];
+        kernel_shared[(((int)threadIdx.x) + 4928)] = kernel[((((((((int)blockIdx.x) / 7) * 589824) + (((((int)threadIdx.x) + 4928) / 72) * 4608)) + (rc_outer_outer * 72)) + ((((((int)threadIdx.x) + 32) % 72) / 3) * 3)) + ((((int)threadIdx.x) + 2) % 3))];
+        kernel_shared[(((int)threadIdx.x) + 4992)] = kernel[((((((((int)blockIdx.x) / 7) * 589824) + (((((int)threadIdx.x) + 4992) / 72) * 4608)) + (rc_outer_outer * 72)) + ((((((int)threadIdx.x) / 3) + 8) % 24) * 3)) + (((int)threadIdx.x) % 3))];
+        kernel_shared[(((int)threadIdx.x) + 5056)] = kernel[((((((((int)blockIdx.x) / 7) * 589824) + (((((int)threadIdx.x) + 5056) / 72) * 4608)) + (rc_outer_outer * 72)) + ((((((int)threadIdx.x) + 16) % 72) / 3) * 3)) + ((((int)threadIdx.x) + 1) % 3))];
+        kernel_shared[(((int)threadIdx.x) + 5120)] = kernel[((((((((int)blockIdx.x) / 7) * 589824) + (((((int)threadIdx.x) + 5120) / 72) * 4608)) + (rc_outer_outer * 72)) + (((((int)threadIdx.x) + 8) / 3) * 3)) + ((((int)threadIdx.x) + 2) % 3))];
+        kernel_shared[(((int)threadIdx.x) + 5184)] = kernel[(((((((int)blockIdx.x) / 7) * 589824) + (rc_outer_outer * 72)) + ((int)threadIdx.x)) + 331776)];
+        kernel_shared[(((int)threadIdx.x) + 5248)] = kernel[((((((((int)blockIdx.x) / 7) * 589824) + (((((int)threadIdx.x) + 5248) / 72) * 4608)) + (rc_outer_outer * 72)) + ((((((int)threadIdx.x) + 64) % 72) / 3) * 3)) + ((((int)threadIdx.x) + 1) % 3))];
+        kernel_shared[(((int)threadIdx.x) + 5312)] = kernel[((((((((int)blockIdx.x) / 7) * 589824) + (((((int)threadIdx.x) + 5312) / 72) * 4608)) + (rc_outer_outer * 72)) + ((((((int)threadIdx.x) + 56) % 72) / 3) * 3)) + ((((int)threadIdx.x) + 2) % 3))];
+        kernel_shared[(((int)threadIdx.x) + 5376)] = kernel[((((((((int)blockIdx.x) / 7) * 589824) + (((((int)threadIdx.x) + 5376) / 72) * 4608)) + (rc_outer_outer * 72)) + ((((((int)threadIdx.x) / 3) + 16) % 24) * 3)) + (((int)threadIdx.x) % 3))];
+        kernel_shared[(((int)threadIdx.x) + 5440)] = kernel[((((((((int)blockIdx.x) / 7) * 589824) + (((((int)threadIdx.x) + 5440) / 72) * 4608)) + (rc_outer_outer * 72)) + ((((((int)threadIdx.x) + 40) % 72) / 3) * 3)) + ((((int)threadIdx.x) + 1) % 3))];
+        kernel_shared[(((int)threadIdx.x) + 5504)] = kernel[((((((((int)blockIdx.x) / 7) * 589824) + (((((int)threadIdx.x) + 5504) / 72) * 4608)) + (rc_outer_outer * 72)) + ((((((int)threadIdx.x) + 32) % 72) / 3) * 3)) + ((((int)threadIdx.x) + 2) % 3))];
+        kernel_shared[(((int)threadIdx.x) + 5568)] = kernel[((((((((int)blockIdx.x) / 7) * 589824) + (((((int)threadIdx.x) + 5568) / 72) * 4608)) + (rc_outer_outer * 72)) + ((((((int)threadIdx.x) / 3) + 8) % 24) * 3)) + (((int)threadIdx.x) % 3))];
+        kernel_shared[(((int)threadIdx.x) + 5632)] = kernel[((((((((int)blockIdx.x) / 7) * 589824) + (((((int)threadIdx.x) + 5632) / 72) * 4608)) + (rc_outer_outer * 72)) + ((((((int)threadIdx.x) + 16) % 72) / 3) * 3)) + ((((int)threadIdx.x) + 1) % 3))];
+        kernel_shared[(((int)threadIdx.x) + 5696)] = kernel[((((((((int)blockIdx.x) / 7) * 589824) + (((((int)threadIdx.x) + 5696) / 72) * 4608)) + (rc_outer_outer * 72)) + (((((int)threadIdx.x) + 8) / 3) * 3)) + ((((int)threadIdx.x) + 2) % 3))];
+        kernel_shared[(((int)threadIdx.x) + 5760)] = kernel[(((((((int)blockIdx.x) / 7) * 589824) + (rc_outer_outer * 72)) + ((int)threadIdx.x)) + 368640)];
+        kernel_shared[(((int)threadIdx.x) + 5824)] = kernel[((((((((int)blockIdx.x) / 7) * 589824) + (((((int)threadIdx.x) + 5824) / 72) * 4608)) + (rc_outer_outer * 72)) + ((((((int)threadIdx.x) + 64) % 72) / 3) * 3)) + ((((int)threadIdx.x) + 1) % 3))];
+        kernel_shared[(((int)threadIdx.x) + 5888)] = kernel[((((((((int)blockIdx.x) / 7) * 589824) + (((((int)threadIdx.x) + 5888) / 72) * 4608)) + (rc_outer_outer * 72)) + ((((((int)threadIdx.x) + 56) % 72) / 3) * 3)) + ((((int)threadIdx.x) + 2) % 3))];
+        kernel_shared[(((int)threadIdx.x) + 5952)] = kernel[((((((((int)blockIdx.x) / 7) * 589824) + (((((int)threadIdx.x) + 5952) / 72) * 4608)) + (rc_outer_outer * 72)) + ((((((int)threadIdx.x) / 3) + 16) % 24) * 3)) + (((int)threadIdx.x) % 3))];
+        kernel_shared[(((int)threadIdx.x) + 6016)] = kernel[((((((((int)blockIdx.x) / 7) * 589824) + (((((int)threadIdx.x) + 6016) / 72) * 4608)) + (rc_outer_outer * 72)) + ((((((int)threadIdx.x) + 40) % 72) / 3) * 3)) + ((((int)threadIdx.x) + 1) % 3))];
+        kernel_shared[(((int)threadIdx.x) + 6080)] = kernel[((((((((int)blockIdx.x) / 7) * 589824) + (((((int)threadIdx.x) + 6080) / 72) * 4608)) + (rc_outer_outer * 72)) + ((((((int)threadIdx.x) + 32) % 72) / 3) * 3)) + ((((int)threadIdx.x) + 2) % 3))];
+        kernel_shared[(((int)threadIdx.x) + 6144)] = kernel[((((((((int)blockIdx.x) / 7) * 589824) + (((((int)threadIdx.x) + 6144) / 72) * 4608)) + (rc_outer_outer * 72)) + ((((((int)threadIdx.x) / 3) + 8) % 24) * 3)) + (((int)threadIdx.x) % 3))];
+        kernel_shared[(((int)threadIdx.x) + 6208)] = kernel[((((((((int)blockIdx.x) / 7) * 589824) + (((((int)threadIdx.x) + 6208) / 72) * 4608)) + (rc_outer_outer * 72)) + ((((((int)threadIdx.x) + 16) % 72) / 3) * 3)) + ((((int)threadIdx.x) + 1) % 3))];
+        kernel_shared[(((int)threadIdx.x) + 6272)] = kernel[((((((((int)blockIdx.x) / 7) * 589824) + (((((int)threadIdx.x) + 6272) / 72) * 4608)) + (rc_outer_outer * 72)) + (((((int)threadIdx.x) + 8) / 3) * 3)) + ((((int)threadIdx.x) + 2) % 3))];
+        kernel_shared[(((int)threadIdx.x) + 6336)] = kernel[(((((((int)blockIdx.x) / 7) * 589824) + (rc_outer_outer * 72)) + ((int)threadIdx.x)) + 405504)];
+        kernel_shared[(((int)threadIdx.x) + 6400)] = kernel[((((((((int)blockIdx.x) / 7) * 589824) + (((((int)threadIdx.x) + 6400) / 72) * 4608)) + (rc_outer_outer * 72)) + ((((((int)threadIdx.x) + 64) % 72) / 3) * 3)) + ((((int)threadIdx.x) + 1) % 3))];
+        kernel_shared[(((int)threadIdx.x) + 6464)] = kernel[((((((((int)blockIdx.x) / 7) * 589824) + (((((int)threadIdx.x) + 6464) / 72) * 4608)) + (rc_outer_outer * 72)) + ((((((int)threadIdx.x) + 56) % 72) / 3) * 3)) + ((((int)threadIdx.x) + 2) % 3))];
+        kernel_shared[(((int)threadIdx.x) + 6528)] = kernel[((((((((int)blockIdx.x) / 7) * 589824) + (((((int)threadIdx.x) + 6528) / 72) * 4608)) + (rc_outer_outer * 72)) + ((((((int)threadIdx.x) / 3) + 16) % 24) * 3)) + (((int)threadIdx.x) % 3))];
+        kernel_shared[(((int)threadIdx.x) + 6592)] = kernel[((((((((int)blockIdx.x) / 7) * 589824) + (((((int)threadIdx.x) + 6592) / 72) * 4608)) + (rc_outer_outer * 72)) + ((((((int)threadIdx.x) + 40) % 72) / 3) * 3)) + ((((int)threadIdx.x) + 1) % 3))];
+        kernel_shared[(((int)threadIdx.x) + 6656)] = kernel[((((((((int)blockIdx.x) / 7) * 589824) + (((((int)threadIdx.x) + 6656) / 72) * 4608)) + (rc_outer_outer * 72)) + ((((((int)threadIdx.x) + 32) % 72) / 3) * 3)) + ((((int)threadIdx.x) + 2) % 3))];
+        kernel_shared[(((int)threadIdx.x) + 6720)] = kernel[((((((((int)blockIdx.x) / 7) * 589824) + (((((int)threadIdx.x) + 6720) / 72) * 4608)) + (rc_outer_outer * 72)) + ((((((int)threadIdx.x) / 3) + 8) % 24) * 3)) + (((int)threadIdx.x) % 3))];
+        kernel_shared[(((int)threadIdx.x) + 6784)] = kernel[((((((((int)blockIdx.x) / 7) * 589824) + (((((int)threadIdx.x) + 6784) / 72) * 4608)) + (rc_outer_outer * 72)) + ((((((int)threadIdx.x) + 16) % 72) / 3) * 3)) + ((((int)threadIdx.x) + 1) % 3))];
+        kernel_shared[(((int)threadIdx.x) + 6848)] = kernel[((((((((int)blockIdx.x) / 7) * 589824) + (((((int)threadIdx.x) + 6848) / 72) * 4608)) + (rc_outer_outer * 72)) + (((((int)threadIdx.x) + 8) / 3) * 3)) + ((((int)threadIdx.x) + 2) % 3))];
+        kernel_shared[(((int)threadIdx.x) + 6912)] = kernel[(((((((int)blockIdx.x) / 7) * 589824) + (rc_outer_outer * 72)) + ((int)threadIdx.x)) + 442368)];
+        kernel_shared[(((int)threadIdx.x) + 6976)] = kernel[((((((((int)blockIdx.x) / 7) * 589824) + (((((int)threadIdx.x) + 6976) / 72) * 4608)) + (rc_outer_outer * 72)) + ((((((int)threadIdx.x) + 64) % 72) / 3) * 3)) + ((((int)threadIdx.x) + 1) % 3))];
+        kernel_shared[(((int)threadIdx.x) + 7040)] = kernel[((((((((int)blockIdx.x) / 7) * 589824) + (((((int)threadIdx.x) + 7040) / 72) * 4608)) + (rc_outer_outer * 72)) + ((((((int)threadIdx.x) + 56) % 72) / 3) * 3)) + ((((int)threadIdx.x) + 2) % 3))];
+        kernel_shared[(((int)threadIdx.x) + 7104)] = kernel[((((((((int)blockIdx.x) / 7) * 589824) + (((((int)threadIdx.x) + 7104) / 72) * 4608)) + (rc_outer_outer * 72)) + ((((((int)threadIdx.x) / 3) + 16) % 24) * 3)) + (((int)threadIdx.x) % 3))];
+        kernel_shared[(((int)threadIdx.x) + 7168)] = kernel[((((((((int)blockIdx.x) / 7) * 589824) + (((((int)threadIdx.x) + 7168) / 72) * 4608)) + (rc_outer_outer * 72)) + ((((((int)threadIdx.x) + 40) % 72) / 3) * 3)) + ((((int)threadIdx.x) + 1) % 3))];
+        kernel_shared[(((int)threadIdx.x) + 7232)] = kernel[((((((((int)blockIdx.x) / 7) * 589824) + (((((int)threadIdx.x) + 7232) / 72) * 4608)) + (rc_outer_outer * 72)) + ((((((int)threadIdx.x) + 32) % 72) / 3) * 3)) + ((((int)threadIdx.x) + 2) % 3))];
+        kernel_shared[(((int)threadIdx.x) + 7296)] = kernel[((((((((int)blockIdx.x) / 7) * 589824) + (((((int)threadIdx.x) + 7296) / 72) * 4608)) + (rc_outer_outer * 72)) + ((((((int)threadIdx.x) / 3) + 8) % 24) * 3)) + (((int)threadIdx.x) % 3))];
+        kernel_shared[(((int)threadIdx.x) + 7360)] = kernel[((((((((int)blockIdx.x) / 7) * 589824) + (((((int)threadIdx.x) + 7360) / 72) * 4608)) + (rc_outer_outer * 72)) + ((((((int)threadIdx.x) + 16) % 72) / 3) * 3)) + ((((int)threadIdx.x) + 1) % 3))];
+        kernel_shared[(((int)threadIdx.x) + 7424)] = kernel[((((((((int)blockIdx.x) / 7) * 589824) + (((((int)threadIdx.x) + 7424) / 72) * 4608)) + (rc_outer_outer * 72)) + (((((int)threadIdx.x) + 8) / 3) * 3)) + ((((int)threadIdx.x) + 2) % 3))];
+        kernel_shared[(((int)threadIdx.x) + 7488)] = kernel[(((((((int)blockIdx.x) / 7) * 589824) + (rc_outer_outer * 72)) + ((int)threadIdx.x)) + 479232)];
+        kernel_shared[(((int)threadIdx.x) + 7552)] = kernel[((((((((int)blockIdx.x) / 7) * 589824) + (((((int)threadIdx.x) + 7552) / 72) * 4608)) + (rc_outer_outer * 72)) + ((((((int)threadIdx.x) + 64) % 72) / 3) * 3)) + ((((int)threadIdx.x) + 1) % 3))];
+        kernel_shared[(((int)threadIdx.x) + 7616)] = kernel[((((((((int)blockIdx.x) / 7) * 589824) + (((((int)threadIdx.x) + 7616) / 72) * 4608)) + (rc_outer_outer * 72)) + ((((((int)threadIdx.x) + 56) % 72) / 3) * 3)) + ((((int)threadIdx.x) + 2) % 3))];
+        kernel_shared[(((int)threadIdx.x) + 7680)] = kernel[((((((((int)blockIdx.x) / 7) * 589824) + (((((int)threadIdx.x) + 7680) / 72) * 4608)) + (rc_outer_outer * 72)) + ((((((int)threadIdx.x) / 3) + 16) % 24) * 3)) + (((int)threadIdx.x) % 3))];
+        kernel_shared[(((int)threadIdx.x) + 7744)] = kernel[((((((((int)blockIdx.x) / 7) * 589824) + (((((int)threadIdx.x) + 7744) / 72) * 4608)) + (rc_outer_outer * 72)) + ((((((int)threadIdx.x) + 40) % 72) / 3) * 3)) + ((((int)threadIdx.x) + 1) % 3))];
+        kernel_shared[(((int)threadIdx.x) + 7808)] = kernel[((((((((int)blockIdx.x) / 7) * 589824) + (((((int)threadIdx.x) + 7808) / 72) * 4608)) + (rc_outer_outer * 72)) + ((((((int)threadIdx.x) + 32) % 72) / 3) * 3)) + ((((int)threadIdx.x) + 2) % 3))];
+        kernel_shared[(((int)threadIdx.x) + 7872)] = kernel[((((((((int)blockIdx.x) / 7) * 589824) + (((((int)threadIdx.x) + 7872) / 72) * 4608)) + (rc_outer_outer * 72)) + ((((((int)threadIdx.x) / 3) + 8) % 24) * 3)) + (((int)threadIdx.x) % 3))];
+        kernel_shared[(((int)threadIdx.x) + 7936)] = kernel[((((((((int)blockIdx.x) / 7) * 589824) + (((((int)threadIdx.x) + 7936) / 72) * 4608)) + (rc_outer_outer * 72)) + ((((((int)threadIdx.x) + 16) % 72) / 3) * 3)) + ((((int)threadIdx.x) + 1) % 3))];
+        kernel_shared[(((int)threadIdx.x) + 8000)] = kernel[((((((((int)blockIdx.x) / 7) * 589824) + (((((int)threadIdx.x) + 8000) / 72) * 4608)) + (rc_outer_outer * 72)) + (((((int)threadIdx.x) + 8) / 3) * 3)) + ((((int)threadIdx.x) + 2) % 3))];
+        kernel_shared[(((int)threadIdx.x) + 8064)] = kernel[(((((((int)blockIdx.x) / 7) * 589824) + (rc_outer_outer * 72)) + ((int)threadIdx.x)) + 516096)];
+        kernel_shared[(((int)threadIdx.x) + 8128)] = kernel[((((((((int)blockIdx.x) / 7) * 589824) + (((((int)threadIdx.x) + 8128) / 72) * 4608)) + (rc_outer_outer * 72)) + ((((((int)threadIdx.x) + 64) % 72) / 3) * 3)) + ((((int)threadIdx.x) + 1) % 3))];
+        kernel_shared[(((int)threadIdx.x) + 8192)] = kernel[((((((((int)blockIdx.x) / 7) * 589824) + (((((int)threadIdx.x) + 8192) / 72) * 4608)) + (rc_outer_outer * 72)) + ((((((int)threadIdx.x) + 56) % 72) / 3) * 3)) + ((((int)threadIdx.x) + 2) % 3))];
+        kernel_shared[(((int)threadIdx.x) + 8256)] = kernel[((((((((int)blockIdx.x) / 7) * 589824) + (((((int)threadIdx.x) + 8256) / 72) * 4608)) + (rc_outer_outer * 72)) + ((((((int)threadIdx.x) / 3) + 16) % 24) * 3)) + (((int)threadIdx.x) % 3))];
+        kernel_shared[(((int)threadIdx.x) + 8320)] = kernel[((((((((int)blockIdx.x) / 7) * 589824) + (((((int)threadIdx.x) + 8320) / 72) * 4608)) + (rc_outer_outer * 72)) + ((((((int)threadIdx.x) + 40) % 72) / 3) * 3)) + ((((int)threadIdx.x) + 1) % 3))];
+        kernel_shared[(((int)threadIdx.x) + 8384)] = kernel[((((((((int)blockIdx.x) / 7) * 589824) + (((((int)threadIdx.x) + 8384) / 72) * 4608)) + (rc_outer_outer * 72)) + ((((((int)threadIdx.x) + 32) % 72) / 3) * 3)) + ((((int)threadIdx.x) + 2) % 3))];
+        kernel_shared[(((int)threadIdx.x) + 8448)] = kernel[((((((((int)blockIdx.x) / 7) * 589824) + (((((int)threadIdx.x) + 8448) / 72) * 4608)) + (rc_outer_outer * 72)) + ((((((int)threadIdx.x) / 3) + 8) % 24) * 3)) + (((int)threadIdx.x) % 3))];
+        kernel_shared[(((int)threadIdx.x) + 8512)] = kernel[((((((((int)blockIdx.x) / 7) * 589824) + (((((int)threadIdx.x) + 8512) / 72) * 4608)) + (rc_outer_outer * 72)) + ((((((int)threadIdx.x) + 16) % 72) / 3) * 3)) + ((((int)threadIdx.x) + 1) % 3))];
+        kernel_shared[(((int)threadIdx.x) + 8576)] = kernel[((((((((int)blockIdx.x) / 7) * 589824) + (((((int)threadIdx.x) + 8576) / 72) * 4608)) + (rc_outer_outer * 72)) + (((((int)threadIdx.x) + 8) / 3) * 3)) + ((((int)threadIdx.x) + 2) % 3))];
+        kernel_shared[(((int)threadIdx.x) + 8640)] = kernel[(((((((int)blockIdx.x) / 7) * 589824) + (rc_outer_outer * 72)) + ((int)threadIdx.x)) + 552960)];
+        kernel_shared[(((int)threadIdx.x) + 8704)] = kernel[((((((((int)blockIdx.x) / 7) * 589824) + (((((int)threadIdx.x) + 8704) / 72) * 4608)) + (rc_outer_outer * 72)) + ((((((int)threadIdx.x) + 64) % 72) / 3) * 3)) + ((((int)threadIdx.x) + 1) % 3))];
+        kernel_shared[(((int)threadIdx.x) + 8768)] = kernel[((((((((int)blockIdx.x) / 7) * 589824) + (((((int)threadIdx.x) + 8768) / 72) * 4608)) + (rc_outer_outer * 72)) + ((((((int)threadIdx.x) + 56) % 72) / 3) * 3)) + ((((int)threadIdx.x) + 2) % 3))];
+        kernel_shared[(((int)threadIdx.x) + 8832)] = kernel[((((((((int)blockIdx.x) / 7) * 589824) + (((((int)threadIdx.x) + 8832) / 72) * 4608)) + (rc_outer_outer * 72)) + ((((((int)threadIdx.x) / 3) + 16) % 24) * 3)) + (((int)threadIdx.x) % 3))];
+        kernel_shared[(((int)threadIdx.x) + 8896)] = kernel[((((((((int)blockIdx.x) / 7) * 589824) + (((((int)threadIdx.x) + 8896) / 72) * 4608)) + (rc_outer_outer * 72)) + ((((((int)threadIdx.x) + 40) % 72) / 3) * 3)) + ((((int)threadIdx.x) + 1) % 3))];
+        kernel_shared[(((int)threadIdx.x) + 8960)] = kernel[((((((((int)blockIdx.x) / 7) * 589824) + (((((int)threadIdx.x) + 8960) / 72) * 4608)) + (rc_outer_outer * 72)) + ((((((int)threadIdx.x) + 32) % 72) / 3) * 3)) + ((((int)threadIdx.x) + 2) % 3))];
+        kernel_shared[(((int)threadIdx.x) + 9024)] = kernel[((((((((int)blockIdx.x) / 7) * 589824) + (((((int)threadIdx.x) + 9024) / 72) * 4608)) + (rc_outer_outer * 72)) + ((((((int)threadIdx.x) / 3) + 8) % 24) * 3)) + (((int)threadIdx.x) % 3))];
+        kernel_shared[(((int)threadIdx.x) + 9088)] = kernel[((((((((int)blockIdx.x) / 7) * 589824) + (((((int)threadIdx.x) + 9088) / 72) * 4608)) + (rc_outer_outer * 72)) + ((((((int)threadIdx.x) + 16) % 72) / 3) * 3)) + ((((int)threadIdx.x) + 1) % 3))];
+        kernel_shared[(((int)threadIdx.x) + 9152)] = kernel[((((((((int)blockIdx.x) / 7) * 589824) + (((((int)threadIdx.x) + 9152) / 72) * 4608)) + (rc_outer_outer * 72)) + (((((int)threadIdx.x) + 8) / 3) * 3)) + ((((int)threadIdx.x) + 2) % 3))];
+        __syncthreads();
+        conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[0] * kernel_shared[(((int)threadIdx.x) * 72)]));
+        conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[1] * kernel_shared[(((int)threadIdx.x) * 72)]));
+        conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[2] * kernel_shared[(((int)threadIdx.x) * 72)]));
+        conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[3] * kernel_shared[(((int)threadIdx.x) * 72)]));
+        conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[4] * kernel_shared[(((int)threadIdx.x) * 72)]));
+        conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[5] * kernel_shared[(((int)threadIdx.x) * 72)]));
+        conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[6] * kernel_shared[(((int)threadIdx.x) * 72)]));
+        conv2d_nchw[7] = (conv2d_nchw[7] + (pad_temp_shared[0] * kernel_shared[((((int)threadIdx.x) * 72) + 4608)]));
+        conv2d_nchw[8] = (conv2d_nchw[8] + (pad_temp_shared[1] * kernel_shared[((((int)threadIdx.x) * 72) + 4608)]));
+        conv2d_nchw[9] = (conv2d_nchw[9] + (pad_temp_shared[2] * kernel_shared[((((int)threadIdx.x) * 72) + 4608)]));
+        conv2d_nchw[10] = (conv2d_nchw[10] + (pad_temp_shared[3] * kernel_shared[((((int)threadIdx.x) * 72) + 4608)]));
+        conv2d_nchw[11] = (conv2d_nchw[11] + (pad_temp_shared[4] * kernel_shared[((((int)threadIdx.x) * 72) + 4608)]));
+        conv2d_nchw[12] = (conv2d_nchw[12] + (pad_temp_shared[5] * kernel_shared[((((int)threadIdx.x) * 72) + 4608)]));
+        conv2d_nchw[13] = (conv2d_nchw[13] + (pad_temp_shared[6] * kernel_shared[((((int)threadIdx.x) * 72) + 4608)]));
+        conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[1] * kernel_shared[((((int)threadIdx.x) * 72) + 1)]));
+        conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[2] * kernel_shared[((((int)threadIdx.x) * 72) + 1)]));
+        conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[3] * kernel_shared[((((int)threadIdx.x) * 72) + 1)]));
+        conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[4] * kernel_shared[((((int)threadIdx.x) * 72) + 1)]));
+        conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[5] * kernel_shared[((((int)threadIdx.x) * 72) + 1)]));
+        conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[6] * kernel_shared[((((int)threadIdx.x) * 72) + 1)]));
+        conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[7] * kernel_shared[((((int)threadIdx.x) * 72) + 1)]));
+        conv2d_nchw[7] = (conv2d_nchw[7] + (pad_temp_shared[1] * kernel_shared[((((int)threadIdx.x) * 72) + 4609)]));
+        conv2d_nchw[8] = (conv2d_nchw[8] + (pad_temp_shared[2] * kernel_shared[((((int)threadIdx.x) * 72) + 4609)]));
+        conv2d_nchw[9] = (conv2d_nchw[9] + (pad_temp_shared[3] * kernel_shared[((((int)threadIdx.x) * 72) + 4609)]));
+        conv2d_nchw[10] = (conv2d_nchw[10] + (pad_temp_shared[4] * kernel_shared[((((int)threadIdx.x) * 72) + 4609)]));
+        conv2d_nchw[11] = (conv2d_nchw[11] + (pad_temp_shared[5] * kernel_shared[((((int)threadIdx.x) * 72) + 4609)]));
+        conv2d_nchw[12] = (conv2d_nchw[12] + (pad_temp_shared[6] * kernel_shared[((((int)threadIdx.x) * 72) + 4609)]));
+        conv2d_nchw[13] = (conv2d_nchw[13] + (pad_temp_shared[7] * kernel_shared[((((int)threadIdx.x) * 72) + 4609)]));
+        conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[2] * kernel_shared[((((int)threadIdx.x) * 72) + 2)]));
+        conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[3] * kernel_shared[((((int)threadIdx.x) * 72) + 2)]));
+        conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[4] * kernel_shared[((((int)threadIdx.x) * 72) + 2)]));
+        conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[5] * kernel_shared[((((int)threadIdx.x) * 72) + 2)]));
+        conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[6] * kernel_shared[((((int)threadIdx.x) * 72) + 2)]));
+        conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[7] * kernel_shared[((((int)threadIdx.x) * 72) + 2)]));
+        conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[8] * kernel_shared[((((int)threadIdx.x) * 72) + 2)]));
+        conv2d_nchw[7] = (conv2d_nchw[7] + (pad_temp_shared[2] * kernel_shared[((((int)threadIdx.x) * 72) + 4610)]));
+        conv2d_nchw[8] = (conv2d_nchw[8] + (pad_temp_shared[3] * kernel_shared[((((int)threadIdx.x) * 72) + 4610)]));
+        conv2d_nchw[9] = (conv2d_nchw[9] + (pad_temp_shared[4] * kernel_shared[((((int)threadIdx.x) * 72) + 4610)]));
+        conv2d_nchw[10] = (conv2d_nchw[10] + (pad_temp_shared[5] * kernel_shared[((((int)threadIdx.x) * 72) + 4610)]));
+        conv2d_nchw[11] = (conv2d_nchw[11] + (pad_temp_shared[6] * kernel_shared[((((int)threadIdx.x) * 72) + 4610)]));
+        conv2d_nchw[12] = (conv2d_nchw[12] + (pad_temp_shared[7] * kernel_shared[((((int)threadIdx.x) * 72) + 4610)]));
+        conv2d_nchw[13] = (conv2d_nchw[13] + (pad_temp_shared[8] * kernel_shared[((((int)threadIdx.x) * 72) + 4610)]));
+        conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[9] * kernel_shared[((((int)threadIdx.x) * 72) + 3)]));
+        conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[10] * kernel_shared[((((int)threadIdx.x) * 72) + 3)]));
+        conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[11] * kernel_shared[((((int)threadIdx.x) * 72) + 3)]));
+        conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[12] * kernel_shared[((((int)threadIdx.x) * 72) + 3)]));
+        conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[13] * kernel_shared[((((int)threadIdx.x) * 72) + 3)]));
+        conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[14] * kernel_shared[((((int)threadIdx.x) * 72) + 3)]));
+        conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[15] * kernel_shared[((((int)threadIdx.x) * 72) + 3)]));
+        conv2d_nchw[7] = (conv2d_nchw[7] + (pad_temp_shared[9] * kernel_shared[((((int)threadIdx.x) * 72) + 4611)]));
+        conv2d_nchw[8] = (conv2d_nchw[8] + (pad_temp_shared[10] * kernel_shared[((((int)threadIdx.x) * 72) + 4611)]));
+        conv2d_nchw[9] = (conv2d_nchw[9] + (pad_temp_shared[11] * kernel_shared[((((int)threadIdx.x) * 72) + 4611)]));
+        conv2d_nchw[10] = (conv2d_nchw[10] + (pad_temp_shared[12] * kernel_shared[((((int)threadIdx.x) * 72) + 4611)]));
+        conv2d_nchw[11] = (conv2d_nchw[11] + (pad_temp_shared[13] * kernel_shared[((((int)threadIdx.x) * 72) + 4611)]));
+        conv2d_nchw[12] = (conv2d_nchw[12] + (pad_temp_shared[14] * kernel_shared[((((int)threadIdx.x) * 72) + 4611)]));
+        conv2d_nchw[13] = (conv2d_nchw[13] + (pad_temp_shared[15] * kernel_shared[((((int)threadIdx.x) * 72) + 4611)]));
+        conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[10] * kernel_shared[((((int)threadIdx.x) * 72) + 4)]));
+        conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[11] * kernel_shared[((((int)threadIdx.x) * 72) + 4)]));
+        conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[12] * kernel_shared[((((int)threadIdx.x) * 72) + 4)]));
+        conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[13] * kernel_shared[((((int)threadIdx.x) * 72) + 4)]));
+        conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[14] * kernel_shared[((((int)threadIdx.x) * 72) + 4)]));
+        conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[15] * kernel_shared[((((int)threadIdx.x) * 72) + 4)]));
+        conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[16] * kernel_shared[((((int)threadIdx.x) * 72) + 4)]));
+        conv2d_nchw[7] = (conv2d_nchw[7] + (pad_temp_shared[10] * kernel_shared[((((int)threadIdx.x) * 72) + 4612)]));
+        conv2d_nchw[8] = (conv2d_nchw[8] + (pad_temp_shared[11] * kernel_shared[((((int)threadIdx.x) * 72) + 4612)]));
+        conv2d_nchw[9] = (conv2d_nchw[9] + (pad_temp_shared[12] * kernel_shared[((((int)threadIdx.x) * 72) + 4612)]));
+        conv2d_nchw[10] = (conv2d_nchw[10] + (pad_temp_shared[13] * kernel_shared[((((int)threadIdx.x) * 72) + 4612)]));
+        conv2d_nchw[11] = (conv2d_nchw[11] + (pad_temp_shared[14] * kernel_shared[((((int)threadIdx.x) * 72) + 4612)]));
+        conv2d_nchw[12] = (conv2d_nchw[12] + (pad_temp_shared[15] * kernel_shared[((((int)threadIdx.x) * 72) + 4612)]));
+        conv2d_nchw[13] = (conv2d_nchw[13] + (pad_temp_shared[16] * kernel_shared[((((int)threadIdx.x) * 72) + 4612)]));
+        conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[11] * kernel_shared[((((int)threadIdx.x) * 72) + 5)]));
+        conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[12] * kernel_shared[((((int)threadIdx.x) * 72) + 5)]));
+        conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[13] * kernel_shared[((((int)threadIdx.x) * 72) + 5)]));
+        conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[14] * kernel_shared[((((int)threadIdx.x) * 72) + 5)]));
+        conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[15] * kernel_shared[((((int)threadIdx.x) * 72) + 5)]));
+        conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[16] * kernel_shared[((((int)threadIdx.x) * 72) + 5)]));
+        conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[17] * kernel_shared[((((int)threadIdx.x) * 72) + 5)]));
+        conv2d_nchw[7] = (conv2d_nchw[7] + (pad_temp_shared[11] * kernel_shared[((((int)threadIdx.x) * 72) + 4613)]));
+        conv2d_nchw[8] = (conv2d_nchw[8] + (pad_temp_shared[12] * kernel_shared[((((int)threadIdx.x) * 72) + 4613)]));
+        conv2d_nchw[9] = (conv2d_nchw[9] + (pad_temp_shared[13] * kernel_shared[((((int)threadIdx.x) * 72) + 4613)]));
+        conv2d_nchw[10] = (conv2d_nchw[10] + (pad_temp_shared[14] * kernel_shared[((((int)threadIdx.x) * 72) + 4613)]));
+        conv2d_nchw[11] = (conv2d_nchw[11] + (pad_temp_shared[15] * kernel_shared[((((int)threadIdx.x) * 72) + 4613)]));
+        conv2d_nchw[12] = (conv2d_nchw[12] + (pad_temp_shared[16] * kernel_shared[((((int)threadIdx.x) * 72) + 4613)]));
+        conv2d_nchw[13] = (conv2d_nchw[13] + (pad_temp_shared[17] * kernel_shared[((((int)threadIdx.x) * 72) + 4613)]));
+        conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[18] * kernel_shared[((((int)threadIdx.x) * 72) + 6)]));
+        conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[19] * kernel_shared[((((int)threadIdx.x) * 72) + 6)]));
+        conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[20] * kernel_shared[((((int)threadIdx.x) * 72) + 6)]));
+        conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[21] * kernel_shared[((((int)threadIdx.x) * 72) + 6)]));
+        conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[22] * kernel_shared[((((int)threadIdx.x) * 72) + 6)]));
+        conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[23] * kernel_shared[((((int)threadIdx.x) * 72) + 6)]));
+        conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[24] * kernel_shared[((((int)threadIdx.x) * 72) + 6)]));
+        conv2d_nchw[7] = (conv2d_nchw[7] + (pad_temp_shared[18] * kernel_shared[((((int)threadIdx.x) * 72) + 4614)]));
+        conv2d_nchw[8] = (conv2d_nchw[8] + (pad_temp_shared[19] * kernel_shared[((((int)threadIdx.x) * 72) + 4614)]));
+        conv2d_nchw[9] = (conv2d_nchw[9] + (pad_temp_shared[20] * kernel_shared[((((int)threadIdx.x) * 72) + 4614)]));
+        conv2d_nchw[10] = (conv2d_nchw[10] + (pad_temp_shared[21] * kernel_shared[((((int)threadIdx.x) * 72) + 4614)]));
+        conv2d_nchw[11] = (conv2d_nchw[11] + (pad_temp_shared[22] * kernel_shared[((((int)threadIdx.x) * 72) + 4614)]));
+        conv2d_nchw[12] = (conv2d_nchw[12] + (pad_temp_shared[23] * kernel_shared[((((int)threadIdx.x) * 72) + 4614)]));
+        conv2d_nchw[13] = (conv2d_nchw[13] + (pad_temp_shared[24] * kernel_shared[((((int)threadIdx.x) * 72) + 4614)]));
+        conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[19] * kernel_shared[((((int)threadIdx.x) * 72) + 7)]));
+        conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[20] * kernel_shared[((((int)threadIdx.x) * 72) + 7)]));
+        conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[21] * kernel_shared[((((int)threadIdx.x) * 72) + 7)]));
+        conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[22] * kernel_shared[((((int)threadIdx.x) * 72) + 7)]));
+        conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[23] * kernel_shared[((((int)threadIdx.x) * 72) + 7)]));
+        conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[24] * kernel_shared[((((int)threadIdx.x) * 72) + 7)]));
+        conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[25] * kernel_shared[((((int)threadIdx.x) * 72) + 7)]));
+        conv2d_nchw[7] = (conv2d_nchw[7] + (pad_temp_shared[19] * kernel_shared[((((int)threadIdx.x) * 72) + 4615)]));
+        conv2d_nchw[8] = (conv2d_nchw[8] + (pad_temp_shared[20] * kernel_shared[((((int)threadIdx.x) * 72) + 4615)]));
+        conv2d_nchw[9] = (conv2d_nchw[9] + (pad_temp_shared[21] * kernel_shared[((((int)threadIdx.x) * 72) + 4615)]));
+        conv2d_nchw[10] = (conv2d_nchw[10] + (pad_temp_shared[22] * kernel_shared[((((int)threadIdx.x) * 72) + 4615)]));
+        conv2d_nchw[11] = (conv2d_nchw[11] + (pad_temp_shared[23] * kernel_shared[((((int)threadIdx.x) * 72) + 4615)]));
+        conv2d_nchw[12] = (conv2d_nchw[12] + (pad_temp_shared[24] * kernel_shared[((((int)threadIdx.x) * 72) + 4615)]));
+        conv2d_nchw[13] = (conv2d_nchw[13] + (pad_temp_shared[25] * kernel_shared[((((int)threadIdx.x) * 72) + 4615)]));
+        conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[20] * kernel_shared[((((int)threadIdx.x) * 72) + 8)]));
+        conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[21] * kernel_shared[((((int)threadIdx.x) * 72) + 8)]));
+        conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[22] * kernel_shared[((((int)threadIdx.x) * 72) + 8)]));
+        conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[23] * kernel_shared[((((int)threadIdx.x) * 72) + 8)]));
+        conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[24] * kernel_shared[((((int)threadIdx.x) * 72) + 8)]));
+        conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[25] * kernel_shared[((((int)threadIdx.x) * 72) + 8)]));
+        conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[26] * kernel_shared[((((int)threadIdx.x) * 72) + 8)]));
+        conv2d_nchw[7] = (conv2d_nchw[7] + (pad_temp_shared[20] * kernel_shared[((((int)threadIdx.x) * 72) + 4616)]));
+        conv2d_nchw[8] = (conv2d_nchw[8] + (pad_temp_shared[21] * kernel_shared[((((int)threadIdx.x) * 72) + 4616)]));
+        conv2d_nchw[9] = (conv2d_nchw[9] + (pad_temp_shared[22] * kernel_shared[((((int)threadIdx.x) * 72) + 4616)]));
+        conv2d_nchw[10] = (conv2d_nchw[10] + (pad_temp_shared[23] * kernel_shared[((((int)threadIdx.x) * 72) + 4616)]));
+        conv2d_nchw[11] = (conv2d_nchw[11] + (pad_temp_shared[24] * kernel_shared[((((int)threadIdx.x) * 72) + 4616)]));
+        conv2d_nchw[12] = (conv2d_nchw[12] + (pad_temp_shared[25] * kernel_shared[((((int)threadIdx.x) * 72) + 4616)]));
+        conv2d_nchw[13] = (conv2d_nchw[13] + (pad_temp_shared[26] * kernel_shared[((((int)threadIdx.x) * 72) + 4616)]));
+        conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[27] * kernel_shared[((((int)threadIdx.x) * 72) + 9)]));
+        conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[28] * kernel_shared[((((int)threadIdx.x) * 72) + 9)]));
+        conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[29] * kernel_shared[((((int)threadIdx.x) * 72) + 9)]));
+        conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[30] * kernel_shared[((((int)threadIdx.x) * 72) + 9)]));
+        conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[31] * kernel_shared[((((int)threadIdx.x) * 72) + 9)]));
+        conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[32] * kernel_shared[((((int)threadIdx.x) * 72) + 9)]));
+        conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[33] * kernel_shared[((((int)threadIdx.x) * 72) + 9)]));
+        conv2d_nchw[7] = (conv2d_nchw[7] + (pad_temp_shared[27] * kernel_shared[((((int)threadIdx.x) * 72) + 4617)]));
+        conv2d_nchw[8] = (conv2d_nchw[8] + (pad_temp_shared[28] * kernel_shared[((((int)threadIdx.x) * 72) + 4617)]));
+        conv2d_nchw[9] = (conv2d_nchw[9] + (pad_temp_shared[29] * kernel_shared[((((int)threadIdx.x) * 72) + 4617)]));
+        conv2d_nchw[10] = (conv2d_nchw[10] + (pad_temp_shared[30] * kernel_shared[((((int)threadIdx.x) * 72) + 4617)]));
+        conv2d_nchw[11] = (conv2d_nchw[11] + (pad_temp_shared[31] * kernel_shared[((((int)threadIdx.x) * 72) + 4617)]));
+        conv2d_nchw[12] = (conv2d_nchw[12] + (pad_temp_shared[32] * kernel_shared[((((int)threadIdx.x) * 72) + 4617)]));
+        conv2d_nchw[13] = (conv2d_nchw[13] + (pad_temp_shared[33] * kernel_shared[((((int)threadIdx.x) * 72) + 4617)]));
+        conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[28] * kernel_shared[((((int)threadIdx.x) * 72) + 10)]));
+        conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[29] * kernel_shared[((((int)threadIdx.x) * 72) + 10)]));
+        conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[30] * kernel_shared[((((int)threadIdx.x) * 72) + 10)]));
+        conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[31] * kernel_shared[((((int)threadIdx.x) * 72) + 10)]));
+        conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[32] * kernel_shared[((((int)threadIdx.x) * 72) + 10)]));
+        conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[33] * kernel_shared[((((int)threadIdx.x) * 72) + 10)]));
+        conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[34] * kernel_shared[((((int)threadIdx.x) * 72) + 10)]));
+        conv2d_nchw[7] = (conv2d_nchw[7] + (pad_temp_shared[28] * kernel_shared[((((int)threadIdx.x) * 72) + 4618)]));
+        conv2d_nchw[8] = (conv2d_nchw[8] + (pad_temp_shared[29] * kernel_shared[((((int)threadIdx.x) * 72) + 4618)]));
+        conv2d_nchw[9] = (conv2d_nchw[9] + (pad_temp_shared[30] * kernel_shared[((((int)threadIdx.x) * 72) + 4618)]));
+        conv2d_nchw[10] = (conv2d_nchw[10] + (pad_temp_shared[31] * kernel_shared[((((int)threadIdx.x) * 72) + 4618)]));
+        conv2d_nchw[11] = (conv2d_nchw[11] + (pad_temp_shared[32] * kernel_shared[((((int)threadIdx.x) * 72) + 4618)]));
+        conv2d_nchw[12] = (conv2d_nchw[12] + (pad_temp_shared[33] * kernel_shared[((((int)threadIdx.x) * 72) + 4618)]));
+        conv2d_nchw[13] = (conv2d_nchw[13] + (pad_temp_shared[34] * kernel_shared[((((int)threadIdx.x) * 72) + 4618)]));
+        conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[29] * kernel_shared[((((int)threadIdx.x) * 72) + 11)]));
+        conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[30] * kernel_shared[((((int)threadIdx.x) * 72) + 11)]));
+        conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[31] * kernel_shared[((((int)threadIdx.x) * 72) + 11)]));
+        conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[32] * kernel_shared[((((int)threadIdx.x) * 72) + 11)]));
+        conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[33] * kernel_shared[((((int)threadIdx.x) * 72) + 11)]));
+        conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[34] * kernel_shared[((((int)threadIdx.x) * 72) + 11)]));
+        conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[35] * kernel_shared[((((int)threadIdx.x) * 72) + 11)]));
+        conv2d_nchw[7] = (conv2d_nchw[7] + (pad_temp_shared[29] * kernel_shared[((((int)threadIdx.x) * 72) + 4619)]));
+        conv2d_nchw[8] = (conv2d_nchw[8] + (pad_temp_shared[30] * kernel_shared[((((int)threadIdx.x) * 72) + 4619)]));
+        conv2d_nchw[9] = (conv2d_nchw[9] + (pad_temp_shared[31] * kernel_shared[((((int)threadIdx.x) * 72) + 4619)]));
+        conv2d_nchw[10] = (conv2d_nchw[10] + (pad_temp_shared[32] * kernel_shared[((((int)threadIdx.x) * 72) + 4619)]));
+        conv2d_nchw[11] = (conv2d_nchw[11] + (pad_temp_shared[33] * kernel_shared[((((int)threadIdx.x) * 72) + 4619)]));
+        conv2d_nchw[12] = (conv2d_nchw[12] + (pad_temp_shared[34] * kernel_shared[((((int)threadIdx.x) * 72) + 4619)]));
+        conv2d_nchw[13] = (conv2d_nchw[13] + (pad_temp_shared[35] * kernel_shared[((((int)threadIdx.x) * 72) + 4619)]));
+        conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[36] * kernel_shared[((((int)threadIdx.x) * 72) + 12)]));
+        conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[37] * kernel_shared[((((int)threadIdx.x) * 72) + 12)]));
+        conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[38] * kernel_shared[((((int)threadIdx.x) * 72) + 12)]));
+        conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[39] * kernel_shared[((((int)threadIdx.x) * 72) + 12)]));
+        conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[40] * kernel_shared[((((int)threadIdx.x) * 72) + 12)]));
+        conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[41] * kernel_shared[((((int)threadIdx.x) * 72) + 12)]));
+        conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[42] * kernel_shared[((((int)threadIdx.x) * 72) + 12)]));
+        conv2d_nchw[7] = (conv2d_nchw[7] + (pad_temp_shared[36] * kernel_shared[((((int)threadIdx.x) * 72) + 4620)]));
+        conv2d_nchw[8] = (conv2d_nchw[8] + (pad_temp_shared[37] * kernel_shared[((((int)threadIdx.x) * 72) + 4620)]));
+        conv2d_nchw[9] = (conv2d_nchw[9] + (pad_temp_shared[38] * kernel_shared[((((int)threadIdx.x) * 72) + 4620)]));
+        conv2d_nchw[10] = (conv2d_nchw[10] + (pad_temp_shared[39] * kernel_shared[((((int)threadIdx.x) * 72) + 4620)]));
+        conv2d_nchw[11] = (conv2d_nchw[11] + (pad_temp_shared[40] * kernel_shared[((((int)threadIdx.x) * 72) + 4620)]));
+        conv2d_nchw[12] = (conv2d_nchw[12] + (pad_temp_shared[41] * kernel_shared[((((int)threadIdx.x) * 72) + 4620)]));
+        conv2d_nchw[13] = (conv2d_nchw[13] + (pad_temp_shared[42] * kernel_shared[((((int)threadIdx.x) * 72) + 4620)]));
+        conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[37] * kernel_shared[((((int)threadIdx.x) * 72) + 13)]));
+        conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[38] * kernel_shared[((((int)threadIdx.x) * 72) + 13)]));
+        conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[39] * kernel_shared[((((int)threadIdx.x) * 72) + 13)]));
+        conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[40] * kernel_shared[((((int)threadIdx.x) * 72) + 13)]));
+        conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[41] * kernel_shared[((((int)threadIdx.x) * 72) + 13)]));
+        conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[42] * kernel_shared[((((int)threadIdx.x) * 72) + 13)]));
+        conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[43] * kernel_shared[((((int)threadIdx.x) * 72) + 13)]));
+        conv2d_nchw[7] = (conv2d_nchw[7] + (pad_temp_shared[37] * kernel_shared[((((int)threadIdx.x) * 72) + 4621)]));
+        conv2d_nchw[8] = (conv2d_nchw[8] + (pad_temp_shared[38] * kernel_shared[((((int)threadIdx.x) * 72) + 4621)]));
+        conv2d_nchw[9] = (conv2d_nchw[9] + (pad_temp_shared[39] * kernel_shared[((((int)threadIdx.x) * 72) + 4621)]));
+        conv2d_nchw[10] = (conv2d_nchw[10] + (pad_temp_shared[40] * kernel_shared[((((int)threadIdx.x) * 72) + 4621)]));
+        conv2d_nchw[11] = (conv2d_nchw[11] + (pad_temp_shared[41] * kernel_shared[((((int)threadIdx.x) * 72) + 4621)]));
+        conv2d_nchw[12] = (conv2d_nchw[12] + (pad_temp_shared[42] * kernel_shared[((((int)threadIdx.x) * 72) + 4621)]));
+        conv2d_nchw[13] = (conv2d_nchw[13] + (pad_temp_shared[43] * kernel_shared[((((int)threadIdx.x) * 72) + 4621)]));
+        conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[38] * kernel_shared[((((int)threadIdx.x) * 72) + 14)]));
+        conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[39] * kernel_shared[((((int)threadIdx.x) * 72) + 14)]));
+        conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[40] * kernel_shared[((((int)threadIdx.x) * 72) + 14)]));
+        conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[41] * kernel_shared[((((int)threadIdx.x) * 72) + 14)]));
+        conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[42] * kernel_shared[((((int)threadIdx.x) * 72) + 14)]));
+        conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[43] * kernel_shared[((((int)threadIdx.x) * 72) + 14)]));
+        conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[44] * kernel_shared[((((int)threadIdx.x) * 72) + 14)]));
+        conv2d_nchw[7] = (conv2d_nchw[7] + (pad_temp_shared[38] * kernel_shared[((((int)threadIdx.x) * 72) + 4622)]));
+        conv2d_nchw[8] = (conv2d_nchw[8] + (pad_temp_shared[39] * kernel_shared[((((int)threadIdx.x) * 72) + 4622)]));
+        conv2d_nchw[9] = (conv2d_nchw[9] + (pad_temp_shared[40] * kernel_shared[((((int)threadIdx.x) * 72) + 4622)]));
+        conv2d_nchw[10] = (conv2d_nchw[10] + (pad_temp_shared[41] * kernel_shared[((((int)threadIdx.x) * 72) + 4622)]));
+        conv2d_nchw[11] = (conv2d_nchw[11] + (pad_temp_shared[42] * kernel_shared[((((int)threadIdx.x) * 72) + 4622)]));
+        conv2d_nchw[12] = (conv2d_nchw[12] + (pad_temp_shared[43] * kernel_shared[((((int)threadIdx.x) * 72) + 4622)]));
+        conv2d_nchw[13] = (conv2d_nchw[13] + (pad_temp_shared[44] * kernel_shared[((((int)threadIdx.x) * 72) + 4622)]));
+        conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[45] * kernel_shared[((((int)threadIdx.x) * 72) + 15)]));
+        conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[46] * kernel_shared[((((int)threadIdx.x) * 72) + 15)]));
+        conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[47] * kernel_shared[((((int)threadIdx.x) * 72) + 15)]));
+        conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[48] * kernel_shared[((((int)threadIdx.x) * 72) + 15)]));
+        conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[49] * kernel_shared[((((int)threadIdx.x) * 72) + 15)]));
+        conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[50] * kernel_shared[((((int)threadIdx.x) * 72) + 15)]));
+        conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[51] * kernel_shared[((((int)threadIdx.x) * 72) + 15)]));
+        conv2d_nchw[7] = (conv2d_nchw[7] + (pad_temp_shared[45] * kernel_shared[((((int)threadIdx.x) * 72) + 4623)]));
+        conv2d_nchw[8] = (conv2d_nchw[8] + (pad_temp_shared[46] * kernel_shared[((((int)threadIdx.x) * 72) + 4623)]));
+        conv2d_nchw[9] = (conv2d_nchw[9] + (pad_temp_shared[47] * kernel_shared[((((int)threadIdx.x) * 72) + 4623)]));
+        conv2d_nchw[10] = (conv2d_nchw[10] + (pad_temp_shared[48] * kernel_shared[((((int)threadIdx.x) * 72) + 4623)]));
+        conv2d_nchw[11] = (conv2d_nchw[11] + (pad_temp_shared[49] * kernel_shared[((((int)threadIdx.x) * 72) + 4623)]));
+        conv2d_nchw[12] = (conv2d_nchw[12] + (pad_temp_shared[50] * kernel_shared[((((int)threadIdx.x) * 72) + 4623)]));
+        conv2d_nchw[13] = (conv2d_nchw[13] + (pad_temp_shared[51] * kernel_shared[((((int)threadIdx.x) * 72) + 4623)]));
+        conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[46] * kernel_shared[((((int)threadIdx.x) * 72) + 16)]));
+        conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[47] * kernel_shared[((((int)threadIdx.x) * 72) + 16)]));
+        conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[48] * kernel_shared[((((int)threadIdx.x) * 72) + 16)]));
+        conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[49] * kernel_shared[((((int)threadIdx.x) * 72) + 16)]));
+        conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[50] * kernel_shared[((((int)threadIdx.x) * 72) + 16)]));
+        conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[51] * kernel_shared[((((int)threadIdx.x) * 72) + 16)]));
+        conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[52] * kernel_shared[((((int)threadIdx.x) * 72) + 16)]));
+        conv2d_nchw[7] = (conv2d_nchw[7] + (pad_temp_shared[46] * kernel_shared[((((int)threadIdx.x) * 72) + 4624)]));
+        conv2d_nchw[8] = (conv2d_nchw[8] + (pad_temp_shared[47] * kernel_shared[((((int)threadIdx.x) * 72) + 4624)]));
+        conv2d_nchw[9] = (conv2d_nchw[9] + (pad_temp_shared[48] * kernel_shared[((((int)threadIdx.x) * 72) + 4624)]));
+        conv2d_nchw[10] = (conv2d_nchw[10] + (pad_temp_shared[49] * kernel_shared[((((int)threadIdx.x) * 72) + 4624)]));
+        conv2d_nchw[11] = (conv2d_nchw[11] + (pad_temp_shared[50] * kernel_shared[((((int)threadIdx.x) * 72) + 4624)]));
+        conv2d_nchw[12] = (conv2d_nchw[12] + (pad_temp_shared[51] * kernel_shared[((((int)threadIdx.x) * 72) + 4624)]));
+        conv2d_nchw[13] = (conv2d_nchw[13] + (pad_temp_shared[52] * kernel_shared[((((int)threadIdx.x) * 72) + 4624)]));
+        conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[47] * kernel_shared[((((int)threadIdx.x) * 72) + 17)]));
+        conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[48] * kernel_shared[((((int)threadIdx.x) * 72) + 17)]));
+        conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[49] * kernel_shared[((((int)threadIdx.x) * 72) + 17)]));
+        conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[50] * kernel_shared[((((int)threadIdx.x) * 72) + 17)]));
+        conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[51] * kernel_shared[((((int)threadIdx.x) * 72) + 17)]));
+        conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[52] * kernel_shared[((((int)threadIdx.x) * 72) + 17)]));
+        conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[53] * kernel_shared[((((int)threadIdx.x) * 72) + 17)]));
+        conv2d_nchw[7] = (conv2d_nchw[7] + (pad_temp_shared[47] * kernel_shared[((((int)threadIdx.x) * 72) + 4625)]));
+        conv2d_nchw[8] = (conv2d_nchw[8] + (pad_temp_shared[48] * kernel_shared[((((int)threadIdx.x) * 72) + 4625)]));
+        conv2d_nchw[9] = (conv2d_nchw[9] + (pad_temp_shared[49] * kernel_shared[((((int)threadIdx.x) * 72) + 4625)]));
+        conv2d_nchw[10] = (conv2d_nchw[10] + (pad_temp_shared[50] * kernel_shared[((((int)threadIdx.x) * 72) + 4625)]));
+        conv2d_nchw[11] = (conv2d_nchw[11] + (pad_temp_shared[51] * kernel_shared[((((int)threadIdx.x) * 72) + 4625)]));
+        conv2d_nchw[12] = (conv2d_nchw[12] + (pad_temp_shared[52] * kernel_shared[((((int)threadIdx.x) * 72) + 4625)]));
+        conv2d_nchw[13] = (conv2d_nchw[13] + (pad_temp_shared[53] * kernel_shared[((((int)threadIdx.x) * 72) + 4625)]));
+        conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[54] * kernel_shared[((((int)threadIdx.x) * 72) + 18)]));
+        conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[55] * kernel_shared[((((int)threadIdx.x) * 72) + 18)]));
+        conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[56] * kernel_shared[((((int)threadIdx.x) * 72) + 18)]));
+        conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[57] * kernel_shared[((((int)threadIdx.x) * 72) + 18)]));
+        conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[58] * kernel_shared[((((int)threadIdx.x) * 72) + 18)]));
+        conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[59] * kernel_shared[((((int)threadIdx.x) * 72) + 18)]));
+        conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[60] * kernel_shared[((((int)threadIdx.x) * 72) + 18)]));
+        conv2d_nchw[7] = (conv2d_nchw[7] + (pad_temp_shared[54] * kernel_shared[((((int)threadIdx.x) * 72) + 4626)]));
+        conv2d_nchw[8] = (conv2d_nchw[8] + (pad_temp_shared[55] * kernel_shared[((((int)threadIdx.x) * 72) + 4626)]));
+        conv2d_nchw[9] = (conv2d_nchw[9] + (pad_temp_shared[56] * kernel_shared[((((int)threadIdx.x) * 72) + 4626)]));
+        conv2d_nchw[10] = (conv2d_nchw[10] + (pad_temp_shared[57] * kernel_shared[((((int)threadIdx.x) * 72) + 4626)]));
+        conv2d_nchw[11] = (conv2d_nchw[11] + (pad_temp_shared[58] * kernel_shared[((((int)threadIdx.x) * 72) + 4626)]));
+        conv2d_nchw[12] = (conv2d_nchw[12] + (pad_temp_shared[59] * kernel_shared[((((int)threadIdx.x) * 72) + 4626)]));
+        conv2d_nchw[13] = (conv2d_nchw[13] + (pad_temp_shared[60] * kernel_shared[((((int)threadIdx.x) * 72) + 4626)]));
+        conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[55] * kernel_shared[((((int)threadIdx.x) * 72) + 19)]));
+        conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[56] * kernel_shared[((((int)threadIdx.x) * 72) + 19)]));
+        conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[57] * kernel_shared[((((int)threadIdx.x) * 72) + 19)]));
+        conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[58] * kernel_shared[((((int)threadIdx.x) * 72) + 19)]));
+        conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[59] * kernel_shared[((((int)threadIdx.x) * 72) + 19)]));
+        conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[60] * kernel_shared[((((int)threadIdx.x) * 72) + 19)]));
+        conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[61] * kernel_shared[((((int)threadIdx.x) * 72) + 19)]));
+        conv2d_nchw[7] = (conv2d_nchw[7] + (pad_temp_shared[55] * kernel_shared[((((int)threadIdx.x) * 72) + 4627)]));
+        conv2d_nchw[8] = (conv2d_nchw[8] + (pad_temp_shared[56] * kernel_shared[((((int)threadIdx.x) * 72) + 4627)]));
+        conv2d_nchw[9] = (conv2d_nchw[9] + (pad_temp_shared[57] * kernel_shared[((((int)threadIdx.x) * 72) + 4627)]));
+        conv2d_nchw[10] = (conv2d_nchw[10] + (pad_temp_shared[58] * kernel_shared[((((int)threadIdx.x) * 72) + 4627)]));
+        conv2d_nchw[11] = (conv2d_nchw[11] + (pad_temp_shared[59] * kernel_shared[((((int)threadIdx.x) * 72) + 4627)]));
+        conv2d_nchw[12] = (conv2d_nchw[12] + (pad_temp_shared[60] * kernel_shared[((((int)threadIdx.x) * 72) + 4627)]));
+        conv2d_nchw[13] = (conv2d_nchw[13] + (pad_temp_shared[61] * kernel_shared[((((int)threadIdx.x) * 72) + 4627)]));
+        conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[56] * kernel_shared[((((int)threadIdx.x) * 72) + 20)]));
+        conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[57] * kernel_shared[((((int)threadIdx.x) * 72) + 20)]));
+        conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[58] * kernel_shared[((((int)threadIdx.x) * 72) + 20)]));
+        conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[59] * kernel_shared[((((int)threadIdx.x) * 72) + 20)]));
+        conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[60] * kernel_shared[((((int)threadIdx.x) * 72) + 20)]));
+        conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[61] * kernel_shared[((((int)threadIdx.x) * 72) + 20)]));
+        conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[62] * kernel_shared[((((int)threadIdx.x) * 72) + 20)]));
+        conv2d_nchw[7] = (conv2d_nchw[7] + (pad_temp_shared[56] * kernel_shared[((((int)threadIdx.x) * 72) + 4628)]));
+        conv2d_nchw[8] = (conv2d_nchw[8] + (pad_temp_shared[57] * kernel_shared[((((int)threadIdx.x) * 72) + 4628)]));
+        conv2d_nchw[9] = (conv2d_nchw[9] + (pad_temp_shared[58] * kernel_shared[((((int)threadIdx.x) * 72) + 4628)]));
+        conv2d_nchw[10] = (conv2d_nchw[10] + (pad_temp_shared[59] * kernel_shared[((((int)threadIdx.x) * 72) + 4628)]));
+        conv2d_nchw[11] = (conv2d_nchw[11] + (pad_temp_shared[60] * kernel_shared[((((int)threadIdx.x) * 72) + 4628)]));
+        conv2d_nchw[12] = (conv2d_nchw[12] + (pad_temp_shared[61] * kernel_shared[((((int)threadIdx.x) * 72) + 4628)]));
+        conv2d_nchw[13] = (conv2d_nchw[13] + (pad_temp_shared[62] * kernel_shared[((((int)threadIdx.x) * 72) + 4628)]));
+        conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[63] * kernel_shared[((((int)threadIdx.x) * 72) + 21)]));
+        conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[64] * kernel_shared[((((int)threadIdx.x) * 72) + 21)]));
+        conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[65] * kernel_shared[((((int)threadIdx.x) * 72) + 21)]));
+        conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[66] * kernel_shared[((((int)threadIdx.x) * 72) + 21)]));
+        conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[67] * kernel_shared[((((int)threadIdx.x) * 72) + 21)]));
+        conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[68] * kernel_shared[((((int)threadIdx.x) * 72) + 21)]));
+        conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[69] * kernel_shared[((((int)threadIdx.x) * 72) + 21)]));
+        conv2d_nchw[7] = (conv2d_nchw[7] + (pad_temp_shared[63] * kernel_shared[((((int)threadIdx.x) * 72) + 4629)]));
+        conv2d_nchw[8] = (conv2d_nchw[8] + (pad_temp_shared[64] * kernel_shared[((((int)threadIdx.x) * 72) + 4629)]));
+        conv2d_nchw[9] = (conv2d_nchw[9] + (pad_temp_shared[65] * kernel_shared[((((int)threadIdx.x) * 72) + 4629)]));
+        conv2d_nchw[10] = (conv2d_nchw[10] + (pad_temp_shared[66] * kernel_shared[((((int)threadIdx.x) * 72) + 4629)]));
+        conv2d_nchw[11] = (conv2d_nchw[11] + (pad_temp_shared[67] * kernel_shared[((((int)threadIdx.x) * 72) + 4629)]));
+        conv2d_nchw[12] = (conv2d_nchw[12] + (pad_temp_shared[68] * kernel_shared[((((int)threadIdx.x) * 72) + 4629)]));
+        conv2d_nchw[13] = (conv2d_nchw[13] + (pad_temp_shared[69] * kernel_shared[((((int)threadIdx.x) * 72) + 4629)]));
+        conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[64] * kernel_shared[((((int)threadIdx.x) * 72) + 22)]));
+        conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[65] * kernel_shared[((((int)threadIdx.x) * 72) + 22)]));
+        conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[66] * kernel_shared[((((int)threadIdx.x) * 72) + 22)]));
+        conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[67] * kernel_shared[((((int)threadIdx.x) * 72) + 22)]));
+        conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[68] * kernel_shared[((((int)threadIdx.x) * 72) + 22)]));
+        conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[69] * kernel_shared[((((int)threadIdx.x) * 72) + 22)]));
+        conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[70] * kernel_shared[((((int)threadIdx.x) * 72) + 22)]));
+        conv2d_nchw[7] = (conv2d_nchw[7] + (pad_temp_shared[64] * kernel_shared[((((int)threadIdx.x) * 72) + 4630)]));
+        conv2d_nchw[8] = (conv2d_nchw[8] + (pad_temp_shared[65] * kernel_shared[((((int)threadIdx.x) * 72) + 4630)]));
+        conv2d_nchw[9] = (conv2d_nchw[9] + (pad_temp_shared[66] * kernel_shared[((((int)threadIdx.x) * 72) + 4630)]));
+        conv2d_nchw[10] = (conv2d_nchw[10] + (pad_temp_shared[67] * kernel_shared[((((int)threadIdx.x) * 72) + 4630)]));
+        conv2d_nchw[11] = (conv2d_nchw[11] + (pad_temp_shared[68] * kernel_shared[((((int)threadIdx.x) * 72) + 4630)]));
+        conv2d_nchw[12] = (conv2d_nchw[12] + (pad_temp_shared[69] * kernel_shared[((((int)threadIdx.x) * 72) + 4630)]));
+        conv2d_nchw[13] = (conv2d_nchw[13] + (pad_temp_shared[70] * kernel_shared[((((int)threadIdx.x) * 72) + 4630)]));
+        conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[65] * kernel_shared[((((int)threadIdx.x) * 72) + 23)]));
+        conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[66] * kernel_shared[((((int)threadIdx.x) * 72) + 23)]));
+        conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[67] * kernel_shared[((((int)threadIdx.x) * 72) + 23)]));
+        conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[68] * kernel_shared[((((int)threadIdx.x) * 72) + 23)]));
+        conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[69] * kernel_shared[((((int)threadIdx.x) * 72) + 23)]));
+        conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[70] * kernel_shared[((((int)threadIdx.x) * 72) + 23)]));
+        conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[71] * kernel_shared[((((int)threadIdx.x) * 72) + 23)]));
+        conv2d_nchw[7] = (conv2d_nchw[7] + (pad_temp_shared[65] * kernel_shared[((((int)threadIdx.x) * 72) + 4631)]));
+        conv2d_nchw[8] = (conv2d_nchw[8] + (pad_temp_shared[66] * kernel_shared[((((int)threadIdx.x) * 72) + 4631)]));
+        conv2d_nchw[9] = (conv2d_nchw[9] + (pad_temp_shared[67] * kernel_shared[((((int)threadIdx.x) * 72) + 4631)]));
+        conv2d_nchw[10] = (conv2d_nchw[10] + (pad_temp_shared[68] * kernel_shared[((((int)threadIdx.x) * 72) + 4631)]));
+        conv2d_nchw[11] = (conv2d_nchw[11] + (pad_temp_shared[69] * kernel_shared[((((int)threadIdx.x) * 72) + 4631)]));
+        conv2d_nchw[12] = (conv2d_nchw[12] + (pad_temp_shared[70] * kernel_shared[((((int)threadIdx.x) * 72) + 4631)]));
+        conv2d_nchw[13] = (conv2d_nchw[13] + (pad_temp_shared[71] * kernel_shared[((((int)threadIdx.x) * 72) + 4631)]));
+        conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[72] * kernel_shared[((((int)threadIdx.x) * 72) + 24)]));
+        conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[73] * kernel_shared[((((int)threadIdx.x) * 72) + 24)]));
+        conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[74] * kernel_shared[((((int)threadIdx.x) * 72) + 24)]));
+        conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[75] * kernel_shared[((((int)threadIdx.x) * 72) + 24)]));
+        conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[76] * kernel_shared[((((int)threadIdx.x) * 72) + 24)]));
+        conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[77] * kernel_shared[((((int)threadIdx.x) * 72) + 24)]));
+        conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[78] * kernel_shared[((((int)threadIdx.x) * 72) + 24)]));
+        conv2d_nchw[7] = (conv2d_nchw[7] + (pad_temp_shared[72] * kernel_shared[((((int)threadIdx.x) * 72) + 4632)]));
+        conv2d_nchw[8] = (conv2d_nchw[8] + (pad_temp_shared[73] * kernel_shared[((((int)threadIdx.x) * 72) + 4632)]));
+        conv2d_nchw[9] = (conv2d_nchw[9] + (pad_temp_shared[74] * kernel_shared[((((int)threadIdx.x) * 72) + 4632)]));
+        conv2d_nchw[10] = (conv2d_nchw[10] + (pad_temp_shared[75] * kernel_shared[((((int)threadIdx.x) * 72) + 4632)]));
+        conv2d_nchw[11] = (conv2d_nchw[11] + (pad_temp_shared[76] * kernel_shared[((((int)threadIdx.x) * 72) + 4632)]));
+        conv2d_nchw[12] = (conv2d_nchw[12] + (pad_temp_shared[77] * kernel_shared[((((int)threadIdx.x) * 72) + 4632)]));
+        conv2d_nchw[13] = (conv2d_nchw[13] + (pad_temp_shared[78] * kernel_shared[((((int)threadIdx.x) * 72) + 4632)]));
+        conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[73] * kernel_shared[((((int)threadIdx.x) * 72) + 25)]));
+        conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[74] * kernel_shared[((((int)threadIdx.x) * 72) + 25)]));
+        conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[75] * kernel_shared[((((int)threadIdx.x) * 72) + 25)]));
+        conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[76] * kernel_shared[((((int)threadIdx.x) * 72) + 25)]));
+        conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[77] * kernel_shared[((((int)threadIdx.x) * 72) + 25)]));
+        conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[78] * kernel_shared[((((int)threadIdx.x) * 72) + 25)]));
+        conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[79] * kernel_shared[((((int)threadIdx.x) * 72) + 25)]));
+        conv2d_nchw[7] = (conv2d_nchw[7] + (pad_temp_shared[73] * kernel_shared[((((int)threadIdx.x) * 72) + 4633)]));
+        conv2d_nchw[8] = (conv2d_nchw[8] + (pad_temp_shared[74] * kernel_shared[((((int)threadIdx.x) * 72) + 4633)]));
+        conv2d_nchw[9] = (conv2d_nchw[9] + (pad_temp_shared[75] * kernel_shared[((((int)threadIdx.x) * 72) + 4633)]));
+        conv2d_nchw[10] = (conv2d_nchw[10] + (pad_temp_shared[76] * kernel_shared[((((int)threadIdx.x) * 72) + 4633)]));
+        conv2d_nchw[11] = (conv2d_nchw[11] + (pad_temp_shared[77] * kernel_shared[((((int)threadIdx.x) * 72) + 4633)]));
+        conv2d_nchw[12] = (conv2d_nchw[12] + (pad_temp_shared[78] * kernel_shared[((((int)threadIdx.x) * 72) + 4633)]));
+        conv2d_nchw[13] = (conv2d_nchw[13] + (pad_temp_shared[79] * kernel_shared[((((int)threadIdx.x) * 72) + 4633)]));
+        conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[74] * kernel_shared[((((int)threadIdx.x) * 72) + 26)]));
+        conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[75] * kernel_shared[((((int)threadIdx.x) * 72) + 26)]));
+        conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[76] * kernel_shared[((((int)threadIdx.x) * 72) + 26)]));
+        conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[77] * kernel_shared[((((int)threadIdx.x) * 72) + 26)]));
+        conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[78] * kernel_shared[((((int)threadIdx.x) * 72) + 26)]));
+        conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[79] * kernel_shared[((((int)threadIdx.x) * 72) + 26)]));
+        conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[80] * kernel_shared[((((int)threadIdx.x) * 72) + 26)]));
+        conv2d_nchw[7] = (conv2d_nchw[7] + (pad_temp_shared[74] * kernel_shared[((((int)threadIdx.x) * 72) + 4634)]));
+        conv2d_nchw[8] = (conv2d_nchw[8] + (pad_temp_shared[75] * kernel_shared[((((int)threadIdx.x) * 72) + 4634)]));
+        conv2d_nchw[9] = (conv2d_nchw[9] + (pad_temp_shared[76] * kernel_shared[((((int)threadIdx.x) * 72) + 4634)]));
+        conv2d_nchw[10] = (conv2d_nchw[10] + (pad_temp_shared[77] * kernel_shared[((((int)threadIdx.x) * 72) + 4634)]));
+        conv2d_nchw[11] = (conv2d_nchw[11] + (pad_temp_shared[78] * kernel_shared[((((int)threadIdx.x) * 72) + 4634)]));
+        conv2d_nchw[12] = (conv2d_nchw[12] + (pad_temp_shared[79] * kernel_shared[((((int)threadIdx.x) * 72) + 4634)]));
+        conv2d_nchw[13] = (conv2d_nchw[13] + (pad_temp_shared[80] * kernel_shared[((((int)threadIdx.x) * 72) + 4634)]));
+        conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[81] * kernel_shared[((((int)threadIdx.x) * 72) + 27)]));
+        conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[82] * kernel_shared[((((int)threadIdx.x) * 72) + 27)]));
+        conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[83] * kernel_shared[((((int)threadIdx.x) * 72) + 27)]));
+        conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[84] * kernel_shared[((((int)threadIdx.x) * 72) + 27)]));
+        conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[85] * kernel_shared[((((int)threadIdx.x) * 72) + 27)]));
+        conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[86] * kernel_shared[((((int)threadIdx.x) * 72) + 27)]));
+        conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[87] * kernel_shared[((((int)threadIdx.x) * 72) + 27)]));
+        conv2d_nchw[7] = (conv2d_nchw[7] + (pad_temp_shared[81] * kernel_shared[((((int)threadIdx.x) * 72) + 4635)]));
+        conv2d_nchw[8] = (conv2d_nchw[8] + (pad_temp_shared[82] * kernel_shared[((((int)threadIdx.x) * 72) + 4635)]));
+        conv2d_nchw[9] = (conv2d_nchw[9] + (pad_temp_shared[83] * kernel_shared[((((int)threadIdx.x) * 72) + 4635)]));
+        conv2d_nchw[10] = (conv2d_nchw[10] + (pad_temp_shared[84] * kernel_shared[((((int)threadIdx.x) * 72) + 4635)]));
+        conv2d_nchw[11] = (conv2d_nchw[11] + (pad_temp_shared[85] * kernel_shared[((((int)threadIdx.x) * 72) + 4635)]));
+        conv2d_nchw[12] = (conv2d_nchw[12] + (pad_temp_shared[86] * kernel_shared[((((int)threadIdx.x) * 72) + 4635)]));
+        conv2d_nchw[13] = (conv2d_nchw[13] + (pad_temp_shared[87] * kernel_shared[((((int)threadIdx.x) * 72) + 4635)]));
+        conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[82] * kernel_shared[((((int)threadIdx.x) * 72) + 28)]));
+        conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[83] * kernel_shared[((((int)threadIdx.x) * 72) + 28)]));
+        conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[84] * kernel_shared[((((int)threadIdx.x) * 72) + 28)]));
+        conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[85] * kernel_shared[((((int)threadIdx.x) * 72) + 28)]));
+        conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[86] * kernel_shared[((((int)threadIdx.x) * 72) + 28)]));
+        conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[87] * kernel_shared[((((int)threadIdx.x) * 72) + 28)]));
+        conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[88] * kernel_shared[((((int)threadIdx.x) * 72) + 28)]));
+        conv2d_nchw[7] = (conv2d_nchw[7] + (pad_temp_shared[82] * kernel_shared[((((int)threadIdx.x) * 72) + 4636)]));
+        conv2d_nchw[8] = (conv2d_nchw[8] + (pad_temp_shared[83] * kernel_shared[((((int)threadIdx.x) * 72) + 4636)]));
+        conv2d_nchw[9] = (conv2d_nchw[9] + (pad_temp_shared[84] * kernel_shared[((((int)threadIdx.x) * 72) + 4636)]));
+        conv2d_nchw[10] = (conv2d_nchw[10] + (pad_temp_shared[85] * kernel_shared[((((int)threadIdx.x) * 72) + 4636)]));
+        conv2d_nchw[11] = (conv2d_nchw[11] + (pad_temp_shared[86] * kernel_shared[((((int)threadIdx.x) * 72) + 4636)]));
+        conv2d_nchw[12] = (conv2d_nchw[12] + (pad_temp_shared[87] * kernel_shared[((((int)threadIdx.x) * 72) + 4636)]));
+        conv2d_nchw[13] = (conv2d_nchw[13] + (pad_temp_shared[88] * kernel_shared[((((int)threadIdx.x) * 72) + 4636)]));
+        conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[83] * kernel_shared[((((int)threadIdx.x) * 72) + 29)]));
+        conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[84] * kernel_shared[((((int)threadIdx.x) * 72) + 29)]));
+        conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[85] * kernel_shared[((((int)threadIdx.x) * 72) + 29)]));
+        conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[86] * kernel_shared[((((int)threadIdx.x) * 72) + 29)]));
+        conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[87] * kernel_shared[((((int)threadIdx.x) * 72) + 29)]));
+        conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[88] * kernel_shared[((((int)threadIdx.x) * 72) + 29)]));
+        conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[89] * kernel_shared[((((int)threadIdx.x) * 72) + 29)]));
+        conv2d_nchw[7] = (conv2d_nchw[7] + (pad_temp_shared[83] * kernel_shared[((((int)threadIdx.x) * 72) + 4637)]));
+        conv2d_nchw[8] = (conv2d_nchw[8] + (pad_temp_shared[84] * kernel_shared[((((int)threadIdx.x) * 72) + 4637)]));
+        conv2d_nchw[9] = (conv2d_nchw[9] + (pad_temp_shared[85] * kernel_shared[((((int)threadIdx.x) * 72) + 4637)]));
+        conv2d_nchw[10] = (conv2d_nchw[10] + (pad_temp_shared[86] * kernel_shared[((((int)threadIdx.x) * 72) + 4637)]));
+        conv2d_nchw[11] = (conv2d_nchw[11] + (pad_temp_shared[87] * kernel_shared[((((int)threadIdx.x) * 72) + 4637)]));
+        conv2d_nchw[12] = (conv2d_nchw[12] + (pad_temp_shared[88] * kernel_shared[((((int)threadIdx.x) * 72) + 4637)]));
+        conv2d_nchw[13] = (conv2d_nchw[13] + (pad_temp_shared[89] * kernel_shared[((((int)threadIdx.x) * 72) + 4637)]));
+        conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[90] * kernel_shared[((((int)threadIdx.x) * 72) + 30)]));
+        conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[91] * kernel_shared[((((int)threadIdx.x) * 72) + 30)]));
+        conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[92] * kernel_shared[((((int)threadIdx.x) * 72) + 30)]));
+        conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[93] * kernel_shared[((((int)threadIdx.x) * 72) + 30)]));
+        conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[94] * kernel_shared[((((int)threadIdx.x) * 72) + 30)]));
+        conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[95] * kernel_shared[((((int)threadIdx.x) * 72) + 30)]));
+        conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[96] * kernel_shared[((((int)threadIdx.x) * 72) + 30)]));
+        conv2d_nchw[7] = (conv2d_nchw[7] + (pad_temp_shared[90] * kernel_shared[((((int)threadIdx.x) * 72) + 4638)]));
+        conv2d_nchw[8] = (conv2d_nchw[8] + (pad_temp_shared[91] * kernel_shared[((((int)threadIdx.x) * 72) + 4638)]));
+        conv2d_nchw[9] = (conv2d_nchw[9] + (pad_temp_shared[92] * kernel_shared[((((int)threadIdx.x) * 72) + 4638)]));
+        conv2d_nchw[10] = (conv2d_nchw[10] + (pad_temp_shared[93] * kernel_shared[((((int)threadIdx.x) * 72) + 4638)]));
+        conv2d_nchw[11] = (conv2d_nchw[11] + (pad_temp_shared[94] * kernel_shared[((((int)threadIdx.x) * 72) + 4638)]));
+        conv2d_nchw[12] = (conv2d_nchw[12] + (pad_temp_shared[95] * kernel_shared[((((int)threadIdx.x) * 72) + 4638)]));
+        conv2d_nchw[13] = (conv2d_nchw[13] + (pad_temp_shared[96] * kernel_shared[((((int)threadIdx.x) * 72) + 4638)]));
+        conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[91] * kernel_shared[((((int)threadIdx.x) * 72) + 31)]));
+        conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[92] * kernel_shared[((((int)threadIdx.x) * 72) + 31)]));
+        conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[93] * kernel_shared[((((int)threadIdx.x) * 72) + 31)]));
+        conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[94] * kernel_shared[((((int)threadIdx.x) * 72) + 31)]));
+        conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[95] * kernel_shared[((((int)threadIdx.x) * 72) + 31)]));
+        conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[96] * kernel_shared[((((int)threadIdx.x) * 72) + 31)]));
+        conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[97] * kernel_shared[((((int)threadIdx.x) * 72) + 31)]));
+        conv2d_nchw[7] = (conv2d_nchw[7] + (pad_temp_shared[91] * kernel_shared[((((int)threadIdx.x) * 72) + 4639)]));
+        conv2d_nchw[8] = (conv2d_nchw[8] + (pad_temp_shared[92] * kernel_shared[((((int)threadIdx.x) * 72) + 4639)]));
+        conv2d_nchw[9] = (conv2d_nchw[9] + (pad_temp_shared[93] * kernel_shared[((((int)threadIdx.x) * 72) + 4639)]));
+        conv2d_nchw[10] = (conv2d_nchw[10] + (pad_temp_shared[94] * kernel_shared[((((int)threadIdx.x) * 72) + 4639)]));
+        conv2d_nchw[11] = (conv2d_nchw[11] + (pad_temp_shared[95] * kernel_shared[((((int)threadIdx.x) * 72) + 4639)]));
+        conv2d_nchw[12] = (conv2d_nchw[12] + (pad_temp_shared[96] * kernel_shared[((((int)threadIdx.x) * 72) + 4639)]));
+        conv2d_nchw[13] = (conv2d_nchw[13] + (pad_temp_shared[97] * kernel_shared[((((int)threadIdx.x) * 72) + 4639)]));
+        conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[92] * kernel_shared[((((int)threadIdx.x) * 72) + 32)]));
+        conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[93] * kernel_shared[((((int)threadIdx.x) * 72) + 32)]));
+        conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[94] * kernel_shared[((((int)threadIdx.x) * 72) + 32)]));
+        conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[95] * kernel_shared[((((int)threadIdx.x) * 72) + 32)]));
+        conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[96] * kernel_shared[((((int)threadIdx.x) * 72) + 32)]));
+        conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[97] * kernel_shared[((((int)threadIdx.x) * 72) + 32)]));
+        conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[98] * kernel_shared[((((int)threadIdx.x) * 72) + 32)]));
+        conv2d_nchw[7] = (conv2d_nchw[7] + (pad_temp_shared[92] * kernel_shared[((((int)threadIdx.x) * 72) + 4640)]));
+        conv2d_nchw[8] = (conv2d_nchw[8] + (pad_temp_shared[93] * kernel_shared[((((int)threadIdx.x) * 72) + 4640)]));
+        conv2d_nchw[9] = (conv2d_nchw[9] + (pad_temp_shared[94] * kernel_shared[((((int)threadIdx.x) * 72) + 4640)]));
+        conv2d_nchw[10] = (conv2d_nchw[10] + (pad_temp_shared[95] * kernel_shared[((((int)threadIdx.x) * 72) + 4640)]));
+        conv2d_nchw[11] = (conv2d_nchw[11] + (pad_temp_shared[96] * kernel_shared[((((int)threadIdx.x) * 72) + 4640)]));
+        conv2d_nchw[12] = (conv2d_nchw[12] + (pad_temp_shared[97] * kernel_shared[((((int)threadIdx.x) * 72) + 4640)]));
+        conv2d_nchw[13] = (conv2d_nchw[13] + (pad_temp_shared[98] * kernel_shared[((((int)threadIdx.x) * 72) + 4640)]));
+        conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[99] * kernel_shared[((((int)threadIdx.x) * 72) + 33)]));
+        conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[100] * kernel_shared[((((int)threadIdx.x) * 72) + 33)]));
+        conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[101] * kernel_shared[((((int)threadIdx.x) * 72) + 33)]));
+        conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[102] * kernel_shared[((((int)threadIdx.x) * 72) + 33)]));
+        conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[103] * kernel_shared[((((int)threadIdx.x) * 72) + 33)]));
+        conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[104] * kernel_shared[((((int)threadIdx.x) * 72) + 33)]));
+        conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[105] * kernel_shared[((((int)threadIdx.x) * 72) + 33)]));
+        conv2d_nchw[7] = (conv2d_nchw[7] + (pad_temp_shared[99] * kernel_shared[((((int)threadIdx.x) * 72) + 4641)]));
+        conv2d_nchw[8] = (conv2d_nchw[8] + (pad_temp_shared[100] * kernel_shared[((((int)threadIdx.x) * 72) + 4641)]));
+        conv2d_nchw[9] = (conv2d_nchw[9] + (pad_temp_shared[101] * kernel_shared[((((int)threadIdx.x) * 72) + 4641)]));
+        conv2d_nchw[10] = (conv2d_nchw[10] + (pad_temp_shared[102] * kernel_shared[((((int)threadIdx.x) * 72) + 4641)]));
+        conv2d_nchw[11] = (conv2d_nchw[11] + (pad_temp_shared[103] * kernel_shared[((((int)threadIdx.x) * 72) + 4641)]));
+        conv2d_nchw[12] = (conv2d_nchw[12] + (pad_temp_shared[104] * kernel_shared[((((int)threadIdx.x) * 72) + 4641)]));
+        conv2d_nchw[13] = (conv2d_nchw[13] + (pad_temp_shared[105] * kernel_shared[((((int)threadIdx.x) * 72) + 4641)]));
+        conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[100] * kernel_shared[((((int)threadIdx.x) * 72) + 34)]));
+        conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[101] * kernel_shared[((((int)threadIdx.x) * 72) + 34)]));
+        conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[102] * kernel_shared[((((int)threadIdx.x) * 72) + 34)]));
+        conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[103] * kernel_shared[((((int)threadIdx.x) * 72) + 34)]));
+        conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[104] * kernel_shared[((((int)threadIdx.x) * 72) + 34)]));
+        conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[105] * kernel_shared[((((int)threadIdx.x) * 72) + 34)]));
+        conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[106] * kernel_shared[((((int)threadIdx.x) * 72) + 34)]));
+        conv2d_nchw[7] = (conv2d_nchw[7] + (pad_temp_shared[100] * kernel_shared[((((int)threadIdx.x) * 72) + 4642)]));
+        conv2d_nchw[8] = (conv2d_nchw[8] + (pad_temp_shared[101] * kernel_shared[((((int)threadIdx.x) * 72) + 4642)]));
+        conv2d_nchw[9] = (conv2d_nchw[9] + (pad_temp_shared[102] * kernel_shared[((((int)threadIdx.x) * 72) + 4642)]));
+        conv2d_nchw[10] = (conv2d_nchw[10] + (pad_temp_shared[103] * kernel_shared[((((int)threadIdx.x) * 72) + 4642)]));
+        conv2d_nchw[11] = (conv2d_nchw[11] + (pad_temp_shared[104] * kernel_shared[((((int)threadIdx.x) * 72) + 4642)]));
+        conv2d_nchw[12] = (conv2d_nchw[12] + (pad_temp_shared[105] * kernel_shared[((((int)threadIdx.x) * 72) + 4642)]));
+        conv2d_nchw[13] = (conv2d_nchw[13] + (pad_temp_shared[106] * kernel_shared[((((int)threadIdx.x) * 72) + 4642)]));
+        conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[101] * kernel_shared[((((int)threadIdx.x) * 72) + 35)]));
+        conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[102] * kernel_shared[((((int)threadIdx.x) * 72) + 35)]));
+        conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[103] * kernel_shared[((((int)threadIdx.x) * 72) + 35)]));
+        conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[104] * kernel_shared[((((int)threadIdx.x) * 72) + 35)]));
+        conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[105] * kernel_shared[((((int)threadIdx.x) * 72) + 35)]));
+        conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[106] * kernel_shared[((((int)threadIdx.x) * 72) + 35)]));
+        conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[107] * kernel_shared[((((int)threadIdx.x) * 72) + 35)]));
+        conv2d_nchw[7] = (conv2d_nchw[7] + (pad_temp_shared[101] * kernel_shared[((((int)threadIdx.x) * 72) + 4643)]));
+        conv2d_nchw[8] = (conv2d_nchw[8] + (pad_temp_shared[102] * kernel_shared[((((int)threadIdx.x) * 72) + 4643)]));
+        conv2d_nchw[9] = (conv2d_nchw[9] + (pad_temp_shared[103] * kernel_shared[((((int)threadIdx.x) * 72) + 4643)]));
+        conv2d_nchw[10] = (conv2d_nchw[10] + (pad_temp_shared[104] * kernel_shared[((((int)threadIdx.x) * 72) + 4643)]));
+        conv2d_nchw[11] = (conv2d_nchw[11] + (pad_temp_shared[105] * kernel_shared[((((int)threadIdx.x) * 72) + 4643)]));
+        conv2d_nchw[12] = (conv2d_nchw[12] + (pad_temp_shared[106] * kernel_shared[((((int)threadIdx.x) * 72) + 4643)]));
+        conv2d_nchw[13] = (conv2d_nchw[13] + (pad_temp_shared[107] * kernel_shared[((((int)threadIdx.x) * 72) + 4643)]));
+        conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[108] * kernel_shared[((((int)threadIdx.x) * 72) + 36)]));
+        conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[109] * kernel_shared[((((int)threadIdx.x) * 72) + 36)]));
+        conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[110] * kernel_shared[((((int)threadIdx.x) * 72) + 36)]));
+        conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[111] * kernel_shared[((((int)threadIdx.x) * 72) + 36)]));
+        conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[112] * kernel_shared[((((int)threadIdx.x) * 72) + 36)]));
+        conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[113] * kernel_shared[((((int)threadIdx.x) * 72) + 36)]));
+        conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[114] * kernel_shared[((((int)threadIdx.x) * 72) + 36)]));
+        conv2d_nchw[7] = (conv2d_nchw[7] + (pad_temp_shared[108] * kernel_shared[((((int)threadIdx.x) * 72) + 4644)]));
+        conv2d_nchw[8] = (conv2d_nchw[8] + (pad_temp_shared[109] * kernel_shared[((((int)threadIdx.x) * 72) + 4644)]));
+        conv2d_nchw[9] = (conv2d_nchw[9] + (pad_temp_shared[110] * kernel_shared[((((int)threadIdx.x) * 72) + 4644)]));
+        conv2d_nchw[10] = (conv2d_nchw[10] + (pad_temp_shared[111] * kernel_shared[((((int)threadIdx.x) * 72) + 4644)]));
+        conv2d_nchw[11] = (conv2d_nchw[11] + (pad_temp_shared[112] * kernel_shared[((((int)threadIdx.x) * 72) + 4644)]));
+        conv2d_nchw[12] = (conv2d_nchw[12] + (pad_temp_shared[113] * kernel_shared[((((int)threadIdx.x) * 72) + 4644)]));
+        conv2d_nchw[13] = (conv2d_nchw[13] + (pad_temp_shared[114] * kernel_shared[((((int)threadIdx.x) * 72) + 4644)]));
+        conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[109] * kernel_shared[((((int)threadIdx.x) * 72) + 37)]));
+        conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[110] * kernel_shared[((((int)threadIdx.x) * 72) + 37)]));
+        conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[111] * kernel_shared[((((int)threadIdx.x) * 72) + 37)]));
+        conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[112] * kernel_shared[((((int)threadIdx.x) * 72) + 37)]));
+        conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[113] * kernel_shared[((((int)threadIdx.x) * 72) + 37)]));
+        conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[114] * kernel_shared[((((int)threadIdx.x) * 72) + 37)]));
+        conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[115] * kernel_shared[((((int)threadIdx.x) * 72) + 37)]));
+        conv2d_nchw[7] = (conv2d_nchw[7] + (pad_temp_shared[109] * kernel_shared[((((int)threadIdx.x) * 72) + 4645)]));
+        conv2d_nchw[8] = (conv2d_nchw[8] + (pad_temp_shared[110] * kernel_shared[((((int)threadIdx.x) * 72) + 4645)]));
+        conv2d_nchw[9] = (conv2d_nchw[9] + (pad_temp_shared[111] * kernel_shared[((((int)threadIdx.x) * 72) + 4645)]));
+        conv2d_nchw[10] = (conv2d_nchw[10] + (pad_temp_shared[112] * kernel_shared[((((int)threadIdx.x) * 72) + 4645)]));
+        conv2d_nchw[11] = (conv2d_nchw[11] + (pad_temp_shared[113] * kernel_shared[((((int)threadIdx.x) * 72) + 4645)]));
+        conv2d_nchw[12] = (conv2d_nchw[12] + (pad_temp_shared[114] * kernel_shared[((((int)threadIdx.x) * 72) + 4645)]));
+        conv2d_nchw[13] = (conv2d_nchw[13] + (pad_temp_shared[115] * kernel_shared[((((int)threadIdx.x) * 72) + 4645)]));
+        conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[110] * kernel_shared[((((int)threadIdx.x) * 72) + 38)]));
+        conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[111] * kernel_shared[((((int)threadIdx.x) * 72) + 38)]));
+        conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[112] * kernel_shared[((((int)threadIdx.x) * 72) + 38)]));
+        conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[113] * kernel_shared[((((int)threadIdx.x) * 72) + 38)]));
+        conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[114] * kernel_shared[((((int)threadIdx.x) * 72) + 38)]));
+        conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[115] * kernel_shared[((((int)threadIdx.x) * 72) + 38)]));
+        conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[116] * kernel_shared[((((int)threadIdx.x) * 72) + 38)]));
+        conv2d_nchw[7] = (conv2d_nchw[7] + (pad_temp_shared[110] * kernel_shared[((((int)threadIdx.x) * 72) + 4646)]));
+        conv2d_nchw[8] = (conv2d_nchw[8] + (pad_temp_shared[111] * kernel_shared[((((int)threadIdx.x) * 72) + 4646)]));
+        conv2d_nchw[9] = (conv2d_nchw[9] + (pad_temp_shared[112] * kernel_shared[((((int)threadIdx.x) * 72) + 4646)]));
+        conv2d_nchw[10] = (conv2d_nchw[10] + (pad_temp_shared[113] * kernel_shared[((((int)threadIdx.x) * 72) + 4646)]));
+        conv2d_nchw[11] = (conv2d_nchw[11] + (pad_temp_shared[114] * kernel_shared[((((int)threadIdx.x) * 72) + 4646)]));
+        conv2d_nchw[12] = (conv2d_nchw[12] + (pad_temp_shared[115] * kernel_shared[((((int)threadIdx.x) * 72) + 4646)]));
+        conv2d_nchw[13] = (conv2d_nchw[13] + (pad_temp_shared[116] * kernel_shared[((((int)threadIdx.x) * 72) + 4646)]));
+        conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[117] * kernel_shared[((((int)threadIdx.x) * 72) + 39)]));
+        conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[118] * kernel_shared[((((int)threadIdx.x) * 72) + 39)]));
+        conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[119] * kernel_shared[((((int)threadIdx.x) * 72) + 39)]));
+        conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[120] * kernel_shared[((((int)threadIdx.x) * 72) + 39)]));
+        conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[121] * kernel_shared[((((int)threadIdx.x) * 72) + 39)]));
+        conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[122] * kernel_shared[((((int)threadIdx.x) * 72) + 39)]));
+        conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[123] * kernel_shared[((((int)threadIdx.x) * 72) + 39)]));
+        conv2d_nchw[7] = (conv2d_nchw[7] + (pad_temp_shared[117] * kernel_shared[((((int)threadIdx.x) * 72) + 4647)]));
+        conv2d_nchw[8] = (conv2d_nchw[8] + (pad_temp_shared[118] * kernel_shared[((((int)threadIdx.x) * 72) + 4647)]));
+        conv2d_nchw[9] = (conv2d_nchw[9] + (pad_temp_shared[119] * kernel_shared[((((int)threadIdx.x) * 72) + 4647)]));
+        conv2d_nchw[10] = (conv2d_nchw[10] + (pad_temp_shared[120] * kernel_shared[((((int)threadIdx.x) * 72) + 4647)]));
+        conv2d_nchw[11] = (conv2d_nchw[11] + (pad_temp_shared[121] * kernel_shared[((((int)threadIdx.x) * 72) + 4647)]));
+        conv2d_nchw[12] = (conv2d_nchw[12] + (pad_temp_shared[122] * kernel_shared[((((int)threadIdx.x) * 72) + 4647)]));
+        conv2d_nchw[13] = (conv2d_nchw[13] + (pad_temp_shared[123] * kernel_shared[((((int)threadIdx.x) * 72) + 4647)]));
+        conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[118] * kernel_shared[((((int)threadIdx.x) * 72) + 40)]));
+        conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[119] * kernel_shared[((((int)threadIdx.x) * 72) + 40)]));
+        conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[120] * kernel_shared[((((int)threadIdx.x) * 72) + 40)]));
+        conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[121] * kernel_shared[((((int)threadIdx.x) * 72) + 40)]));
+        conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[122] * kernel_shared[((((int)threadIdx.x) * 72) + 40)]));
+        conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[123] * kernel_shared[((((int)threadIdx.x) * 72) + 40)]));
+        conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[124] * kernel_shared[((((int)threadIdx.x) * 72) + 40)]));
+        conv2d_nchw[7] = (conv2d_nchw[7] + (pad_temp_shared[118] * kernel_shared[((((int)threadIdx.x) * 72) + 4648)]));
+        conv2d_nchw[8] = (conv2d_nchw[8] + (pad_temp_shared[119] * kernel_shared[((((int)threadIdx.x) * 72) + 4648)]));
+        conv2d_nchw[9] = (conv2d_nchw[9] + (pad_temp_shared[120] * kernel_shared[((((int)threadIdx.x) * 72) + 4648)]));
+        conv2d_nchw[10] = (conv2d_nchw[10] + (pad_temp_shared[121] * kernel_shared[((((int)threadIdx.x) * 72) + 4648)]));
+        conv2d_nchw[11] = (conv2d_nchw[11] + (pad_temp_shared[122] * kernel_shared[((((int)threadIdx.x) * 72) + 4648)]));
+        conv2d_nchw[12] = (conv2d_nchw[12] + (pad_temp_shared[123] * kernel_shared[((((int)threadIdx.x) * 72) + 4648)]));
+        conv2d_nchw[13] = (conv2d_nchw[13] + (pad_temp_shared[124] * kernel_shared[((((int)threadIdx.x) * 72) + 4648)]));
+        conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[119] * kernel_shared[((((int)threadIdx.x) * 72) + 41)]));
+        conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[120] * kernel_shared[((((int)threadIdx.x) * 72) + 41)]));
+        conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[121] * kernel_shared[((((int)threadIdx.x) * 72) + 41)]));
+        conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[122] * kernel_shared[((((int)threadIdx.x) * 72) + 41)]));
+        conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[123] * kernel_shared[((((int)threadIdx.x) * 72) + 41)]));
+        conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[124] * kernel_shared[((((int)threadIdx.x) * 72) + 41)]));
+        conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[125] * kernel_shared[((((int)threadIdx.x) * 72) + 41)]));
+        conv2d_nchw[7] = (conv2d_nchw[7] + (pad_temp_shared[119] * kernel_shared[((((int)threadIdx.x) * 72) + 4649)]));
+        conv2d_nchw[8] = (conv2d_nchw[8] + (pad_temp_shared[120] * kernel_shared[((((int)threadIdx.x) * 72) + 4649)]));
+        conv2d_nchw[9] = (conv2d_nchw[9] + (pad_temp_shared[121] * kernel_shared[((((int)threadIdx.x) * 72) + 4649)]));
+        conv2d_nchw[10] = (conv2d_nchw[10] + (pad_temp_shared[122] * kernel_shared[((((int)threadIdx.x) * 72) + 4649)]));
+        conv2d_nchw[11] = (conv2d_nchw[11] + (pad_temp_shared[123] * kernel_shared[((((int)threadIdx.x) * 72) + 4649)]));
+        conv2d_nchw[12] = (conv2d_nchw[12] + (pad_temp_shared[124] * kernel_shared[((((int)threadIdx.x) * 72) + 4649)]));
+        conv2d_nchw[13] = (conv2d_nchw[13] + (pad_temp_shared[125] * kernel_shared[((((int)threadIdx.x) * 72) + 4649)]));
+        conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[126] * kernel_shared[((((int)threadIdx.x) * 72) + 42)]));
+        conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[127] * kernel_shared[((((int)threadIdx.x) * 72) + 42)]));
+        conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[128] * kernel_shared[((((int)threadIdx.x) * 72) + 42)]));
+        conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[129] * kernel_shared[((((int)threadIdx.x) * 72) + 42)]));
+        conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[130] * kernel_shared[((((int)threadIdx.x) * 72) + 42)]));
+        conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[131] * kernel_shared[((((int)threadIdx.x) * 72) + 42)]));
+        conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[132] * kernel_shared[((((int)threadIdx.x) * 72) + 42)]));
+        conv2d_nchw[7] = (conv2d_nchw[7] + (pad_temp_shared[126] * kernel_shared[((((int)threadIdx.x) * 72) + 4650)]));
+        conv2d_nchw[8] = (conv2d_nchw[8] + (pad_temp_shared[127] * kernel_shared[((((int)threadIdx.x) * 72) + 4650)]));
+        conv2d_nchw[9] = (conv2d_nchw[9] + (pad_temp_shared[128] * kernel_shared[((((int)threadIdx.x) * 72) + 4650)]));
+        conv2d_nchw[10] = (conv2d_nchw[10] + (pad_temp_shared[129] * kernel_shared[((((int)threadIdx.x) * 72) + 4650)]));
+        conv2d_nchw[11] = (conv2d_nchw[11] + (pad_temp_shared[130] * kernel_shared[((((int)threadIdx.x) * 72) + 4650)]));
+        conv2d_nchw[12] = (conv2d_nchw[12] + (pad_temp_shared[131] * kernel_shared[((((int)threadIdx.x) * 72) + 4650)]));
+        conv2d_nchw[13] = (conv2d_nchw[13] + (pad_temp_shared[132] * kernel_shared[((((int)threadIdx.x) * 72) + 4650)]));
+        conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[127] * kernel_shared[((((int)threadIdx.x) * 72) + 43)]));
+        conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[128] * kernel_shared[((((int)threadIdx.x) * 72) + 43)]));
+        conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[129] * kernel_shared[((((int)threadIdx.x) * 72) + 43)]));
+        conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[130] * kernel_shared[((((int)threadIdx.x) * 72) + 43)]));
+        conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[131] * kernel_shared[((((int)threadIdx.x) * 72) + 43)]));
+        conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[132] * kernel_shared[((((int)threadIdx.x) * 72) + 43)]));
+        conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[133] * kernel_shared[((((int)threadIdx.x) * 72) + 43)]));
+        conv2d_nchw[7] = (conv2d_nchw[7] + (pad_temp_shared[127] * kernel_shared[((((int)threadIdx.x) * 72) + 4651)]));
+        conv2d_nchw[8] = (conv2d_nchw[8] + (pad_temp_shared[128] * kernel_shared[((((int)threadIdx.x) * 72) + 4651)]));
+        conv2d_nchw[9] = (conv2d_nchw[9] + (pad_temp_shared[129] * kernel_shared[((((int)threadIdx.x) * 72) + 4651)]));
+        conv2d_nchw[10] = (conv2d_nchw[10] + (pad_temp_shared[130] * kernel_shared[((((int)threadIdx.x) * 72) + 4651)]));
+        conv2d_nchw[11] = (conv2d_nchw[11] + (pad_temp_shared[131] * kernel_shared[((((int)threadIdx.x) * 72) + 4651)]));
+        conv2d_nchw[12] = (conv2d_nchw[12] + (pad_temp_shared[132] * kernel_shared[((((int)threadIdx.x) * 72) + 4651)]));
+        conv2d_nchw[13] = (conv2d_nchw[13] + (pad_temp_shared[133] * kernel_shared[((((int)threadIdx.x) * 72) + 4651)]));
+        conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[128] * kernel_shared[((((int)threadIdx.x) * 72) + 44)]));
+        conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[129] * kernel_shared[((((int)threadIdx.x) * 72) + 44)]));
+        conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[130] * kernel_shared[((((int)threadIdx.x) * 72) + 44)]));
+        conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[131] * kernel_shared[((((int)threadIdx.x) * 72) + 44)]));
+        conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[132] * kernel_shared[((((int)threadIdx.x) * 72) + 44)]));
+        conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[133] * kernel_shared[((((int)threadIdx.x) * 72) + 44)]));
+        conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[134] * kernel_shared[((((int)threadIdx.x) * 72) + 44)]));
+        conv2d_nchw[7] = (conv2d_nchw[7] + (pad_temp_shared[128] * kernel_shared[((((int)threadIdx.x) * 72) + 4652)]));
+        conv2d_nchw[8] = (conv2d_nchw[8] + (pad_temp_shared[129] * kernel_shared[((((int)threadIdx.x) * 72) + 4652)]));
+        conv2d_nchw[9] = (conv2d_nchw[9] + (pad_temp_shared[130] * kernel_shared[((((int)threadIdx.x) * 72) + 4652)]));
+        conv2d_nchw[10] = (conv2d_nchw[10] + (pad_temp_shared[131] * kernel_shared[((((int)threadIdx.x) * 72) + 4652)]));
+        conv2d_nchw[11] = (conv2d_nchw[11] + (pad_temp_shared[132] * kernel_shared[((((int)threadIdx.x) * 72) + 4652)]));
+        conv2d_nchw[12] = (conv2d_nchw[12] + (pad_temp_shared[133] * kernel_shared[((((int)threadIdx.x) * 72) + 4652)]));
+        conv2d_nchw[13] = (conv2d_nchw[13] + (pad_temp_shared[134] * kernel_shared[((((int)threadIdx.x) * 72) + 4652)]));
+        conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[135] * kernel_shared[((((int)threadIdx.x) * 72) + 45)]));
+        conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[136] * kernel_shared[((((int)threadIdx.x) * 72) + 45)]));
+        conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[137] * kernel_shared[((((int)threadIdx.x) * 72) + 45)]));
+        conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[138] * kernel_shared[((((int)threadIdx.x) * 72) + 45)]));
+        conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[139] * kernel_shared[((((int)threadIdx.x) * 72) + 45)]));
+        conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[140] * kernel_shared[((((int)threadIdx.x) * 72) + 45)]));
+        conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[141] * kernel_shared[((((int)threadIdx.x) * 72) + 45)]));
+        conv2d_nchw[7] = (conv2d_nchw[7] + (pad_temp_shared[135] * kernel_shared[((((int)threadIdx.x) * 72) + 4653)]));
+        conv2d_nchw[8] = (conv2d_nchw[8] + (pad_temp_shared[136] * kernel_shared[((((int)threadIdx.x) * 72) + 4653)]));
+        conv2d_nchw[9] = (conv2d_nchw[9] + (pad_temp_shared[137] * kernel_shared[((((int)threadIdx.x) * 72) + 4653)]));
+        conv2d_nchw[10] = (conv2d_nchw[10] + (pad_temp_shared[138] * kernel_shared[((((int)threadIdx.x) * 72) + 4653)]));
+        conv2d_nchw[11] = (conv2d_nchw[11] + (pad_temp_shared[139] * kernel_shared[((((int)threadIdx.x) * 72) + 4653)]));
+        conv2d_nchw[12] = (conv2d_nchw[12] + (pad_temp_shared[140] * kernel_shared[((((int)threadIdx.x) * 72) + 4653)]));
+        conv2d_nchw[13] = (conv2d_nchw[13] + (pad_temp_shared[141] * kernel_shared[((((int)threadIdx.x) * 72) + 4653)]));
+        conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[136] * kernel_shared[((((int)threadIdx.x) * 72) + 46)]));
+        conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[137] * kernel_shared[((((int)threadIdx.x) * 72) + 46)]));
+        conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[138] * kernel_shared[((((int)threadIdx.x) * 72) + 46)]));
+        conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[139] * kernel_shared[((((int)threadIdx.x) * 72) + 46)]));
+        conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[140] * kernel_shared[((((int)threadIdx.x) * 72) + 46)]));
+        conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[141] * kernel_shared[((((int)threadIdx.x) * 72) + 46)]));
+        conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[142] * kernel_shared[((((int)threadIdx.x) * 72) + 46)]));
+        conv2d_nchw[7] = (conv2d_nchw[7] + (pad_temp_shared[136] * kernel_shared[((((int)threadIdx.x) * 72) + 4654)]));
+        conv2d_nchw[8] = (conv2d_nchw[8] + (pad_temp_shared[137] * kernel_shared[((((int)threadIdx.x) * 72) + 4654)]));
+        conv2d_nchw[9] = (conv2d_nchw[9] + (pad_temp_shared[138] * kernel_shared[((((int)threadIdx.x) * 72) + 4654)]));
+        conv2d_nchw[10] = (conv2d_nchw[10] + (pad_temp_shared[139] * kernel_shared[((((int)threadIdx.x) * 72) + 4654)]));
+        conv2d_nchw[11] = (conv2d_nchw[11] + (pad_temp_shared[140] * kernel_shared[((((int)threadIdx.x) * 72) + 4654)]));
+        conv2d_nchw[12] = (conv2d_nchw[12] + (pad_temp_shared[141] * kernel_shared[((((int)threadIdx.x) * 72) + 4654)]));
+        conv2d_nchw[13] = (conv2d_nchw[13] + (pad_temp_shared[142] * kernel_shared[((((int)threadIdx.x) * 72) + 4654)]));
+        conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[137] * kernel_shared[((((int)threadIdx.x) * 72) + 47)]));
+        conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[138] * kernel_shared[((((int)threadIdx.x) * 72) + 47)]));
+        conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[139] * kernel_shared[((((int)threadIdx.x) * 72) + 47)]));
+        conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[140] * kernel_shared[((((int)threadIdx.x) * 72) + 47)]));
+        conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[141] * kernel_shared[((((int)threadIdx.x) * 72) + 47)]));
+        conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[142] * kernel_shared[((((int)threadIdx.x) * 72) + 47)]));
+        conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[143] * kernel_shared[((((int)threadIdx.x) * 72) + 47)]));
+        conv2d_nchw[7] = (conv2d_nchw[7] + (pad_temp_shared[137] * kernel_shared[((((int)threadIdx.x) * 72) + 4655)]));
+        conv2d_nchw[8] = (conv2d_nchw[8] + (pad_temp_shared[138] * kernel_shared[((((int)threadIdx.x) * 72) + 4655)]));
+        conv2d_nchw[9] = (conv2d_nchw[9] + (pad_temp_shared[139] * kernel_shared[((((int)threadIdx.x) * 72) + 4655)]));
+        conv2d_nchw[10] = (conv2d_nchw[10] + (pad_temp_shared[140] * kernel_shared[((((int)threadIdx.x) * 72) + 4655)]));
+        conv2d_nchw[11] = (conv2d_nchw[11] + (pad_temp_shared[141] * kernel_shared[((((int)threadIdx.x) * 72) + 4655)]));
+        conv2d_nchw[12] = (conv2d_nchw[12] + (pad_temp_shared[142] * kernel_shared[((((int)threadIdx.x) * 72) + 4655)]));
+        conv2d_nchw[13] = (conv2d_nchw[13] + (pad_temp_shared[143] * kernel_shared[((((int)threadIdx.x) * 72) + 4655)]));
+        conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[144] * kernel_shared[((((int)threadIdx.x) * 72) + 48)]));
+        conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[145] * kernel_shared[((((int)threadIdx.x) * 72) + 48)]));
+        conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[146] * kernel_shared[((((int)threadIdx.x) * 72) + 48)]));
+        conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[147] * kernel_shared[((((int)threadIdx.x) * 72) + 48)]));
+        conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[148] * kernel_shared[((((int)threadIdx.x) * 72) + 48)]));
+        conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[149] * kernel_shared[((((int)threadIdx.x) * 72) + 48)]));
+        conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[150] * kernel_shared[((((int)threadIdx.x) * 72) + 48)]));
+        conv2d_nchw[7] = (conv2d_nchw[7] + (pad_temp_shared[144] * kernel_shared[((((int)threadIdx.x) * 72) + 4656)]));
+        conv2d_nchw[8] = (conv2d_nchw[8] + (pad_temp_shared[145] * kernel_shared[((((int)threadIdx.x) * 72) + 4656)]));
+        conv2d_nchw[9] = (conv2d_nchw[9] + (pad_temp_shared[146] * kernel_shared[((((int)threadIdx.x) * 72) + 4656)]));
+        conv2d_nchw[10] = (conv2d_nchw[10] + (pad_temp_shared[147] * kernel_shared[((((int)threadIdx.x) * 72) + 4656)]));
+        conv2d_nchw[11] = (conv2d_nchw[11] + (pad_temp_shared[148] * kernel_shared[((((int)threadIdx.x) * 72) + 4656)]));
+        conv2d_nchw[12] = (conv2d_nchw[12] + (pad_temp_shared[149] * kernel_shared[((((int)threadIdx.x) * 72) + 4656)]));
+        conv2d_nchw[13] = (conv2d_nchw[13] + (pad_temp_shared[150] * kernel_shared[((((int)threadIdx.x) * 72) + 4656)]));
+        conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[145] * kernel_shared[((((int)threadIdx.x) * 72) + 49)]));
+        conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[146] * kernel_shared[((((int)threadIdx.x) * 72) + 49)]));
+        conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[147] * kernel_shared[((((int)threadIdx.x) * 72) + 49)]));
+        conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[148] * kernel_shared[((((int)threadIdx.x) * 72) + 49)]));
+        conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[149] * kernel_shared[((((int)threadIdx.x) * 72) + 49)]));
+        conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[150] * kernel_shared[((((int)threadIdx.x) * 72) + 49)]));
+        conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[151] * kernel_shared[((((int)threadIdx.x) * 72) + 49)]));
+        conv2d_nchw[7] = (conv2d_nchw[7] + (pad_temp_shared[145] * kernel_shared[((((int)threadIdx.x) * 72) + 4657)]));
+        conv2d_nchw[8] = (conv2d_nchw[8] + (pad_temp_shared[146] * kernel_shared[((((int)threadIdx.x) * 72) + 4657)]));
+        conv2d_nchw[9] = (conv2d_nchw[9] + (pad_temp_shared[147] * kernel_shared[((((int)threadIdx.x) * 72) + 4657)]));
+        conv2d_nchw[10] = (conv2d_nchw[10] + (pad_temp_shared[148] * kernel_shared[((((int)threadIdx.x) * 72) + 4657)]));
+        conv2d_nchw[11] = (conv2d_nchw[11] + (pad_temp_shared[149] * kernel_shared[((((int)threadIdx.x) * 72) + 4657)]));
+        conv2d_nchw[12] = (conv2d_nchw[12] + (pad_temp_shared[150] * kernel_shared[((((int)threadIdx.x) * 72) + 4657)]));
+        conv2d_nchw[13] = (conv2d_nchw[13] + (pad_temp_shared[151] * kernel_shared[((((int)threadIdx.x) * 72) + 4657)]));
+        conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[146] * kernel_shared[((((int)threadIdx.x) * 72) + 50)]));
+        conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[147] * kernel_shared[((((int)threadIdx.x) * 72) + 50)]));
+        conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[148] * kernel_shared[((((int)threadIdx.x) * 72) + 50)]));
+        conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[149] * kernel_shared[((((int)threadIdx.x) * 72) + 50)]));
+        conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[150] * kernel_shared[((((int)threadIdx.x) * 72) + 50)]));
+        conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[151] * kernel_shared[((((int)threadIdx.x) * 72) + 50)]));
+        conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[152] * kernel_shared[((((int)threadIdx.x) * 72) + 50)]));
+        conv2d_nchw[7] = (conv2d_nchw[7] + (pad_temp_shared[146] * kernel_shared[((((int)threadIdx.x) * 72) + 4658)]));
+        conv2d_nchw[8] = (conv2d_nchw[8] + (pad_temp_shared[147] * kernel_shared[((((int)threadIdx.x) * 72) + 4658)]));
+        conv2d_nchw[9] = (conv2d_nchw[9] + (pad_temp_shared[148] * kernel_shared[((((int)threadIdx.x) * 72) + 4658)]));
+        conv2d_nchw[10] = (conv2d_nchw[10] + (pad_temp_shared[149] * kernel_shared[((((int)threadIdx.x) * 72) + 4658)]));
+        conv2d_nchw[11] = (conv2d_nchw[11] + (pad_temp_shared[150] * kernel_shared[((((int)threadIdx.x) * 72) + 4658)]));
+        conv2d_nchw[12] = (conv2d_nchw[12] + (pad_temp_shared[151] * kernel_shared[((((int)threadIdx.x) * 72) + 4658)]));
+        conv2d_nchw[13] = (conv2d_nchw[13] + (pad_temp_shared[152] * kernel_shared[((((int)threadIdx.x) * 72) + 4658)]));
+        conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[153] * kernel_shared[((((int)threadIdx.x) * 72) + 51)]));
+        conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[154] * kernel_shared[((((int)threadIdx.x) * 72) + 51)]));
+        conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[155] * kernel_shared[((((int)threadIdx.x) * 72) + 51)]));
+        conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[156] * kernel_shared[((((int)threadIdx.x) * 72) + 51)]));
+        conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[157] * kernel_shared[((((int)threadIdx.x) * 72) + 51)]));
+        conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[158] * kernel_shared[((((int)threadIdx.x) * 72) + 51)]));
+        conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[159] * kernel_shared[((((int)threadIdx.x) * 72) + 51)]));
+        conv2d_nchw[7] = (conv2d_nchw[7] + (pad_temp_shared[153] * kernel_shared[((((int)threadIdx.x) * 72) + 4659)]));
+        conv2d_nchw[8] = (conv2d_nchw[8] + (pad_temp_shared[154] * kernel_shared[((((int)threadIdx.x) * 72) + 4659)]));
+        conv2d_nchw[9] = (conv2d_nchw[9] + (pad_temp_shared[155] * kernel_shared[((((int)threadIdx.x) * 72) + 4659)]));
+        conv2d_nchw[10] = (conv2d_nchw[10] + (pad_temp_shared[156] * kernel_shared[((((int)threadIdx.x) * 72) + 4659)]));
+        conv2d_nchw[11] = (conv2d_nchw[11] + (pad_temp_shared[157] * kernel_shared[((((int)threadIdx.x) * 72) + 4659)]));
+        conv2d_nchw[12] = (conv2d_nchw[12] + (pad_temp_shared[158] * kernel_shared[((((int)threadIdx.x) * 72) + 4659)]));
+        conv2d_nchw[13] = (conv2d_nchw[13] + (pad_temp_shared[159] * kernel_shared[((((int)threadIdx.x) * 72) + 4659)]));
+        conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[154] * kernel_shared[((((int)threadIdx.x) * 72) + 52)]));
+        conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[155] * kernel_shared[((((int)threadIdx.x) * 72) + 52)]));
+        conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[156] * kernel_shared[((((int)threadIdx.x) * 72) + 52)]));
+        conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[157] * kernel_shared[((((int)threadIdx.x) * 72) + 52)]));
+        conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[158] * kernel_shared[((((int)threadIdx.x) * 72) + 52)]));
+        conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[159] * kernel_shared[((((int)threadIdx.x) * 72) + 52)]));
+        conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[160] * kernel_shared[((((int)threadIdx.x) * 72) + 52)]));
+        conv2d_nchw[7] = (conv2d_nchw[7] + (pad_temp_shared[154] * kernel_shared[((((int)threadIdx.x) * 72) + 4660)]));
+        conv2d_nchw[8] = (conv2d_nchw[8] + (pad_temp_shared[155] * kernel_shared[((((int)threadIdx.x) * 72) + 4660)]));
+        conv2d_nchw[9] = (conv2d_nchw[9] + (pad_temp_shared[156] * kernel_shared[((((int)threadIdx.x) * 72) + 4660)]));
+        conv2d_nchw[10] = (conv2d_nchw[10] + (pad_temp_shared[157] * kernel_shared[((((int)threadIdx.x) * 72) + 4660)]));
+        conv2d_nchw[11] = (conv2d_nchw[11] + (pad_temp_shared[158] * kernel_shared[((((int)threadIdx.x) * 72) + 4660)]));
+        conv2d_nchw[12] = (conv2d_nchw[12] + (pad_temp_shared[159] * kernel_shared[((((int)threadIdx.x) * 72) + 4660)]));
+        conv2d_nchw[13] = (conv2d_nchw[13] + (pad_temp_shared[160] * kernel_shared[((((int)threadIdx.x) * 72) + 4660)]));
+        conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[155] * kernel_shared[((((int)threadIdx.x) * 72) + 53)]));
+        conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[156] * kernel_shared[((((int)threadIdx.x) * 72) + 53)]));
+        conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[157] * kernel_shared[((((int)threadIdx.x) * 72) + 53)]));
+        conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[158] * kernel_shared[((((int)threadIdx.x) * 72) + 53)]));
+        conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[159] * kernel_shared[((((int)threadIdx.x) * 72) + 53)]));
+        conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[160] * kernel_shared[((((int)threadIdx.x) * 72) + 53)]));
+        conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[161] * kernel_shared[((((int)threadIdx.x) * 72) + 53)]));
+        conv2d_nchw[7] = (conv2d_nchw[7] + (pad_temp_shared[155] * kernel_shared[((((int)threadIdx.x) * 72) + 4661)]));
+        conv2d_nchw[8] = (conv2d_nchw[8] + (pad_temp_shared[156] * kernel_shared[((((int)threadIdx.x) * 72) + 4661)]));
+        conv2d_nchw[9] = (conv2d_nchw[9] + (pad_temp_shared[157] * kernel_shared[((((int)threadIdx.x) * 72) + 4661)]));
+        conv2d_nchw[10] = (conv2d_nchw[10] + (pad_temp_shared[158] * kernel_shared[((((int)threadIdx.x) * 72) + 4661)]));
+        conv2d_nchw[11] = (conv2d_nchw[11] + (pad_temp_shared[159] * kernel_shared[((((int)threadIdx.x) * 72) + 4661)]));
+        conv2d_nchw[12] = (conv2d_nchw[12] + (pad_temp_shared[160] * kernel_shared[((((int)threadIdx.x) * 72) + 4661)]));
+        conv2d_nchw[13] = (conv2d_nchw[13] + (pad_temp_shared[161] * kernel_shared[((((int)threadIdx.x) * 72) + 4661)]));
+        conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[162] * kernel_shared[((((int)threadIdx.x) * 72) + 54)]));
+        conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[163] * kernel_shared[((((int)threadIdx.x) * 72) + 54)]));
+        conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[164] * kernel_shared[((((int)threadIdx.x) * 72) + 54)]));
+        conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[165] * kernel_shared[((((int)threadIdx.x) * 72) + 54)]));
+        conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[166] * kernel_shared[((((int)threadIdx.x) * 72) + 54)]));
+        conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[167] * kernel_shared[((((int)threadIdx.x) * 72) + 54)]));
+        conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[168] * kernel_shared[((((int)threadIdx.x) * 72) + 54)]));
+        conv2d_nchw[7] = (conv2d_nchw[7] + (pad_temp_shared[162] * kernel_shared[((((int)threadIdx.x) * 72) + 4662)]));
+        conv2d_nchw[8] = (conv2d_nchw[8] + (pad_temp_shared[163] * kernel_shared[((((int)threadIdx.x) * 72) + 4662)]));
+        conv2d_nchw[9] = (conv2d_nchw[9] + (pad_temp_shared[164] * kernel_shared[((((int)threadIdx.x) * 72) + 4662)]));
+        conv2d_nchw[10] = (conv2d_nchw[10] + (pad_temp_shared[165] * kernel_shared[((((int)threadIdx.x) * 72) + 4662)]));
+        conv2d_nchw[11] = (conv2d_nchw[11] + (pad_temp_shared[166] * kernel_shared[((((int)threadIdx.x) * 72) + 4662)]));
+        conv2d_nchw[12] = (conv2d_nchw[12] + (pad_temp_shared[167] * kernel_shared[((((int)threadIdx.x) * 72) + 4662)]));
+        conv2d_nchw[13] = (conv2d_nchw[13] + (pad_temp_shared[168] * kernel_shared[((((int)threadIdx.x) * 72) + 4662)]));
+        conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[163] * kernel_shared[((((int)threadIdx.x) * 72) + 55)]));
+        conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[164] * kernel_shared[((((int)threadIdx.x) * 72) + 55)]));
+        conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[165] * kernel_shared[((((int)threadIdx.x) * 72) + 55)]));
+        conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[166] * kernel_shared[((((int)threadIdx.x) * 72) + 55)]));
+        conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[167] * kernel_shared[((((int)threadIdx.x) * 72) + 55)]));
+        conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[168] * kernel_shared[((((int)threadIdx.x) * 72) + 55)]));
+        conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[169] * kernel_shared[((((int)threadIdx.x) * 72) + 55)]));
+        conv2d_nchw[7] = (conv2d_nchw[7] + (pad_temp_shared[163] * kernel_shared[((((int)threadIdx.x) * 72) + 4663)]));
+        conv2d_nchw[8] = (conv2d_nchw[8] + (pad_temp_shared[164] * kernel_shared[((((int)threadIdx.x) * 72) + 4663)]));
+        conv2d_nchw[9] = (conv2d_nchw[9] + (pad_temp_shared[165] * kernel_shared[((((int)threadIdx.x) * 72) + 4663)]));
+        conv2d_nchw[10] = (conv2d_nchw[10] + (pad_temp_shared[166] * kernel_shared[((((int)threadIdx.x) * 72) + 4663)]));
+        conv2d_nchw[11] = (conv2d_nchw[11] + (pad_temp_shared[167] * kernel_shared[((((int)threadIdx.x) * 72) + 4663)]));
+        conv2d_nchw[12] = (conv2d_nchw[12] + (pad_temp_shared[168] * kernel_shared[((((int)threadIdx.x) * 72) + 4663)]));
+        conv2d_nchw[13] = (conv2d_nchw[13] + (pad_temp_shared[169] * kernel_shared[((((int)threadIdx.x) * 72) + 4663)]));
+        conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[164] * kernel_shared[((((int)threadIdx.x) * 72) + 56)]));
+        conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[165] * kernel_shared[((((int)threadIdx.x) * 72) + 56)]));
+        conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[166] * kernel_shared[((((int)threadIdx.x) * 72) + 56)]));
+        conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[167] * kernel_shared[((((int)threadIdx.x) * 72) + 56)]));
+        conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[168] * kernel_shared[((((int)threadIdx.x) * 72) + 56)]));
+        conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[169] * kernel_shared[((((int)threadIdx.x) * 72) + 56)]));
+        conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[170] * kernel_shared[((((int)threadIdx.x) * 72) + 56)]));
+        conv2d_nchw[7] = (conv2d_nchw[7] + (pad_temp_shared[164] * kernel_shared[((((int)threadIdx.x) * 72) + 4664)]));
+        conv2d_nchw[8] = (conv2d_nchw[8] + (pad_temp_shared[165] * kernel_shared[((((int)threadIdx.x) * 72) + 4664)]));
+        conv2d_nchw[9] = (conv2d_nchw[9] + (pad_temp_shared[166] * kernel_shared[((((int)threadIdx.x) * 72) + 4664)]));
+        conv2d_nchw[10] = (conv2d_nchw[10] + (pad_temp_shared[167] * kernel_shared[((((int)threadIdx.x) * 72) + 4664)]));
+        conv2d_nchw[11] = (conv2d_nchw[11] + (pad_temp_shared[168] * kernel_shared[((((int)threadIdx.x) * 72) + 4664)]));
+        conv2d_nchw[12] = (conv2d_nchw[12] + (pad_temp_shared[169] * kernel_shared[((((int)threadIdx.x) * 72) + 4664)]));
+        conv2d_nchw[13] = (conv2d_nchw[13] + (pad_temp_shared[170] * kernel_shared[((((int)threadIdx.x) * 72) + 4664)]));
+        conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[171] * kernel_shared[((((int)threadIdx.x) * 72) + 57)]));
+        conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[172] * kernel_shared[((((int)threadIdx.x) * 72) + 57)]));
+        conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[173] * kernel_shared[((((int)threadIdx.x) * 72) + 57)]));
+        conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[174] * kernel_shared[((((int)threadIdx.x) * 72) + 57)]));
+        conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[175] * kernel_shared[((((int)threadIdx.x) * 72) + 57)]));
+        conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[176] * kernel_shared[((((int)threadIdx.x) * 72) + 57)]));
+        conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[177] * kernel_shared[((((int)threadIdx.x) * 72) + 57)]));
+        conv2d_nchw[7] = (conv2d_nchw[7] + (pad_temp_shared[171] * kernel_shared[((((int)threadIdx.x) * 72) + 4665)]));
+        conv2d_nchw[8] = (conv2d_nchw[8] + (pad_temp_shared[172] * kernel_shared[((((int)threadIdx.x) * 72) + 4665)]));
+        conv2d_nchw[9] = (conv2d_nchw[9] + (pad_temp_shared[173] * kernel_shared[((((int)threadIdx.x) * 72) + 4665)]));
+        conv2d_nchw[10] = (conv2d_nchw[10] + (pad_temp_shared[174] * kernel_shared[((((int)threadIdx.x) * 72) + 4665)]));
+        conv2d_nchw[11] = (conv2d_nchw[11] + (pad_temp_shared[175] * kernel_shared[((((int)threadIdx.x) * 72) + 4665)]));
+        conv2d_nchw[12] = (conv2d_nchw[12] + (pad_temp_shared[176] * kernel_shared[((((int)threadIdx.x) * 72) + 4665)]));
+        conv2d_nchw[13] = (conv2d_nchw[13] + (pad_temp_shared[177] * kernel_shared[((((int)threadIdx.x) * 72) + 4665)]));
+        conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[172] * kernel_shared[((((int)threadIdx.x) * 72) + 58)]));
+        conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[173] * kernel_shared[((((int)threadIdx.x) * 72) + 58)]));
+        conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[174] * kernel_shared[((((int)threadIdx.x) * 72) + 58)]));
+        conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[175] * kernel_shared[((((int)threadIdx.x) * 72) + 58)]));
+        conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[176] * kernel_shared[((((int)threadIdx.x) * 72) + 58)]));
+        conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[177] * kernel_shared[((((int)threadIdx.x) * 72) + 58)]));
+        conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[178] * kernel_shared[((((int)threadIdx.x) * 72) + 58)]));
+        conv2d_nchw[7] = (conv2d_nchw[7] + (pad_temp_shared[172] * kernel_shared[((((int)threadIdx.x) * 72) + 4666)]));
+        conv2d_nchw[8] = (conv2d_nchw[8] + (pad_temp_shared[173] * kernel_shared[((((int)threadIdx.x) * 72) + 4666)]));
+        conv2d_nchw[9] = (conv2d_nchw[9] + (pad_temp_shared[174] * kernel_shared[((((int)threadIdx.x) * 72) + 4666)]));
+        conv2d_nchw[10] = (conv2d_nchw[10] + (pad_temp_shared[175] * kernel_shared[((((int)threadIdx.x) * 72) + 4666)]));
+        conv2d_nchw[11] = (conv2d_nchw[11] + (pad_temp_shared[176] * kernel_shared[((((int)threadIdx.x) * 72) + 4666)]));
+        conv2d_nchw[12] = (conv2d_nchw[12] + (pad_temp_shared[177] * kernel_shared[((((int)threadIdx.x) * 72) + 4666)]));
+        conv2d_nchw[13] = (conv2d_nchw[13] + (pad_temp_shared[178] * kernel_shared[((((int)threadIdx.x) * 72) + 4666)]));
+        conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[173] * kernel_shared[((((int)threadIdx.x) * 72) + 59)]));
+        conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[174] * kernel_shared[((((int)threadIdx.x) * 72) + 59)]));
+        conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[175] * kernel_shared[((((int)threadIdx.x) * 72) + 59)]));
+        conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[176] * kernel_shared[((((int)threadIdx.x) * 72) + 59)]));
+        conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[177] * kernel_shared[((((int)threadIdx.x) * 72) + 59)]));
+        conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[178] * kernel_shared[((((int)threadIdx.x) * 72) + 59)]));
+        conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[179] * kernel_shared[((((int)threadIdx.x) * 72) + 59)]));
+        conv2d_nchw[7] = (conv2d_nchw[7] + (pad_temp_shared[173] * kernel_shared[((((int)threadIdx.x) * 72) + 4667)]));
+        conv2d_nchw[8] = (conv2d_nchw[8] + (pad_temp_shared[174] * kernel_shared[((((int)threadIdx.x) * 72) + 4667)]));
+        conv2d_nchw[9] = (conv2d_nchw[9] + (pad_temp_shared[175] * kernel_shared[((((int)threadIdx.x) * 72) + 4667)]));
+        conv2d_nchw[10] = (conv2d_nchw[10] + (pad_temp_shared[176] * kernel_shared[((((int)threadIdx.x) * 72) + 4667)]));
+        conv2d_nchw[11] = (conv2d_nchw[11] + (pad_temp_shared[177] * kernel_shared[((((int)threadIdx.x) * 72) + 4667)]));
+        conv2d_nchw[12] = (conv2d_nchw[12] + (pad_temp_shared[178] * kernel_shared[((((int)threadIdx.x) * 72) + 4667)]));
+        conv2d_nchw[13] = (conv2d_nchw[13] + (pad_temp_shared[179] * kernel_shared[((((int)threadIdx.x) * 72) + 4667)]));
+        conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[180] * kernel_shared[((((int)threadIdx.x) * 72) + 60)]));
+        conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[181] * kernel_shared[((((int)threadIdx.x) * 72) + 60)]));
+        conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[182] * kernel_shared[((((int)threadIdx.x) * 72) + 60)]));
+        conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[183] * kernel_shared[((((int)threadIdx.x) * 72) + 60)]));
+        conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[184] * kernel_shared[((((int)threadIdx.x) * 72) + 60)]));
+        conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[185] * kernel_shared[((((int)threadIdx.x) * 72) + 60)]));
+        conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[186] * kernel_shared[((((int)threadIdx.x) * 72) + 60)]));
+        conv2d_nchw[7] = (conv2d_nchw[7] + (pad_temp_shared[180] * kernel_shared[((((int)threadIdx.x) * 72) + 4668)]));
+        conv2d_nchw[8] = (conv2d_nchw[8] + (pad_temp_shared[181] * kernel_shared[((((int)threadIdx.x) * 72) + 4668)]));
+        conv2d_nchw[9] = (conv2d_nchw[9] + (pad_temp_shared[182] * kernel_shared[((((int)threadIdx.x) * 72) + 4668)]));
+        conv2d_nchw[10] = (conv2d_nchw[10] + (pad_temp_shared[183] * kernel_shared[((((int)threadIdx.x) * 72) + 4668)]));
+        conv2d_nchw[11] = (conv2d_nchw[11] + (pad_temp_shared[184] * kernel_shared[((((int)threadIdx.x) * 72) + 4668)]));
+        conv2d_nchw[12] = (conv2d_nchw[12] + (pad_temp_shared[185] * kernel_shared[((((int)threadIdx.x) * 72) + 4668)]));
+        conv2d_nchw[13] = (conv2d_nchw[13] + (pad_temp_shared[186] * kernel_shared[((((int)threadIdx.x) * 72) + 4668)]));
+        conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[181] * kernel_shared[((((int)threadIdx.x) * 72) + 61)]));
+        conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[182] * kernel_shared[((((int)threadIdx.x) * 72) + 61)]));
+        conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[183] * kernel_shared[((((int)threadIdx.x) * 72) + 61)]));
+        conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[184] * kernel_shared[((((int)threadIdx.x) * 72) + 61)]));
+        conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[185] * kernel_shared[((((int)threadIdx.x) * 72) + 61)]));
+        conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[186] * kernel_shared[((((int)threadIdx.x) * 72) + 61)]));
+        conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[187] * kernel_shared[((((int)threadIdx.x) * 72) + 61)]));
+        conv2d_nchw[7] = (conv2d_nchw[7] + (pad_temp_shared[181] * kernel_shared[((((int)threadIdx.x) * 72) + 4669)]));
+        conv2d_nchw[8] = (conv2d_nchw[8] + (pad_temp_shared[182] * kernel_shared[((((int)threadIdx.x) * 72) + 4669)]));
+        conv2d_nchw[9] = (conv2d_nchw[9] + (pad_temp_shared[183] * kernel_shared[((((int)threadIdx.x) * 72) + 4669)]));
+        conv2d_nchw[10] = (conv2d_nchw[10] + (pad_temp_shared[184] * kernel_shared[((((int)threadIdx.x) * 72) + 4669)]));
+        conv2d_nchw[11] = (conv2d_nchw[11] + (pad_temp_shared[185] * kernel_shared[((((int)threadIdx.x) * 72) + 4669)]));
+        conv2d_nchw[12] = (conv2d_nchw[12] + (pad_temp_shared[186] * kernel_shared[((((int)threadIdx.x) * 72) + 4669)]));
+        conv2d_nchw[13] = (conv2d_nchw[13] + (pad_temp_shared[187] * kernel_shared[((((int)threadIdx.x) * 72) + 4669)]));
+        conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[182] * kernel_shared[((((int)threadIdx.x) * 72) + 62)]));
+        conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[183] * kernel_shared[((((int)threadIdx.x) * 72) + 62)]));
+        conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[184] * kernel_shared[((((int)threadIdx.x) * 72) + 62)]));
+        conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[185] * kernel_shared[((((int)threadIdx.x) * 72) + 62)]));
+        conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[186] * kernel_shared[((((int)threadIdx.x) * 72) + 62)]));
+        conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[187] * kernel_shared[((((int)threadIdx.x) * 72) + 62)]));
+        conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[188] * kernel_shared[((((int)threadIdx.x) * 72) + 62)]));
+        conv2d_nchw[7] = (conv2d_nchw[7] + (pad_temp_shared[182] * kernel_shared[((((int)threadIdx.x) * 72) + 4670)]));
+        conv2d_nchw[8] = (conv2d_nchw[8] + (pad_temp_shared[183] * kernel_shared[((((int)threadIdx.x) * 72) + 4670)]));
+        conv2d_nchw[9] = (conv2d_nchw[9] + (pad_temp_shared[184] * kernel_shared[((((int)threadIdx.x) * 72) + 4670)]));
+        conv2d_nchw[10] = (conv2d_nchw[10] + (pad_temp_shared[185] * kernel_shared[((((int)threadIdx.x) * 72) + 4670)]));
+        conv2d_nchw[11] = (conv2d_nchw[11] + (pad_temp_shared[186] * kernel_shared[((((int)threadIdx.x) * 72) + 4670)]));
+        conv2d_nchw[12] = (conv2d_nchw[12] + (pad_temp_shared[187] * kernel_shared[((((int)threadIdx.x) * 72) + 4670)]));
+        conv2d_nchw[13] = (conv2d_nchw[13] + (pad_temp_shared[188] * kernel_shared[((((int)threadIdx.x) * 72) + 4670)]));
+        conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[189] * kernel_shared[((((int)threadIdx.x) * 72) + 63)]));
+        conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[190] * kernel_shared[((((int)threadIdx.x) * 72) + 63)]));
+        conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[191] * kernel_shared[((((int)threadIdx.x) * 72) + 63)]));
+        conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[192] * kernel_shared[((((int)threadIdx.x) * 72) + 63)]));
+        conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[193] * kernel_shared[((((int)threadIdx.x) * 72) + 63)]));
+        conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[194] * kernel_shared[((((int)threadIdx.x) * 72) + 63)]));
+        conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[195] * kernel_shared[((((int)threadIdx.x) * 72) + 63)]));
+        conv2d_nchw[7] = (conv2d_nchw[7] + (pad_temp_shared[189] * kernel_shared[((((int)threadIdx.x) * 72) + 4671)]));
+        conv2d_nchw[8] = (conv2d_nchw[8] + (pad_temp_shared[190] * kernel_shared[((((int)threadIdx.x) * 72) + 4671)]));
+        conv2d_nchw[9] = (conv2d_nchw[9] + (pad_temp_shared[191] * kernel_shared[((((int)threadIdx.x) * 72) + 4671)]));
+        conv2d_nchw[10] = (conv2d_nchw[10] + (pad_temp_shared[192] * kernel_shared[((((int)threadIdx.x) * 72) + 4671)]));
+        conv2d_nchw[11] = (conv2d_nchw[11] + (pad_temp_shared[193] * kernel_shared[((((int)threadIdx.x) * 72) + 4671)]));
+        conv2d_nchw[12] = (conv2d_nchw[12] + (pad_temp_shared[194] * kernel_shared[((((int)threadIdx.x) * 72) + 4671)]));
+        conv2d_nchw[13] = (conv2d_nchw[13] + (pad_temp_shared[195] * kernel_shared[((((int)threadIdx.x) * 72) + 4671)]));
+        conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[190] * kernel_shared[((((int)threadIdx.x) * 72) + 64)]));
+        conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[191] * kernel_shared[((((int)threadIdx.x) * 72) + 64)]));
+        conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[192] * kernel_shared[((((int)threadIdx.x) * 72) + 64)]));
+        conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[193] * kernel_shared[((((int)threadIdx.x) * 72) + 64)]));
+        conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[194] * kernel_shared[((((int)threadIdx.x) * 72) + 64)]));
+        conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[195] * kernel_shared[((((int)threadIdx.x) * 72) + 64)]));
+        conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[196] * kernel_shared[((((int)threadIdx.x) * 72) + 64)]));
+        conv2d_nchw[7] = (conv2d_nchw[7] + (pad_temp_shared[190] * kernel_shared[((((int)threadIdx.x) * 72) + 4672)]));
+        conv2d_nchw[8] = (conv2d_nchw[8] + (pad_temp_shared[191] * kernel_shared[((((int)threadIdx.x) * 72) + 4672)]));
+        conv2d_nchw[9] = (conv2d_nchw[9] + (pad_temp_shared[192] * kernel_shared[((((int)threadIdx.x) * 72) + 4672)]));
+        conv2d_nchw[10] = (conv2d_nchw[10] + (pad_temp_shared[193] * kernel_shared[((((int)threadIdx.x) * 72) + 4672)]));
+        conv2d_nchw[11] = (conv2d_nchw[11] + (pad_temp_shared[194] * kernel_shared[((((int)threadIdx.x) * 72) + 4672)]));
+        conv2d_nchw[12] = (conv2d_nchw[12] + (pad_temp_shared[195] * kernel_shared[((((int)threadIdx.x) * 72) + 4672)]));
+        conv2d_nchw[13] = (conv2d_nchw[13] + (pad_temp_shared[196] * kernel_shared[((((int)threadIdx.x) * 72) + 4672)]));
+        conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[191] * kernel_shared[((((int)threadIdx.x) * 72) + 65)]));
+        conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[192] * kernel_shared[((((int)threadIdx.x) * 72) + 65)]));
+        conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[193] * kernel_shared[((((int)threadIdx.x) * 72) + 65)]));
+        conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[194] * kernel_shared[((((int)threadIdx.x) * 72) + 65)]));
+        conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[195] * kernel_shared[((((int)threadIdx.x) * 72) + 65)]));
+        conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[196] * kernel_shared[((((int)threadIdx.x) * 72) + 65)]));
+        conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[197] * kernel_shared[((((int)threadIdx.x) * 72) + 65)]));
+        conv2d_nchw[7] = (conv2d_nchw[7] + (pad_temp_shared[191] * kernel_shared[((((int)threadIdx.x) * 72) + 4673)]));
+        conv2d_nchw[8] = (conv2d_nchw[8] + (pad_temp_shared[192] * kernel_shared[((((int)threadIdx.x) * 72) + 4673)]));
+        conv2d_nchw[9] = (conv2d_nchw[9] + (pad_temp_shared[193] * kernel_shared[((((int)threadIdx.x) * 72) + 4673)]));
+        conv2d_nchw[10] = (conv2d_nchw[10] + (pad_temp_shared[194] * kernel_shared[((((int)threadIdx.x) * 72) + 4673)]));
+        conv2d_nchw[11] = (conv2d_nchw[11] + (pad_temp_shared[195] * kernel_shared[((((int)threadIdx.x) * 72) + 4673)]));
+        conv2d_nchw[12] = (conv2d_nchw[12] + (pad_temp_shared[196] * kernel_shared[((((int)threadIdx.x) * 72) + 4673)]));
+        conv2d_nchw[13] = (conv2d_nchw[13] + (pad_temp_shared[197] * kernel_shared[((((int)threadIdx.x) * 72) + 4673)]));
+        conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[198] * kernel_shared[((((int)threadIdx.x) * 72) + 66)]));
+        conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[199] * kernel_shared[((((int)threadIdx.x) * 72) + 66)]));
+        conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[200] * kernel_shared[((((int)threadIdx.x) * 72) + 66)]));
+        conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[201] * kernel_shared[((((int)threadIdx.x) * 72) + 66)]));
+        conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[202] * kernel_shared[((((int)threadIdx.x) * 72) + 66)]));
+        conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[203] * kernel_shared[((((int)threadIdx.x) * 72) + 66)]));
+        conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[204] * kernel_shared[((((int)threadIdx.x) * 72) + 66)]));
+        conv2d_nchw[7] = (conv2d_nchw[7] + (pad_temp_shared[198] * kernel_shared[((((int)threadIdx.x) * 72) + 4674)]));
+        conv2d_nchw[8] = (conv2d_nchw[8] + (pad_temp_shared[199] * kernel_shared[((((int)threadIdx.x) * 72) + 4674)]));
+        conv2d_nchw[9] = (conv2d_nchw[9] + (pad_temp_shared[200] * kernel_shared[((((int)threadIdx.x) * 72) + 4674)]));
+        conv2d_nchw[10] = (conv2d_nchw[10] + (pad_temp_shared[201] * kernel_shared[((((int)threadIdx.x) * 72) + 4674)]));
+        conv2d_nchw[11] = (conv2d_nchw[11] + (pad_temp_shared[202] * kernel_shared[((((int)threadIdx.x) * 72) + 4674)]));
+        conv2d_nchw[12] = (conv2d_nchw[12] + (pad_temp_shared[203] * kernel_shared[((((int)threadIdx.x) * 72) + 4674)]));
+        conv2d_nchw[13] = (conv2d_nchw[13] + (pad_temp_shared[204] * kernel_shared[((((int)threadIdx.x) * 72) + 4674)]));
+        conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[199] * kernel_shared[((((int)threadIdx.x) * 72) + 67)]));
+        conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[200] * kernel_shared[((((int)threadIdx.x) * 72) + 67)]));
+        conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[201] * kernel_shared[((((int)threadIdx.x) * 72) + 67)]));
+        conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[202] * kernel_shared[((((int)threadIdx.x) * 72) + 67)]));
+        conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[203] * kernel_shared[((((int)threadIdx.x) * 72) + 67)]));
+        conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[204] * kernel_shared[((((int)threadIdx.x) * 72) + 67)]));
+        conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[205] * kernel_shared[((((int)threadIdx.x) * 72) + 67)]));
+        conv2d_nchw[7] = (conv2d_nchw[7] + (pad_temp_shared[199] * kernel_shared[((((int)threadIdx.x) * 72) + 4675)]));
+        conv2d_nchw[8] = (conv2d_nchw[8] + (pad_temp_shared[200] * kernel_shared[((((int)threadIdx.x) * 72) + 4675)]));
+        conv2d_nchw[9] = (conv2d_nchw[9] + (pad_temp_shared[201] * kernel_shared[((((int)threadIdx.x) * 72) + 4675)]));
+        conv2d_nchw[10] = (conv2d_nchw[10] + (pad_temp_shared[202] * kernel_shared[((((int)threadIdx.x) * 72) + 4675)]));
+        conv2d_nchw[11] = (conv2d_nchw[11] + (pad_temp_shared[203] * kernel_shared[((((int)threadIdx.x) * 72) + 4675)]));
+        conv2d_nchw[12] = (conv2d_nchw[12] + (pad_temp_shared[204] * kernel_shared[((((int)threadIdx.x) * 72) + 4675)]));
+        conv2d_nchw[13] = (conv2d_nchw[13] + (pad_temp_shared[205] * kernel_shared[((((int)threadIdx.x) * 72) + 4675)]));
+        conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[200] * kernel_shared[((((int)threadIdx.x) * 72) + 68)]));
+        conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[201] * kernel_shared[((((int)threadIdx.x) * 72) + 68)]));
+        conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[202] * kernel_shared[((((int)threadIdx.x) * 72) + 68)]));
+        conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[203] * kernel_shared[((((int)threadIdx.x) * 72) + 68)]));
+        conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[204] * kernel_shared[((((int)threadIdx.x) * 72) + 68)]));
+        conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[205] * kernel_shared[((((int)threadIdx.x) * 72) + 68)]));
+        conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[206] * kernel_shared[((((int)threadIdx.x) * 72) + 68)]));
+        conv2d_nchw[7] = (conv2d_nchw[7] + (pad_temp_shared[200] * kernel_shared[((((int)threadIdx.x) * 72) + 4676)]));
+        conv2d_nchw[8] = (conv2d_nchw[8] + (pad_temp_shared[201] * kernel_shared[((((int)threadIdx.x) * 72) + 4676)]));
+        conv2d_nchw[9] = (conv2d_nchw[9] + (pad_temp_shared[202] * kernel_shared[((((int)threadIdx.x) * 72) + 4676)]));
+        conv2d_nchw[10] = (conv2d_nchw[10] + (pad_temp_shared[203] * kernel_shared[((((int)threadIdx.x) * 72) + 4676)]));
+        conv2d_nchw[11] = (conv2d_nchw[11] + (pad_temp_shared[204] * kernel_shared[((((int)threadIdx.x) * 72) + 4676)]));
+        conv2d_nchw[12] = (conv2d_nchw[12] + (pad_temp_shared[205] * kernel_shared[((((int)threadIdx.x) * 72) + 4676)]));
+        conv2d_nchw[13] = (conv2d_nchw[13] + (pad_temp_shared[206] * kernel_shared[((((int)threadIdx.x) * 72) + 4676)]));
+        conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[207] * kernel_shared[((((int)threadIdx.x) * 72) + 69)]));
+        conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[208] * kernel_shared[((((int)threadIdx.x) * 72) + 69)]));
+        conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[209] * kernel_shared[((((int)threadIdx.x) * 72) + 69)]));
+        conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[210] * kernel_shared[((((int)threadIdx.x) * 72) + 69)]));
+        conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[211] * kernel_shared[((((int)threadIdx.x) * 72) + 69)]));
+        conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[212] * kernel_shared[((((int)threadIdx.x) * 72) + 69)]));
+        conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[213] * kernel_shared[((((int)threadIdx.x) * 72) + 69)]));
+        conv2d_nchw[7] = (conv2d_nchw[7] + (pad_temp_shared[207] * kernel_shared[((((int)threadIdx.x) * 72) + 4677)]));
+        conv2d_nchw[8] = (conv2d_nchw[8] + (pad_temp_shared[208] * kernel_shared[((((int)threadIdx.x) * 72) + 4677)]));
+        conv2d_nchw[9] = (conv2d_nchw[9] + (pad_temp_shared[209] * kernel_shared[((((int)threadIdx.x) * 72) + 4677)]));
+        conv2d_nchw[10] = (conv2d_nchw[10] + (pad_temp_shared[210] * kernel_shared[((((int)threadIdx.x) * 72) + 4677)]));
+        conv2d_nchw[11] = (conv2d_nchw[11] + (pad_temp_shared[211] * kernel_shared[((((int)threadIdx.x) * 72) + 4677)]));
+        conv2d_nchw[12] = (conv2d_nchw[12] + (pad_temp_shared[212] * kernel_shared[((((int)threadIdx.x) * 72) + 4677)]));
+        conv2d_nchw[13] = (conv2d_nchw[13] + (pad_temp_shared[213] * kernel_shared[((((int)threadIdx.x) * 72) + 4677)]));
+        conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[208] * kernel_shared[((((int)threadIdx.x) * 72) + 70)]));
+        conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[209] * kernel_shared[((((int)threadIdx.x) * 72) + 70)]));
+        conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[210] * kernel_shared[((((int)threadIdx.x) * 72) + 70)]));
+        conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[211] * kernel_shared[((((int)threadIdx.x) * 72) + 70)]));
+        conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[212] * kernel_shared[((((int)threadIdx.x) * 72) + 70)]));
+        conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[213] * kernel_shared[((((int)threadIdx.x) * 72) + 70)]));
+        conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[214] * kernel_shared[((((int)threadIdx.x) * 72) + 70)]));
+        conv2d_nchw[7] = (conv2d_nchw[7] + (pad_temp_shared[208] * kernel_shared[((((int)threadIdx.x) * 72) + 4678)]));
+        conv2d_nchw[8] = (conv2d_nchw[8] + (pad_temp_shared[209] * kernel_shared[((((int)threadIdx.x) * 72) + 4678)]));
+        conv2d_nchw[9] = (conv2d_nchw[9] + (pad_temp_shared[210] * kernel_shared[((((int)threadIdx.x) * 72) + 4678)]));
+        conv2d_nchw[10] = (conv2d_nchw[10] + (pad_temp_shared[211] * kernel_shared[((((int)threadIdx.x) * 72) + 4678)]));
+        conv2d_nchw[11] = (conv2d_nchw[11] + (pad_temp_shared[212] * kernel_shared[((((int)threadIdx.x) * 72) + 4678)]));
+        conv2d_nchw[12] = (conv2d_nchw[12] + (pad_temp_shared[213] * kernel_shared[((((int)threadIdx.x) * 72) + 4678)]));
+        conv2d_nchw[13] = (conv2d_nchw[13] + (pad_temp_shared[214] * kernel_shared[((((int)threadIdx.x) * 72) + 4678)]));
+        conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[209] * kernel_shared[((((int)threadIdx.x) * 72) + 71)]));
+        conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[210] * kernel_shared[((((int)threadIdx.x) * 72) + 71)]));
+        conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[211] * kernel_shared[((((int)threadIdx.x) * 72) + 71)]));
+        conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[212] * kernel_shared[((((int)threadIdx.x) * 72) + 71)]));
+        conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[213] * kernel_shared[((((int)threadIdx.x) * 72) + 71)]));
+        conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[214] * kernel_shared[((((int)threadIdx.x) * 72) + 71)]));
+        conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[215] * kernel_shared[((((int)threadIdx.x) * 72) + 71)]));
+        conv2d_nchw[7] = (conv2d_nchw[7] + (pad_temp_shared[209] * kernel_shared[((((int)threadIdx.x) * 72) + 4679)]));
+        conv2d_nchw[8] = (conv2d_nchw[8] + (pad_temp_shared[210] * kernel_shared[((((int)threadIdx.x) * 72) + 4679)]));
+        conv2d_nchw[9] = (conv2d_nchw[9] + (pad_temp_shared[211] * kernel_shared[((((int)threadIdx.x) * 72) + 4679)]));
+        conv2d_nchw[10] = (conv2d_nchw[10] + (pad_temp_shared[212] * kernel_shared[((((int)threadIdx.x) * 72) + 4679)]));
+        conv2d_nchw[11] = (conv2d_nchw[11] + (pad_temp_shared[213] * kernel_shared[((((int)threadIdx.x) * 72) + 4679)]));
+        conv2d_nchw[12] = (conv2d_nchw[12] + (pad_temp_shared[214] * kernel_shared[((((int)threadIdx.x) * 72) + 4679)]));
+        conv2d_nchw[13] = (conv2d_nchw[13] + (pad_temp_shared[215] * kernel_shared[((((int)threadIdx.x) * 72) + 4679)]));
       }
+      compute[((((((int)blockIdx.x) / 7) * 6272) + (((int)threadIdx.x) * 49)) + ((((int)blockIdx.x) % 7) * 7))] = max((conv2d_nchw[0] + bias[(((((int)blockIdx.x) / 7) * 128) + ((int)threadIdx.x))]), 0.000000e+00f);
+      compute[(((((((int)blockIdx.x) / 7) * 6272) + (((int)threadIdx.x) * 49)) + ((((int)blockIdx.x) % 7) * 7)) + 1)] = max((conv2d_nchw[1] + bias[(((((int)blockIdx.x) / 7) * 128) + ((int)threadIdx.x))]), 0.000000e+00f);
+      compute[(((((((int)blockIdx.x) / 7) * 6272) + (((int)threadIdx.x) * 49)) + ((((int)blockIdx.x) % 7) * 7)) + 2)] = max((conv2d_nchw[2] + bias[(((((int)blockIdx.x) / 7) * 128) + ((int)threadIdx.x))]), 0.000000e+00f);
+      compute[(((((((int)blockIdx.x) / 7) * 6272) + (((int)threadIdx.x) * 49)) + ((((int)blockIdx.x) % 7) * 7)) + 3)] = max((conv2d_nchw[3] + bias[(((((int)blockIdx.x) / 7) * 128) + ((int)threadIdx.x))]), 0.000000e+00f);
+      compute[(((((((int)blockIdx.x) / 7) * 6272) + (((int)threadIdx.x) * 49)) + ((((int)blockIdx.x) % 7) * 7)) + 4)] = max((conv2d_nchw[4] + bias[(((((int)blockIdx.x) / 7) * 128) + ((int)threadIdx.x))]), 0.000000e+00f);
+      compute[(((((((int)blockIdx.x) / 7) * 6272) + (((int)threadIdx.x) * 49)) + ((((int)blockIdx.x) % 7) * 7)) + 5)] = max((conv2d_nchw[5] + bias[(((((int)blockIdx.x) / 7) * 128) + ((int)threadIdx.x))]), 0.000000e+00f);
+      compute[(((((((int)blockIdx.x) / 7) * 6272) + (((int)threadIdx.x) * 49)) + ((((int)blockIdx.x) % 7) * 7)) + 6)] = max((conv2d_nchw[6] + bias[(((((int)blockIdx.x) / 7) * 128) + ((int)threadIdx.x))]), 0.000000e+00f);
+      compute[(((((((int)blockIdx.x) / 7) * 6272) + (((int)threadIdx.x) * 49)) + ((((int)blockIdx.x) % 7) * 7)) + 3136)] = max((conv2d_nchw[7] + bias[((((((int)blockIdx.x) / 7) * 128) + ((int)threadIdx.x)) + 64)]), 0.000000e+00f);
+      compute[(((((((int)blockIdx.x) / 7) * 6272) + (((int)threadIdx.x) * 49)) + ((((int)blockIdx.x) % 7) * 7)) + 3137)] = max((conv2d_nchw[8] + bias[((((((int)blockIdx.x) / 7) * 128) + ((int)threadIdx.x)) + 64)]), 0.000000e+00f);
+      compute[(((((((int)blockIdx.x) / 7) * 6272) + (((int)threadIdx.x) * 49)) + ((((int)blockIdx.x) % 7) * 7)) + 3138)] = max((conv2d_nchw[9] + bias[((((((int)blockIdx.x) / 7) * 128) + ((int)threadIdx.x)) + 64)]), 0.000000e+00f);
+      compute[(((((((int)blockIdx.x) / 7) * 6272) + (((int)threadIdx.x) * 49)) + ((((int)blockIdx.x) % 7) * 7)) + 3139)] = max((conv2d_nchw[10] + bias[((((((int)blockIdx.x) / 7) * 128) + ((int)threadIdx.x)) + 64)]), 0.000000e+00f);
+      compute[(((((((int)blockIdx.x) / 7) * 6272) + (((int)threadIdx.x) * 49)) + ((((int)blockIdx.x) % 7) * 7)) + 3140)] = max((conv2d_nchw[11] + bias[((((((int)blockIdx.x) / 7) * 128) + ((int)threadIdx.x)) + 64)]), 0.000000e+00f);
+      compute[(((((((int)blockIdx.x) / 7) * 6272) + (((int)threadIdx.x) * 49)) + ((((int)blockIdx.x) % 7) * 7)) + 3141)] = max((conv2d_nchw[12] + bias[((((((int)blockIdx.x) / 7) * 128) + ((int)threadIdx.x)) + 64)]), 0.000000e+00f);
+      compute[(((((((int)blockIdx.x) / 7) * 6272) + (((int)threadIdx.x) * 49)) + ((((int)blockIdx.x) % 7) * 7)) + 3142)] = max((conv2d_nchw[13] + bias[((((((int)blockIdx.x) / 7) * 128) + ((int)threadIdx.x)) + 64)]), 0.000000e+00f);
     }
 
 
@@ -1388,7 +3017,7 @@ In the example below we resume the status and do more 5 trials.
 
 .. rst-class:: sphx-glr-timing
 
-   **Total running time of the script:** ( 5 minutes  43.471 seconds)
+   **Total running time of the script:** ( 5 minutes  37.856 seconds)
 
 
 .. _sphx_glr_download_how_to_tune_with_autoscheduler_tune_conv2d_layer_cuda.py:
diff --git a/docs/_sources/how_to/tune_with_autoscheduler/tune_network_cuda.rst.txt b/docs/_sources/how_to/tune_with_autoscheduler/tune_network_cuda.rst.txt
index cd26f4d9aa..3ad5cab773 100644
--- a/docs/_sources/how_to/tune_with_autoscheduler/tune_network_cuda.rst.txt
+++ b/docs/_sources/how_to/tune_with_autoscheduler/tune_network_cuda.rst.txt
@@ -647,7 +647,7 @@ so we can read the log file and load the best schedules.
     Evaluate inference time cost...
     Execution time summary:
      mean (ms)   median (ms)    max (ms)     min (ms)     std (ms)  
-       7.9234       7.9221       7.9270       7.9211       0.0025   
+       7.9140       7.9132       7.9191       7.9097       0.0039   
                
 
 
@@ -675,7 +675,7 @@ Other Tips
 
 .. rst-class:: sphx-glr-timing
 
-   **Total running time of the script:** ( 1 minutes  2.866 seconds)
+   **Total running time of the script:** ( 1 minutes  1.645 seconds)
 
 
 .. _sphx_glr_download_how_to_tune_with_autoscheduler_tune_network_cuda.py:
diff --git a/docs/_sources/how_to/tune_with_autoscheduler/tune_network_x86.rst.txt b/docs/_sources/how_to/tune_with_autoscheduler/tune_network_x86.rst.txt
index a1dc9e84d2..7a2d922a85 100644
--- a/docs/_sources/how_to/tune_with_autoscheduler/tune_network_x86.rst.txt
+++ b/docs/_sources/how_to/tune_with_autoscheduler/tune_network_x86.rst.txt
@@ -666,7 +666,7 @@ so we can read the log file and load the best schedules.
     Evaluate inference time cost...
     Execution time summary:
      mean (ms)   median (ms)    max (ms)     min (ms)     std (ms)  
-      770.1799     771.8335     772.9309     765.7752      3.1466   
+      747.4894     747.7091     747.9152     746.8437      0.4642   
                
 
 
@@ -694,7 +694,7 @@ Other Tips
 
 .. rst-class:: sphx-glr-timing
 
-   **Total running time of the script:** ( 1 minutes  34.285 seconds)
+   **Total running time of the script:** ( 1 minutes  31.696 seconds)
 
 
 .. _sphx_glr_download_how_to_tune_with_autoscheduler_tune_network_x86.py:
diff --git a/docs/_sources/how_to/tune_with_autoscheduler/tune_sparse_x86.rst.txt b/docs/_sources/how_to/tune_with_autoscheduler/tune_sparse_x86.rst.txt
index 40639e8b07..ed66796131 100644
--- a/docs/_sources/how_to/tune_with_autoscheduler/tune_sparse_x86.rst.txt
+++ b/docs/_sources/how_to/tune_with_autoscheduler/tune_sparse_x86.rst.txt
@@ -390,77 +390,27 @@ layout transformation, parallelization, vectorization, unrolling, and operator f
                  placeholder_4: Buffer(placeholder_14: Pointer(float32), float32, [128, 512], []),
                  compute: Buffer(compute_2: Pointer(float32), float32, [128, 512], [])}
       buffer_map = {placeholder_5: placeholder, placeholder_6: placeholder_1, placeholder_7: placeholder_2, placeholder_8: placeholder_3, placeholder_9: placeholder_4, compute_1: compute} {
-      for (i0.outer.i1.outer.fused: int32, 0, 32) "parallel" {
-        allocate(compute_3: Pointer(global float32), float32, [2048]), storage_scope = global {
-          for (i.outer.inner: int32, 0, 8) {
-            for (nb_j.inner: int32, 0, 2) {
-              for (i.inner.init: int32, 0, 8) {
-                let cse_var_1: int32 = (((i.outer.inner*256) + (i.inner.init*32)) + (nb_j.inner*16))
-                 {
-                  compute_4: Buffer(compute_3, float32, [2048], [])[cse_var_1] = 0f32
-                  compute_4[(cse_var_1 + 1)] = 0f32
-                  compute_4[(cse_var_1 + 2)] = 0f32
-                  compute_4[(cse_var_1 + 3)] = 0f32
-                  compute_4[(cse_var_1 + 4)] = 0f32
-                  compute_4[(cse_var_1 + 5)] = 0f32
-                  compute_4[(cse_var_1 + 6)] = 0f32
-                  compute_4[(cse_var_1 + 7)] = 0f32
-                  compute_4[(cse_var_1 + 8)] = 0f32
-                  compute_4[(cse_var_1 + 9)] = 0f32
-                  compute_4[(cse_var_1 + 10)] = 0f32
-                  compute_4[(cse_var_1 + 11)] = 0f32
-                  compute_4[(cse_var_1 + 12)] = 0f32
-                  compute_4[(cse_var_1 + 13)] = 0f32
-                  compute_4[(cse_var_1 + 14)] = 0f32
-                  compute_4[(cse_var_1 + 15)] = 0f32
-                }
-              }
-              for (elem_idx: int32, 0, let cse_var_2: int32 = ((floormod(i0.outer.i1.outer.fused, 16)*2) + nb_j.inner) in (placeholder_15: Buffer(placeholder_13, int32, [33], [])[(cse_var_2 + 1)] - placeholder_15[cse_var_2])) {
-                for (i.inner: int32, 0, 8) {
-                  let cse_var_21: int32 = (elem_idx*16)
-                  let cse_var_20: int32 = ((floormod(i0.outer.i1.outer.fused, 16)*2) + nb_j.inner)
-                  let cse_var_19: int32 = (((i.outer.inner*256) + (i.inner*32)) + (nb_j.inner*16))
-                  let cse_var_18: int32 = (((floordiv(i0.outer.i1.outer.fused, 16)*16384) + (i.outer.inner*2048)) + (i.inner*256))
-                  let cse_var_17: int32 = (cse_var_19 + 9)
-                  let cse_var_16: int32 = (cse_var_19 + 8)
-                  let cse_var_15: int32 = (cse_var_19 + 7)
-                  let cse_var_14: int32 = (cse_var_19 + 6)
-                  let cse_var_13: int32 = (cse_var_19 + 5)
-                  let cse_var_12: int32 = (cse_var_19 + 4)
-                  let cse_var_11: int32 = (cse_var_19 + 3)
-                  let cse_var_10: int32 = (cse_var_19 + 2)
-                  let cse_var_9: int32 = (cse_var_19 + 15)
-                  let cse_var_8: int32 = (cse_var_19 + 14)
-                  let cse_var_7: int32 = (cse_var_19 + 13)
-                  let cse_var_6: int32 = (cse_var_19 + 12)
-                  let cse_var_5: int32 = (cse_var_19 + 11)
-                  let cse_var_4: int32 = (cse_var_19 + 10)
-                  let cse_var_3: int32 = (cse_var_19 + 1)
-                   {
-                    compute_4[cse_var_19] = (compute_4[cse_var_19] + (placeholder_16: Buffer(placeholder_11, float32, [78656], [])[((placeholder_15[cse_var_20]*16) + cse_var_21)]*max(placeholder_17: Buffer(placeholder_10, float32, [32768], [])[(cse_var_18 + placeholder_18: Buffer(placeholder_12, int32, [4916], [])[(placeholder_15[cse_var_20] + elem_idx)])], 0f32)))
-                    compute_4[cse_var_3] = (compute_4[cse_var_3] + (placeholder_16[(((placeholder_15[cse_var_20]*16) + cse_var_21) + 1)]*max(placeholder_17[(cse_var_18 + placeholder_18[(placeholder_15[cse_var_20] + elem_idx)])], 0f32)))
-                    compute_4[cse_var_10] = (compute_4[cse_var_10] + (placeholder_16[(((placeholder_15[cse_var_20]*16) + cse_var_21) + 2)]*max(placeholder_17[(cse_var_18 + placeholder_18[(placeholder_15[cse_var_20] + elem_idx)])], 0f32)))
-                    compute_4[cse_var_11] = (compute_4[cse_var_11] + (placeholder_16[(((placeholder_15[cse_var_20]*16) + cse_var_21) + 3)]*max(placeholder_17[(cse_var_18 + placeholder_18[(placeholder_15[cse_var_20] + elem_idx)])], 0f32)))
-                    compute_4[cse_var_12] = (compute_4[cse_var_12] + (placeholder_16[(((placeholder_15[cse_var_20]*16) + cse_var_21) + 4)]*max(placeholder_17[(cse_var_18 + placeholder_18[(placeholder_15[cse_var_20] + elem_idx)])], 0f32)))
-                    compute_4[cse_var_13] = (compute_4[cse_var_13] + (placeholder_16[(((placeholder_15[cse_var_20]*16) + cse_var_21) + 5)]*max(placeholder_17[(cse_var_18 + placeholder_18[(placeholder_15[cse_var_20] + elem_idx)])], 0f32)))
-                    compute_4[cse_var_14] = (compute_4[cse_var_14] + (placeholder_16[(((placeholder_15[cse_var_20]*16) + cse_var_21) + 6)]*max(placeholder_17[(cse_var_18 + placeholder_18[(placeholder_15[cse_var_20] + elem_idx)])], 0f32)))
-                    compute_4[cse_var_15] = (compute_4[cse_var_15] + (placeholder_16[(((placeholder_15[cse_var_20]*16) + cse_var_21) + 7)]*max(placeholder_17[(cse_var_18 + placeholder_18[(placeholder_15[cse_var_20] + elem_idx)])], 0f32)))
-                    compute_4[cse_var_16] = (compute_4[cse_var_16] + (placeholder_16[(((placeholder_15[cse_var_20]*16) + cse_var_21) + 8)]*max(placeholder_17[(cse_var_18 + placeholder_18[(placeholder_15[cse_var_20] + elem_idx)])], 0f32)))
-                    compute_4[cse_var_17] = (compute_4[cse_var_17] + (placeholder_16[(((placeholder_15[cse_var_20]*16) + cse_var_21) + 9)]*max(placeholder_17[(cse_var_18 + placeholder_18[(placeholder_15[cse_var_20] + elem_idx)])], 0f32)))
-                    compute_4[cse_var_4] = (compute_4[cse_var_4] + (placeholder_16[(((placeholder_15[cse_var_20]*16) + cse_var_21) + 10)]*max(placeholder_17[(cse_var_18 + placeholder_18[(placeholder_15[cse_var_20] + elem_idx)])], 0f32)))
-                    compute_4[cse_var_5] = (compute_4[cse_var_5] + (placeholder_16[(((placeholder_15[cse_var_20]*16) + cse_var_21) + 11)]*max(placeholder_17[(cse_var_18 + placeholder_18[(placeholder_15[cse_var_20] + elem_idx)])], 0f32)))
-                    compute_4[cse_var_6] = (compute_4[cse_var_6] + (placeholder_16[(((placeholder_15[cse_var_20]*16) + cse_var_21) + 12)]*max(placeholder_17[(cse_var_18 + placeholder_18[(placeholder_15[cse_var_20] + elem_idx)])], 0f32)))
-                    compute_4[cse_var_7] = (compute_4[cse_var_7] + (placeholder_16[(((placeholder_15[cse_var_20]*16) + cse_var_21) + 13)]*max(placeholder_17[(cse_var_18 + placeholder_18[(placeholder_15[cse_var_20] + elem_idx)])], 0f32)))
-                    compute_4[cse_var_8] = (compute_4[cse_var_8] + (placeholder_16[(((placeholder_15[cse_var_20]*16) + cse_var_21) + 14)]*max(placeholder_17[(cse_var_18 + placeholder_18[(placeholder_15[cse_var_20] + elem_idx)])], 0f32)))
-                    compute_4[cse_var_9] = (compute_4[cse_var_9] + (placeholder_16[(((placeholder_15[cse_var_20]*16) + cse_var_21) + 15)]*max(placeholder_17[(cse_var_18 + placeholder_18[(placeholder_15[cse_var_20] + elem_idx)])], 0f32)))
-                  }
+      for (i0.outer.i1.outer.fused: int32, 0, 512) "parallel" {
+        allocate(compute_3: Pointer(global float32), float32, [128]), storage_scope = global {
+          for (i.inner.init: int32, 0, 8) {
+            for (j.init: int32, 0, 16) {
+              compute_4: Buffer(compute_3, float32, [128], [])[((i.inner.init*16) + j.init)] = 0f32
+            }
+          }
+          for (elem_idx: int32, 0, let cse_var_1: int32 = floormod(i0.outer.i1.outer.fused, 32) in (placeholder_15: Buffer(placeholder_13, int32, [33], [])[(cse_var_1 + 1)] - placeholder_15[cse_var_1])) {
+            for (i.inner: int32, 0, 8) {
+              for (j: int32, 0, 16) {
+                let cse_var_2: int32 = floormod(i0.outer.i1.outer.fused, 32)
+                if @tir.likely((elem_idx < (placeholder_15[(cse_var_2 + 1)] - placeholder_15[cse_var_2])), dtype=bool) {
+                  let cse_var_3: int32 = ((i.inner*16) + j)
+                  compute_4[cse_var_3] = (compute_4[cse_var_3] + (placeholder_16: Buffer(placeholder_11, float32, [78656], [])[(((placeholder_15[cse_var_2]*16) + (elem_idx*16)) + j)]*max(placeholder_17: Buffer(placeholder_10, float32, [32768], [])[(((floordiv(i0.outer.i1.outer.fused, 32)*2048) + (i.inner*256)) + placeholder_18: Buffer(placeholder_12, int32, [4916], [])[(placeholder_15[cse_var_2] + elem_idx)])], 0f32)))
                 }
               }
             }
           }
-          for (i0.inner: int32, 0, 64) {
-            let cse_var_22: int32 = (((floordiv(i0.outer.i1.outer.fused, 16)*32768) + (i0.inner*512)) + (floormod(i0.outer.i1.outer.fused, 16)*32))
-            compute_5: Buffer(compute_2, float32, [65536], [])[ramp(cse_var_22, 1, 32)] = max((compute_4[ramp((i0.inner*32), 1, 32)] + placeholder_19: Buffer(placeholder_14, float32, [65536], [])[ramp(cse_var_22, 1, 32)]), broadcast(0f32, 32))
+          for (i0.inner: int32, 0, 8) {
+            let cse_var_4: int32 = (((floordiv(i0.outer.i1.outer.fused, 32)*4096) + (i0.inner*512)) + (floormod(i0.outer.i1.outer.fused, 32)*16))
+            compute_5: Buffer(compute_2, float32, [65536], [])[ramp(cse_var_4, 1, 16)] = max((compute_4[ramp((i0.inner*16), 1, 16)] + placeholder_19: Buffer(placeholder_14, float32, [65536], [])[ramp(cse_var_4, 1, 16)]), broadcast(0f32, 16))
           }
         }
       }
@@ -516,7 +466,7 @@ We build the binary and check its correctness and performance.
 
  .. code-block:: none
 
-    Execution time of this operator: 1.852 ms
+    Execution time of this operator: 1.535 ms
 
 
 
diff --git a/docs/_sources/how_to/tune_with_autotvm/sg_execution_times.rst.txt b/docs/_sources/how_to/tune_with_autotvm/sg_execution_times.rst.txt
index da233b3cef..4683a09f15 100644
--- a/docs/_sources/how_to/tune_with_autotvm/sg_execution_times.rst.txt
+++ b/docs/_sources/how_to/tune_with_autotvm/sg_execution_times.rst.txt
@@ -5,12 +5,12 @@
 
 Computation times
 =================
-**00:46.613** total execution time for **how_to_tune_with_autotvm** files:
+**00:39.677** total execution time for **how_to_tune_with_autotvm** files:
 
 +--------------------------------------------------------------------------------------------------+-----------+--------+
-| :ref:`sphx_glr_how_to_tune_with_autotvm_tune_conv2d_cuda.py` (``tune_conv2d_cuda.py``)           | 00:46.576 | 0.0 MB |
+| :ref:`sphx_glr_how_to_tune_with_autotvm_tune_conv2d_cuda.py` (``tune_conv2d_cuda.py``)           | 00:39.641 | 0.0 MB |
 +--------------------------------------------------------------------------------------------------+-----------+--------+
-| :ref:`sphx_glr_how_to_tune_with_autotvm_tune_relay_x86.py` (``tune_relay_x86.py``)               | 00:00.022 | 0.0 MB |
+| :ref:`sphx_glr_how_to_tune_with_autotvm_tune_relay_x86.py` (``tune_relay_x86.py``)               | 00:00.021 | 0.0 MB |
 +--------------------------------------------------------------------------------------------------+-----------+--------+
 | :ref:`sphx_glr_how_to_tune_with_autotvm_tune_relay_cuda.py` (``tune_relay_cuda.py``)             | 00:00.005 | 0.0 MB |
 +--------------------------------------------------------------------------------------------------+-----------+--------+
diff --git a/docs/_sources/how_to/tune_with_autotvm/tune_conv2d_cuda.rst.txt b/docs/_sources/how_to/tune_with_autotvm/tune_conv2d_cuda.rst.txt
index 7d7d2f6184..13d0bb666d 100644
--- a/docs/_sources/how_to/tune_with_autotvm/tune_conv2d_cuda.rst.txt
+++ b/docs/_sources/how_to/tune_with_autotvm/tune_conv2d_cuda.rst.txt
@@ -391,7 +391,7 @@ for this template
       File "tvm/_ffi/_cython/./packed_func.pxi", line 56, in tvm._ffi._cy3.core.tvm_callback
       File "/workspace/python/tvm/autotvm/measure/measure_methods.py", line 875, in verify_pass
         raise InstantiationError("Skipped because of invalid gpu kernel")
-    tvm.autotvm.task.space.InstantiationError: Skipped because of invalid gpu kernel        [('tile_f', [-1, 1, 1, 16]), ('tile_y', [-1, 1, 7, 1]), ('tile_x', [-1, 7, 1, 1]), ('tile_rc', [-1, 512, 1]), ('tile_ry', [-1, 1, 1]), ('tile_rx', [-1, 1, 1]), ('auto_unroll_max_step', 512), ('unroll_explicit', 0)],None,1775564
+    tvm.autotvm.task.space.InstantiationError: Skipped because of invalid gpu kernel        [('tile_f', [-1, 8, 8, 4]), ('tile_y', [-1, 1, 7, 1]), ('tile_x', [-1, 7, 1, 1]), ('tile_rc', [-1, 32, 4]), ('tile_ry', [-1, 1, 1]), ('tile_rx', [-1, 1, 3]), ('auto_unroll_max_step', 0), ('unroll_explicit', 0)],None,1247524
     No: 2   GFLOPS: 0.00/0.00       result: Traceback (most recent call last):
       File "/workspace/python/tvm/autotvm/measure/measure_methods.py", line 592, in __call__
         func, arg_info = _build_func_common(measure_input, self.runtime, **kwargs)
@@ -514,9 +514,8 @@ for this template
       File "tvm/_ffi/_cython/./packed_func.pxi", line 56, in tvm._ffi._cy3.core.tvm_callback
       File "/workspace/python/tvm/autotvm/measure/measure_methods.py", line 875, in verify_pass
         raise InstantiationError("Skipped because of invalid gpu kernel")
-    tvm.autotvm.task.space.InstantiationError: Skipped because of invalid gpu kernel        [('tile_f', [-1, 64, 2, 2]), ('tile_y', [-1, 1, 7, 1]), ('tile_x', [-1, 1, 1, 1]), ('tile_rc', [-1, 2, 16]), ('tile_ry', [-1, 1, 1]), ('tile_rx', [-1, 1, 3]), ('auto_unroll_max_step', 512), ('unroll_explicit', 1)],None,8254910
-    No: 3   GFLOPS: 57.82/57.82     result: MeasureResult(costs=(0.0040036236400000005,), error_no=MeasureErrorNo.NO_ERROR, all_cost=1.583892822265625, timestamp=1673392864.772967)        [('tile_f', [-1, 2, 2, 2]), ('tile_y', [-1, 1, 1, 1]), ('tile_x', [-1, 1, 7, 1]), ('tile_rc', [-1, 4, 8]), ('tile_ry', [-1, 1, 3]), ('tile_rx', [-1, 3, 1]), ('auto_unroll_max_step', 0), ('unroll_explicit', 0)],None,1071905
-    No: 4   GFLOPS: 0.00/57.82      result: Traceback (most recent call last):
+    tvm.autotvm.task.space.InstantiationError: Skipped because of invalid gpu kernel        [('tile_f', [-1, 1, 1, 32]), ('tile_y', [-1, 1, 1, 7]), ('tile_x', [-1, 1, 1, 1]), ('tile_rc', [-1, 64, 1]), ('tile_ry', [-1, 3, 1]), ('tile_rx', [-1, 3, 1]), ('auto_unroll_max_step', 0), ('unroll_explicit', 1)],None,6023565
+    No: 3   GFLOPS: 0.00/0.00       result: Traceback (most recent call last):
       File "/workspace/python/tvm/autotvm/measure/measure_methods.py", line 592, in __call__
         func, arg_info = _build_func_common(measure_input, self.runtime, **kwargs)
       File "/workspace/python/tvm/autotvm/measure/measure_methods.py", line 544, in _build_func_common
@@ -638,8 +637,8 @@ for this template
       File "tvm/_ffi/_cython/./packed_func.pxi", line 56, in tvm._ffi._cy3.core.tvm_callback
       File "/workspace/python/tvm/autotvm/measure/measure_methods.py", line 875, in verify_pass
         raise InstantiationError("Skipped because of invalid gpu kernel")
-    tvm.autotvm.task.space.InstantiationError: Skipped because of invalid gpu kernel        [('tile_f', [-1, 1, 128, 4]), ('tile_y', [-1, 7, 1, 1]), ('tile_x', [-1, 1, 1, 7]), ('tile_rc', [-1, 64, 2]), ('tile_ry', [-1, 3, 1]), ('tile_rx', [-1, 1, 1]), ('auto_unroll_max_step', 1500), ('unroll_explicit', 1)],None,8964915
-    No: 5   GFLOPS: 0.00/57.82      result: Traceback (most recent call last):
+    tvm.autotvm.task.space.InstantiationError: Skipped because of invalid gpu kernel        [('tile_f', [-1, 2, 4, 32]), ('tile_y', [-1, 1, 7, 1]), ('tile_x', [-1, 1, 7, 1]), ('tile_rc', [-1, 8, 1]), ('tile_ry', [-1, 1, 3]), ('tile_rx', [-1, 1, 3]), ('auto_unroll_max_step', 1500), ('unroll_explicit', 0)],None,5046555
+    No: 4   GFLOPS: 0.00/0.00       result: Traceback (most recent call last):
       File "/workspace/python/tvm/autotvm/measure/measure_methods.py", line 592, in __call__
         func, arg_info = _build_func_common(measure_input, self.runtime, **kwargs)
       File "/workspace/python/tvm/autotvm/measure/measure_methods.py", line 544, in _build_func_common
@@ -761,161 +760,9 @@ for this template
       File "tvm/_ffi/_cython/./packed_func.pxi", line 56, in tvm._ffi._cy3.core.tvm_callback
       File "/workspace/python/tvm/autotvm/measure/measure_methods.py", line 875, in verify_pass
         raise InstantiationError("Skipped because of invalid gpu kernel")
-    tvm.autotvm.task.space.InstantiationError: Skipped because of invalid gpu kernel        [('tile_f', [-1, 1, 1, 8]), ('tile_y', [-1, 7, 1, 1]), ('tile_x', [-1, 1, 1, 1]), ('tile_rc', [-1, 8, 32]), ('tile_ry', [-1, 1, 3]), ('tile_rx', [-1, 3, 1]), ('auto_unroll_max_step', 1500), ('unroll_explicit', 1)],None,9831716
-    No: 6   GFLOPS: 0.00/57.82      result: Traceback (most recent call last):
-      File "/workspace/python/tvm/autotvm/measure/measure_methods.py", line 742, in __call__
-        yield remote, remote.load_module(os.path.split(build_result.filename)[1])
-      File "/workspace/python/tvm/autotvm/measure/measure_methods.py", line 706, in run_through_rpc
-        costs = time_f(*args).results
-      File "/workspace/python/tvm/runtime/module.py", line 357, in evaluator
-        blob = feval(*args)
-      File "tvm/_ffi/_cython/./packed_func.pxi", line 331, in tvm._ffi._cy3.core.PackedFuncBase.__call__
-      File "tvm/_ffi/_cython/./packed_func.pxi", line 262, in tvm._ffi._cy3.core.FuncCall
-      File "tvm/_ffi/_cython/./packed_func.pxi", line 251, in tvm._ffi._cy3.core.FuncCall3
-      File "tvm/_ffi/_cython/./base.pxi", line 181, in tvm._ffi._cy3.core.CHECK_CALL
-    tvm._ffi.base.TVMError: Traceback (most recent call last):
-      4: TVMFuncCall
-            at ../src/runtime/c_runtime_api.cc:477
-      3: tvm::runtime::PackedFuncObj::CallPacked(tvm::runtime::TVMArgs, tvm::runtime::TVMRetValue*) const
-            at ../include/tvm/runtime/packed_func.h:1217
-      2: tvm::runtime::RPCWrappedFunc::operator()(tvm::runtime::TVMArgs, tvm::runtime::TVMRetValue*) const
-            at ../src/runtime/rpc/rpc_module.cc:129
-      1: tvm::runtime::RPCClientSession::CallFunc(void*, TVMValue const*, int const*, int, std::function<void (tvm::runtime::TVMArgs)> const&)
-            at ../src/runtime/rpc/rpc_endpoint.cc:1012
-      0: tvm::runtime::RPCEndpoint::CallFunc(void*, TVMValue const*, int const*, int, std::function<void (tvm::runtime::TVMArgs)>)
-            at ../src/runtime/rpc/rpc_endpoint.cc:804
-      File "../src/runtime/rpc/rpc_endpoint.cc", line 804
-    TVMError: 
-    ---------------------------------------------------------------
-    An error occurred during the execution of TVM.
-    For more information, please see: https://tvm.apache.org/docs/errors.html
-    ---------------------------------------------------------------
-      Check failed: (code == RPCCode::kReturn) is false: code=kShutdown
-
-    During handling of the above exception, another exception occurred:
-
-    Traceback (most recent call last):
-      File "/workspace/python/tvm/autotvm/measure/measure_methods.py", line 706, in run_through_rpc
-        costs = time_f(*args).results
-      File "/usr/lib/python3.7/contextlib.py", line 130, in __exit__
-        self.gen.throw(type, value, traceback)
-      File "/workspace/python/tvm/autotvm/measure/measure_methods.py", line 746, in __call__
-        remote.remove(build_result.filename)
-      File "/workspace/python/tvm/rpc/client.py", line 144, in remove
-        self._remote_funcs["remove"] = self.get_function("tvm.rpc.server.remove")
-      File "/workspace/python/tvm/rpc/client.py", line 72, in get_function
-        return self._sess.get_function(name)
-      File "/workspace/python/tvm/runtime/module.py", line 171, in get_function
-        self.handle, c_str(name), ctypes.c_int(query_imports), ctypes.byref(ret_handle)
-      File "/workspace/python/tvm/_ffi/base.py", line 348, in check_call
-        raise get_last_ffi_error()
-    tvm._ffi.base.TVMError: Traceback (most recent call last):
-      52: 0xffffffffffffffff
-      51: _start
-      50: __libc_start_main
-      49: _Py_UnixMain
-      48: 0x0000000000650da0
-      47: 0x0000000000650afa
-      46: _PyFunction_FastCallDict
-      45: _PyEval_EvalCodeWithName
-      44: _PyEval_EvalFrameDefault
-      43: _PyFunction_FastCallKeywords
-      42: _PyEval_EvalCodeWithName
-      41: _PyEval_EvalFrameDefault
-      40: _PyMethodDef_RawFastCallKeywords
-      39: 0x0000000000546369
-      38: _PyEval_EvalCodeWithName
-      37: _PyEval_EvalFrameDefault
-      36: _PyFunction_FastCallKeywords
-      35: _PyEval_EvalCodeWithName
-      34: _PyEval_EvalFrameDefault
-      33: _PyFunction_FastCallDict
-      32: _PyEval_EvalCodeWithName
-      31: _PyEval_EvalFrameDefault
-      30: _PyObject_FastCallDict
-      29: 0x00000000004c06e1
-      28: _PyFunction_FastCallDict
-      27: _PyEval_EvalFrameDefault
-      26: _PyMethodDescr_FastCallKeywords
-      25: 0x00000000005dcb58
-      24: 0x00000000005dc83f
-      23: 0x00000000004ba127
-      22: _PyEval_EvalFrameDefault
-      21: _PyFunction_FastCallKeywords
-      20: _PyEval_EvalFrameDefault
-      19: _PyFunction_FastCallKeywords
-      18: _PyEval_EvalFrameDefault
-      17: _PyFunction_FastCallKeywords
-      16: _PyEval_EvalCodeWithName
-      15: _PyEval_EvalFrameDefault
-      14: 0x0000000000537c30
-      13: _PyObject_FastCallKeywords
-      12: 0x00007f1e3ba10fa2
-      11: _ctypes_callproc
-      10: ffi_call
-      9: ffi_call_unix64
-      8: TVMModGetFunction
-            at ../src/runtime/c_runtime_api.cc:408
-      7: tvm::runtime::ModuleNode::GetFunction(std::__cxx11::basic_string<char, std::char_traits<char>, std::allocator<char> > const&, bool)
-            at ../src/runtime/module.cc:66
-      6: tvm::runtime::RPCModuleNode::GetFunction(std::__cxx11::basic_string<char, std::char_traits<char>, std::allocator<char> > const&, tvm::runtime::ObjectPtr<tvm::runtime::Object> const&)
-            at ../src/runtime/rpc/rpc_module.cc:185
-      5: tvm::runtime::RPCClientSession::GetFunction(std::__cxx11::basic_string<char, std::char_traits<char>, std::allocator<char> > const&)
-            at ../src/runtime/rpc/rpc_endpoint.cc:1007
-      4: tvm::runtime::TVMRetValue tvm::runtime::RPCEndpoint::SysCallRemote<std::__cxx11::basic_string<char, std::char_traits<char>, std::allocator<char> > const&>(tvm::runtime::RPCCode, std::__cxx11::basic_string<char, std::char_traits<char>, std::allocator<char> > const&)
-            at ../src/runtime/rpc/rpc_endpoint.h:223
-      3: tvm::runtime::TVMRetValue tvm::runtime::PackedFunc::operator()<int, std::__cxx11::basic_string<char, std::char_traits<char>, std::allocator<char> > const&>(int&&, std::__cxx11::basic_string<char, std::char_traits<char>, std::allocator<char> > const&) const
-            at ../include/tvm/runtime/packed_func.h:1617
-      2: tvm::runtime::PackedFuncObj::CallPacked(tvm::runtime::TVMArgs, tvm::runtime::TVMRetValue*) const
-            at ../include/tvm/runtime/packed_func.h:1217
-      1: Call
-            at ../include/tvm/runtime/packed_func.h:1213
-      0: operator()
-            at ../src/runtime/rpc/rpc_endpoint.cc:684
-      File "../src/runtime/rpc/rpc_endpoint.cc", line 684
-    TVMError: 
-    ---------------------------------------------------------------
-    An error occurred during the execution of TVM.
-    For more information, please see: https://tvm.apache.org/docs/errors.html
-    ---------------------------------------------------------------
-      Check failed: (code == RPCCode::kReturn) is false: code=1
-
-    Traceback (most recent call last):
-      52: 0xffffffffffffffff
-      51: _start
-      50: __libc_start_main
-      49: _Py_UnixMain
-      48: 0x0000000000650da0
-      47: 0x0000000000650afa
-      46: _PyFunction_FastCallDict
-      45: _PyEval_EvalCodeWithName
-      44: _PyEval_EvalFrameDefault
-      43: _PyFunction_FastCallKeywords
-      42: _PyEval_EvalCodeWithName
-      41: _PyEval_EvalFrameDefault
-      40: _PyMethodDef_RawFastCallKeywords
-      39: 0x0000000000546369
-      38: _PyEval_EvalCodeWithName
-      37: _PyEval_EvalFrameDefault
-      36: _PyFunction_FastCallKeywords
-      35: _PyEval_EvalCodeWithName
-      34: _PyEval_EvalFrameDefault
-      33: _PyFunction_FastCallDict
-      32: _PyEval_EvalCodeWithName
-      31: _PyEval_EvalFrameDefault
-      30: _PyObject_FastCallDict
-      29: 0x00000000004c06e1
-      28: _PyFunction_FastCallDict
-      27: _PyEval_EvalFrameDefault
-      26: _PyMethodDescr_FastCallKeywords
-      25: 0x00000000005dcb58
-      24: 0x00000000005dc83f
-      23: 0x00000000004ba127
-      22: _PyEval_EvalFrameDefault
-      21: _PyFunction_FastCallKeywords
-      20: _PyEval_EvalFrameDefault
-      19: _PyFunction_FastCall      [('tile_f', [-1, 2, 1, 16]), ('tile_y', [-1, 1, 1, 7]), ('tile_x', [-1, 1, 1, 7]), ('tile_rc', [-1, 1, 4]), ('tile_ry', [-1, 3, 1]), ('tile_rx', [-1, 1, 3]), ('auto_unroll_max_step', 512), ('unroll_explicit', 1)],None,8395145
-    No: 7   GFLOPS: 0.00/57.82      result: Traceback (most recent call last):
+    tvm.autotvm.task.space.InstantiationError: Skipped because of invalid gpu kernel        [('tile_f', [-1, 4, 128, 1]), ('tile_y', [-1, 7, 1, 1]), ('tile_x', [-1, 1, 1, 7]), ('tile_rc', [-1, 4, 16]), ('tile_ry', [-1, 1, 3]), ('tile_rx', [-1, 1, 3]), ('auto_unroll_max_step', 512), ('unroll_explicit', 0)],None,3420831
+    No: 5   GFLOPS: 32.37/32.37     result: MeasureResult(costs=(0.007151188529411765,), error_no=MeasureErrorNo.NO_ERROR, all_cost=3.697486400604248, timestamp=1673400515.123984) [('tile_f', [-1, 1, 1, 32]), ('tile_y', [-1, 1, 1, 7]), ('tile_x', [-1, 1, 7, 1]), ('tile_rc', [-1, 2, 2]), ('tile_ry', [-1, 1, 1]), ('tile_rx', [-1, 3, 1]), ('auto_unroll_max_step', 1500), ('unroll_explicit', 1)],None,9334125
+    No: 6   GFLOPS: 0.00/32.37      result: Traceback (most recent call last):
       File "/workspace/python/tvm/autotvm/measure/measure_methods.py", line 592, in __call__
         func, arg_info = _build_func_common(measure_input, self.runtime, **kwargs)
       File "/workspace/python/tvm/autotvm/measure/measure_methods.py", line 544, in _build_func_common
@@ -1037,8 +884,9 @@ for this template
       File "tvm/_ffi/_cython/./packed_func.pxi", line 56, in tvm._ffi._cy3.core.tvm_callback
       File "/workspace/python/tvm/autotvm/measure/measure_methods.py", line 875, in verify_pass
         raise InstantiationError("Skipped because of invalid gpu kernel")
-    tvm.autotvm.task.space.InstantiationError: Skipped because of invalid gpu kernel        [('tile_f', [-1, 1, 16, 16]), ('tile_y', [-1, 1, 1, 7]), ('tile_x', [-1, 1, 7, 1]), ('tile_rc', [-1, 128, 4]), ('tile_ry', [-1, 1, 3]), ('tile_rx', [-1, 1, 1]), ('auto_unroll_max_step', 1500), ('unroll_explicit', 0)],None,3966122
-    No: 8   GFLOPS: 0.00/57.82      result: Traceback (most recent call last):
+    tvm.autotvm.task.space.InstantiationError: Skipped because of invalid gpu kernel        [('tile_f', [-1, 32, 1, 4]), ('tile_y', [-1, 1, 7, 1]), ('tile_x', [-1, 1, 1, 7]), ('tile_rc', [-1, 64, 1]), ('tile_ry', [-1, 1, 3]), ('tile_rx', [-1, 3, 1]), ('auto_unroll_max_step', 512), ('unroll_explicit', 1)],None,7961905
+    No: 7   GFLOPS: 212.78/212.78   result: MeasureResult(costs=(0.0010879592826086957,), error_no=MeasureErrorNo.NO_ERROR, all_cost=1.0672731399536133, timestamp=1673400516.8288145)      [('tile_f', [-1, 2, 8, 8]), ('tile_y', [-1, 1, 7, 1]), ('tile_x', [-1, 1, 1, 1]), ('tile_rc', [-1, 1, 1]), ('tile_ry', [-1, 1, 1]), ('tile_rx', [-1, 1, 3]), ('auto_unroll_max_step', 512), ('unroll_explicit', 0)],None,2904595
+    No: 8   GFLOPS: 0.00/212.78     result: Traceback (most recent call last):
       File "/workspace/python/tvm/autotvm/measure/measure_methods.py", line 592, in __call__
         func, arg_info = _build_func_common(measure_input, self.runtime, **kwargs)
       File "/workspace/python/tvm/autotvm/measure/measure_methods.py", line 544, in _build_func_common
@@ -1160,8 +1008,8 @@ for this template
       File "tvm/_ffi/_cython/./packed_func.pxi", line 56, in tvm._ffi._cy3.core.tvm_callback
       File "/workspace/python/tvm/autotvm/measure/measure_methods.py", line 875, in verify_pass
         raise InstantiationError("Skipped because of invalid gpu kernel")
-    tvm.autotvm.task.space.InstantiationError: Skipped because of invalid gpu kernel        [('tile_f', [-1, 16, 2, 2]), ('tile_y', [-1, 7, 1, 1]), ('tile_x', [-1, 1, 1, 7]), ('tile_rc', [-1, 16, 32]), ('tile_ry', [-1, 3, 1]), ('tile_rx', [-1, 1, 1]), ('auto_unroll_max_step', 0), ('unroll_explicit', 0)],None,351408
-    No: 9   GFLOPS: 0.00/57.82      result: Traceback (most recent call last):
+    tvm.autotvm.task.space.InstantiationError: Skipped because of invalid gpu kernel        [('tile_f', [-1, 1, 256, 2]), ('tile_y', [-1, 1, 1, 1]), ('tile_x', [-1, 1, 1, 7]), ('tile_rc', [-1, 1, 64]), ('tile_ry', [-1, 3, 1]), ('tile_rx', [-1, 1, 1]), ('auto_unroll_max_step', 512), ('unroll_explicit', 0)],None,2097139
+    No: 9   GFLOPS: 0.00/212.78     result: Traceback (most recent call last):
       File "/workspace/python/tvm/autotvm/measure/measure_methods.py", line 592, in __call__
         func, arg_info = _build_func_common(measure_input, self.runtime, **kwargs)
       File "/workspace/python/tvm/autotvm/measure/measure_methods.py", line 544, in _build_func_common
@@ -1283,26 +1131,8 @@ for this template
       File "tvm/_ffi/_cython/./packed_func.pxi", line 56, in tvm._ffi._cy3.core.tvm_callback
       File "/workspace/python/tvm/autotvm/measure/measure_methods.py", line 875, in verify_pass
         raise InstantiationError("Skipped because of invalid gpu kernel")
-    tvm.autotvm.task.space.InstantiationError: Skipped because of invalid gpu kernel        [('tile_f', [-1, 2, 4, 32]), ('tile_y', [-1, 1, 1, 1]), ('tile_x', [-1, 1, 1, 1]), ('tile_rc', [-1, 64, 4]), ('tile_ry', [-1, 3, 1]), ('tile_rx', [-1, 3, 1]), ('auto_unroll_max_step', 0), ('unroll_explicit', 0)],None,862595
-    No: 10  GFLOPS: 0.00/57.82      result: Traceback (most recent call last):
-      File "/workspace/python/tvm/autotvm/measure/measure_methods.py", line 142, in build
-        res = future.result()
-      File "/usr/lib/python3.7/concurrent/futures/_base.py", line 435, in result
-        return self.__get_result()
-      File "/usr/lib/python3.7/concurrent/futures/_base.py", line 384, in __get_result
-        raise self._exception
-      File "/usr/lib/python3.7/concurrent/futures/thread.py", line 57, in run
-        result = self.fn(*self.args, **self.kwargs)
-      File "/workspace/python/tvm/contrib/popen_pool.py", line 432, in <lambda>
-        worker = lambda *args: self._worker_run(*args)
-      File "/workspace/python/tvm/contrib/popen_pool.py", line 401, in _worker_run
-        return proc.recv()
-      File "/workspace/python/tvm/contrib/popen_pool.py", line 309, in recv
-        raise TimeoutError()
-    TimeoutError
-
-            [('tile_f', [-1, 2, 8, 8]), ('tile_y', [-1, 7, 1, 1]), ('tile_x', [-1, 7, 1, 1]), ('tile_rc', [-1, 64, 1]), ('tile_ry', [-1, 1, 1]), ('tile_rx', [-1, 1, 1]), ('auto_unroll_max_step', 1500), ('unroll_explicit', 1)],None,8734375
-    No: 11  GFLOPS: 0.00/57.82      result: Traceback (most recent call last):
+    tvm.autotvm.task.space.InstantiationError: Skipped because of invalid gpu kernel        [('tile_f', [-1, 128, 1, 1]), ('tile_y', [-1, 1, 7, 1]), ('tile_x', [-1, 1, 1, 7]), ('tile_rc', [-1, 1, 32]), ('tile_ry', [-1, 1, 3]), ('tile_rx', [-1, 1, 1]), ('auto_unroll_max_step', 0), ('unroll_explicit', 0)],None,531087
+    No: 10  GFLOPS: 0.00/212.78     result: Traceback (most recent call last):
       File "/workspace/python/tvm/autotvm/measure/measure_methods.py", line 592, in __call__
         func, arg_info = _build_func_common(measure_input, self.runtime, **kwargs)
       File "/workspace/python/tvm/autotvm/measure/measure_methods.py", line 544, in _build_func_common
@@ -1424,8 +1254,8 @@ for this template
       File "tvm/_ffi/_cython/./packed_func.pxi", line 56, in tvm._ffi._cy3.core.tvm_callback
       File "/workspace/python/tvm/autotvm/measure/measure_methods.py", line 875, in verify_pass
         raise InstantiationError("Skipped because of invalid gpu kernel")
-    tvm.autotvm.task.space.InstantiationError: Skipped because of invalid gpu kernel        [('tile_f', [-1, 16, 2, 2]), ('tile_y', [-1, 1, 1, 1]), ('tile_x', [-1, 1, 7, 1]), ('tile_rc', [-1, 8, 16]), ('tile_ry', [-1, 1, 3]), ('tile_rx', [-1, 1, 1]), ('auto_unroll_max_step', 1500), ('unroll_explicit', 1)],None,9231268
-    No: 12  GFLOPS: 0.00/57.82      result: Traceback (most recent call last):
+    tvm.autotvm.task.space.InstantiationError: Skipped because of invalid gpu kernel        [('tile_f', [-1, 1, 4, 32]), ('tile_y', [-1, 7, 1, 1]), ('tile_x', [-1, 1, 1, 1]), ('tile_rc', [-1, 4, 128]), ('tile_ry', [-1, 1, 3]), ('tile_rx', [-1, 1, 1]), ('auto_unroll_max_step', 512), ('unroll_explicit', 1)],None,7536734
+    No: 11  GFLOPS: 0.00/212.78     result: Traceback (most recent call last):
       File "/workspace/python/tvm/autotvm/measure/measure_methods.py", line 592, in __call__
         func, arg_info = _build_func_common(measure_input, self.runtime, **kwargs)
       File "/workspace/python/tvm/autotvm/measure/measure_methods.py", line 544, in _build_func_common
@@ -1547,8 +1377,27 @@ for this template
       File "tvm/_ffi/_cython/./packed_func.pxi", line 56, in tvm._ffi._cy3.core.tvm_callback
       File "/workspace/python/tvm/autotvm/measure/measure_methods.py", line 875, in verify_pass
         raise InstantiationError("Skipped because of invalid gpu kernel")
-    tvm.autotvm.task.space.InstantiationError: Skipped because of invalid gpu kernel        [('tile_f', [-1, 1, 1, 8]), ('tile_y', [-1, 1, 7, 1]), ('tile_x', [-1, 1, 7, 1]), ('tile_rc', [-1, 16, 16]), ('tile_ry', [-1, 1, 3]), ('tile_rx', [-1, 3, 1]), ('auto_unroll_max_step', 512), ('unroll_explicit', 0)],None,2846496
-    No: 13  GFLOPS: 0.00/57.82      result: Traceback (most recent call last):
+    tvm.autotvm.task.space.InstantiationError: Skipped because of invalid gpu kernel        [('tile_f', [-1, 256, 1, 1]), ('tile_y', [-1, 7, 1, 1]), ('tile_x', [-1, 1, 1, 7]), ('tile_rc', [-1, 1, 256]), ('tile_ry', [-1, 3, 1]), ('tile_rx', [-1, 1, 3]), ('auto_unroll_max_step', 512), ('unroll_explicit', 0)],None,3283508
+    No: 12  GFLOPS: 189.23/212.78   result: MeasureResult(costs=(0.0012233553536585365,), error_no=MeasureErrorNo.NO_ERROR, all_cost=1.4947834014892578, timestamp=1673400518.5619686)      [('tile_f', [-1, 1, 32, 4]), ('tile_y', [-1, 1, 1, 7]), ('tile_x', [-1, 1, 1, 1]), ('tile_rc', [-1, 2, 4]), ('tile_ry', [-1, 3, 1]), ('tile_rx', [-1, 1, 3]), ('auto_unroll_max_step', 512), ('unroll_explicit', 0)],None,3168790
+    No: 13  GFLOPS: 0.00/212.78     result: Traceback (most recent call last):
+      File "/workspace/python/tvm/autotvm/measure/measure_methods.py", line 142, in build
+        res = future.result()
+      File "/usr/lib/python3.7/concurrent/futures/_base.py", line 435, in result
+        return self.__get_result()
+      File "/usr/lib/python3.7/concurrent/futures/_base.py", line 384, in __get_result
+        raise self._exception
+      File "/usr/lib/python3.7/concurrent/futures/thread.py", line 57, in run
+        result = self.fn(*self.args, **self.kwargs)
+      File "/workspace/python/tvm/contrib/popen_pool.py", line 432, in <lambda>
+        worker = lambda *args: self._worker_run(*args)
+      File "/workspace/python/tvm/contrib/popen_pool.py", line 401, in _worker_run
+        return proc.recv()
+      File "/workspace/python/tvm/contrib/popen_pool.py", line 309, in recv
+        raise TimeoutError()
+    TimeoutError
+
+            [('tile_f', [-1, 1, 4, 4]), ('tile_y', [-1, 1, 1, 1]), ('tile_x', [-1, 7, 1, 1]), ('tile_rc', [-1, 2, 16]), ('tile_ry', [-1, 1, 3]), ('tile_rx', [-1, 3, 1]), ('auto_unroll_max_step', 1500), ('unroll_explicit', 1)],None,9804195
+    No: 14  GFLOPS: 0.00/212.78     result: Traceback (most recent call last):
       File "/workspace/python/tvm/autotvm/measure/measure_methods.py", line 592, in __call__
         func, arg_info = _build_func_common(measure_input, self.runtime, **kwargs)
       File "/workspace/python/tvm/autotvm/measure/measure_methods.py", line 544, in _build_func_common
@@ -1670,9 +1519,8 @@ for this template
       File "tvm/_ffi/_cython/./packed_func.pxi", line 56, in tvm._ffi._cy3.core.tvm_callback
       File "/workspace/python/tvm/autotvm/measure/measure_methods.py", line 875, in verify_pass
         raise InstantiationError("Skipped because of invalid gpu kernel")
-    tvm.autotvm.task.space.InstantiationError: Skipped because of invalid gpu kernel        [('tile_f', [-1, 64, 2, 1]), ('tile_y', [-1, 7, 1, 1]), ('tile_x', [-1, 1, 1, 1]), ('tile_rc', [-1, 16, 4]), ('tile_ry', [-1, 1, 3]), ('tile_rx', [-1, 3, 1]), ('auto_unroll_max_step', 0), ('unroll_explicit', 1)],None,6276396
-    No: 14  GFLOPS: 6.98/57.82      result: MeasureResult(costs=(0.03316578275,), error_no=MeasureErrorNo.NO_ERROR, all_cost=2.9679858684539795, timestamp=1673392889.1873596)      [('tile_f', [-1, 4, 8, 16]), ('tile_y', [-1, 7, 1, 1]), ('tile_x', [-1, 1, 7, 1]), ('tile_rc', [-1, 8, 2]), ('tile_ry', [-1, 1, 1]), ('tile_rx', [-1, 1, 1]), ('auto_unroll_max_step', 512), ('unroll_explicit', 0)],None,1790321
-    No: 15  GFLOPS: 0.00/57.82      result: Traceback (most recent call last):
+    tvm.autotvm.task.space.InstantiationError: Skipped because of invalid gpu kernel        [('tile_f', [-1, 1, 4, 32]), ('tile_y', [-1, 7, 1, 1]), ('tile_x', [-1, 1, 1, 7]), ('tile_rc', [-1, 256, 2]), ('tile_ry', [-1, 1, 1]), ('tile_rx', [-1, 1, 3]), ('auto_unroll_max_step', 1500), ('unroll_explicit', 1)],None,9940014
+    No: 15  GFLOPS: 0.00/212.78     result: Traceback (most recent call last):
       File "/workspace/python/tvm/autotvm/measure/measure_methods.py", line 592, in __call__
         func, arg_info = _build_func_common(measure_input, self.runtime, **kwargs)
       File "/workspace/python/tvm/autotvm/measure/measure_methods.py", line 544, in _build_func_common
@@ -1794,8 +1642,8 @@ for this template
       File "tvm/_ffi/_cython/./packed_func.pxi", line 56, in tvm._ffi._cy3.core.tvm_callback
       File "/workspace/python/tvm/autotvm/measure/measure_methods.py", line 875, in verify_pass
         raise InstantiationError("Skipped because of invalid gpu kernel")
-    tvm.autotvm.task.space.InstantiationError: Skipped because of invalid gpu kernel        [('tile_f', [-1, 2, 128, 1]), ('tile_y', [-1, 1, 7, 1]), ('tile_x', [-1, 1, 1, 7]), ('tile_rc', [-1, 1, 1]), ('tile_ry', [-1, 1, 1]), ('tile_rx', [-1, 3, 1]), ('auto_unroll_max_step', 0), ('unroll_explicit', 0)],None,583930
-    No: 16  GFLOPS: 0.00/57.82      result: Traceback (most recent call last):
+    tvm.autotvm.task.space.InstantiationError: Skipped because of invalid gpu kernel        [('tile_f', [-1, 4, 8, 16]), ('tile_y', [-1, 1, 1, 1]), ('tile_x', [-1, 1, 1, 7]), ('tile_rc', [-1, 256, 2]), ('tile_ry', [-1, 1, 3]), ('tile_rx', [-1, 1, 1]), ('auto_unroll_max_step', 1500), ('unroll_explicit', 0)],None,3938181
+    No: 16  GFLOPS: 0.00/212.78     result: Traceback (most recent call last):
       File "/workspace/python/tvm/autotvm/measure/measure_methods.py", line 592, in __call__
         func, arg_info = _build_func_common(measure_input, self.runtime, **kwargs)
       File "/workspace/python/tvm/autotvm/measure/measure_methods.py", line 544, in _build_func_common
@@ -1917,8 +1765,8 @@ for this template
       File "tvm/_ffi/_cython/./packed_func.pxi", line 56, in tvm._ffi._cy3.core.tvm_callback
       File "/workspace/python/tvm/autotvm/measure/measure_methods.py", line 875, in verify_pass
         raise InstantiationError("Skipped because of invalid gpu kernel")
-    tvm.autotvm.task.space.InstantiationError: Skipped because of invalid gpu kernel        [('tile_f', [-1, 4, 8, 2]), ('tile_y', [-1, 1, 7, 1]), ('tile_x', [-1, 1, 7, 1]), ('tile_rc', [-1, 1, 128]), ('tile_ry', [-1, 1, 3]), ('tile_rx', [-1, 3, 1]), ('auto_unroll_max_step', 1500), ('unroll_explicit', 1)],None,9854761
-    No: 17  GFLOPS: 0.00/57.82      result: Traceback (most recent call last):
+    tvm.autotvm.task.space.InstantiationError: Skipped because of invalid gpu kernel        [('tile_f', [-1, 1, 32, 1]), ('tile_y', [-1, 1, 1, 1]), ('tile_x', [-1, 1, 1, 7]), ('tile_rc', [-1, 64, 8]), ('tile_ry', [-1, 1, 3]), ('tile_rx', [-1, 3, 1]), ('auto_unroll_max_step', 0), ('unroll_explicit', 0)],None,1086840
+    No: 17  GFLOPS: 0.00/212.78     result: Traceback (most recent call last):
       File "/workspace/python/tvm/autotvm/measure/measure_methods.py", line 592, in __call__
         func, arg_info = _build_func_common(measure_input, self.runtime, **kwargs)
       File "/workspace/python/tvm/autotvm/measure/measure_methods.py", line 544, in _build_func_common
@@ -2040,8 +1888,8 @@ for this template
       File "tvm/_ffi/_cython/./packed_func.pxi", line 56, in tvm._ffi._cy3.core.tvm_callback
       File "/workspace/python/tvm/autotvm/measure/measure_methods.py", line 875, in verify_pass
         raise InstantiationError("Skipped because of invalid gpu kernel")
-    tvm.autotvm.task.space.InstantiationError: Skipped because of invalid gpu kernel        [('tile_f', [-1, 2, 128, 2]), ('tile_y', [-1, 1, 1, 1]), ('tile_x', [-1, 1, 1, 1]), ('tile_rc', [-1, 1, 8]), ('tile_ry', [-1, 3, 1]), ('tile_rx', [-1, 1, 1]), ('auto_unroll_max_step', 1500), ('unroll_explicit', 1)],None,9000738
-    No: 18  GFLOPS: 0.00/57.82      result: Traceback (most recent call last):
+    tvm.autotvm.task.space.InstantiationError: Skipped because of invalid gpu kernel        [('tile_f', [-1, 32, 8, 1]), ('tile_y', [-1, 1, 7, 1]), ('tile_x', [-1, 7, 1, 1]), ('tile_rc', [-1, 16, 1]), ('tile_ry', [-1, 3, 1]), ('tile_rx', [-1, 1, 1]), ('auto_unroll_max_step', 1500), ('unroll_explicit', 1)],None,8921032
+    No: 18  GFLOPS: 0.00/212.78     result: Traceback (most recent call last):
       File "/workspace/python/tvm/autotvm/measure/measure_methods.py", line 592, in __call__
         func, arg_info = _build_func_common(measure_input, self.runtime, **kwargs)
       File "/workspace/python/tvm/autotvm/measure/measure_methods.py", line 544, in _build_func_common
@@ -2163,9 +2011,9 @@ for this template
       File "tvm/_ffi/_cython/./packed_func.pxi", line 56, in tvm._ffi._cy3.core.tvm_callback
       File "/workspace/python/tvm/autotvm/measure/measure_methods.py", line 875, in verify_pass
         raise InstantiationError("Skipped because of invalid gpu kernel")
-    tvm.autotvm.task.space.InstantiationError: Skipped because of invalid gpu kernel        [('tile_f', [-1, 2, 128, 2]), ('tile_y', [-1, 1, 1, 1]), ('tile_x', [-1, 7, 1, 1]), ('tile_rc', [-1, 4, 4]), ('tile_ry', [-1, 1, 1]), ('tile_rx', [-1, 1, 1]), ('auto_unroll_max_step', 512), ('unroll_explicit', 0)],None,1817298
-    No: 19  GFLOPS: 4.59/57.82      result: MeasureResult(costs=(0.050386360000000005,), error_no=MeasureErrorNo.NO_ERROR, all_cost=2.615889549255371, timestamp=1673392892.0286698)        [('tile_f', [-1, 8, 2, 32]), ('tile_y', [-1, 1, 1, 1]), ('tile_x', [-1, 1, 1, 1]), ('tile_rc', [-1, 2, 4]), ('tile_ry', [-1, 1, 1]), ('tile_rx', [-1, 1, 1]), ('auto_unroll_max_step', 1500), ('unroll_explicit', 0)],None,3555393
-    No: 20  GFLOPS: 0.00/57.82      result: Traceback (most recent call last):
+    tvm.autotvm.task.space.InstantiationError: Skipped because of invalid gpu kernel        [('tile_f', [-1, 32, 1, 8]), ('tile_y', [-1, 1, 1, 1]), ('tile_x', [-1, 1, 1, 7]), ('tile_rc', [-1, 16, 4]), ('tile_ry', [-1, 1, 1]), ('tile_rx', [-1, 1, 1]), ('auto_unroll_max_step', 0), ('unroll_explicit', 1)],None,5310941
+    No: 19  GFLOPS: 1.11/212.78     result: MeasureResult(costs=(0.20837886025,), error_no=MeasureErrorNo.NO_ERROR, all_cost=4.230514049530029, timestamp=1673400533.3823516)       [('tile_f', [-1, 1, 1, 1]), ('tile_y', [-1, 7, 1, 1]), ('tile_x', [-1, 1, 1, 7]), ('tile_rc', [-1, 2, 32]), ('tile_ry', [-1, 1, 1]), ('tile_rx', [-1, 1, 1]), ('auto_unroll_max_step', 0), ('unroll_explicit', 1)],None,5374380
+    No: 20  GFLOPS: 0.00/212.78     result: Traceback (most recent call last):
       File "/workspace/python/tvm/autotvm/measure/measure_methods.py", line 592, in __call__
         func, arg_info = _build_func_common(measure_input, self.runtime, **kwargs)
       File "/workspace/python/tvm/autotvm/measure/measure_methods.py", line 544, in _build_func_common
@@ -2287,7 +2135,7 @@ for this template
       File "tvm/_ffi/_cython/./packed_func.pxi", line 56, in tvm._ffi._cy3.core.tvm_callback
       File "/workspace/python/tvm/autotvm/measure/measure_methods.py", line 875, in verify_pass
         raise InstantiationError("Skipped because of invalid gpu kernel")
-    tvm.autotvm.task.space.InstantiationError: Skipped because of invalid gpu kernel        [('tile_f', [-1, 2, 16, 4]), ('tile_y', [-1, 1, 1, 1]), ('tile_x', [-1, 1, 1, 7]), ('tile_rc', [-1, 4, 128]), ('tile_ry', [-1, 3, 1]), ('tile_rx', [-1, 1, 3]), ('auto_unroll_max_step', 1500), ('unroll_explicit', 1)],None,10249487
+    tvm.autotvm.task.space.InstantiationError: Skipped because of invalid gpu kernel        [('tile_f', [-1, 128, 4, 1]), ('tile_y', [-1, 1, 1, 7]), ('tile_x', [-1, 7, 1, 1]), ('tile_rc', [-1, 16, 8]), ('tile_ry', [-1, 1, 3]), ('tile_rx', [-1, 1, 1]), ('auto_unroll_max_step', 1500), ('unroll_explicit', 1)],None,9209886
 
 
 
@@ -2342,9 +2190,9 @@ and measure running time.
     Finish loading 20 records
 
     Best config:
-    [('tile_f', [-1, 2, 2, 2]), ('tile_y', [-1, 1, 1, 1]), ('tile_x', [-1, 1, 7, 1]), ('tile_rc', [-1, 4, 8]), ('tile_ry', [-1, 1, 3]), ('tile_rx', [-1, 3, 1]), ('auto_unroll_max_step', 0), ('unroll_explicit', 0)],None,1071905
+    [('tile_f', [-1, 2, 8, 8]), ('tile_y', [-1, 1, 7, 1]), ('tile_x', [-1, 1, 1, 1]), ('tile_rc', [-1, 1, 1]), ('tile_ry', [-1, 1, 1]), ('tile_rx', [-1, 1, 3]), ('auto_unroll_max_step', 512), ('unroll_explicit', 0)],None,2904595
     Finish loading 20 records
-    Time cost of this operator: 0.002055
+    Time cost of this operator: 0.001817
 
 
 
diff --git a/docs/_sources/how_to/work_with_microtvm/micro_autotune.rst.txt b/docs/_sources/how_to/work_with_microtvm/micro_autotune.rst.txt
index 56fdf6c06c..a267dec377 100644
--- a/docs/_sources/how_to/work_with_microtvm/micro_autotune.rst.txt
+++ b/docs/_sources/how_to/work_with_microtvm/micro_autotune.rst.txt
@@ -368,10 +368,10 @@ Timing the untuned program
     ########## Build without Autotuning ##########
     Node Name                                     Ops                                           Time(us)  Time(%)  Shape              Inputs  Outputs  Measurements(us)  
     ---------                                     ---                                           --------  -------  -----              ------  -------  ----------------  
-    tvmgen_default_fused_nn_contrib_conv2d_NCHWc  tvmgen_default_fused_nn_contrib_conv2d_NCHWc  311.6     98.731   (1, 2, 10, 10, 3)  2       1        [311.6]           
-    tvmgen_default_fused_layout_transform_1       tvmgen_default_fused_layout_transform_1       3.024     0.958    (1, 6, 10, 10)     1       1        [3.024]           
-    tvmgen_default_fused_layout_transform         tvmgen_default_fused_layout_transform         0.982     0.311    (1, 1, 10, 10, 3)  1       1        [0.982]           
-    Total_time                                    -                                             315.606   -        -                  -       -        -                 
+    tvmgen_default_fused_nn_contrib_conv2d_NCHWc  tvmgen_default_fused_nn_contrib_conv2d_NCHWc  310.5     98.715   (1, 2, 10, 10, 3)  2       1        [310.5]           
+    tvmgen_default_fused_layout_transform_1       tvmgen_default_fused_layout_transform_1       3.078     0.979    (1, 6, 10, 10)     1       1        [3.078]           
+    tvmgen_default_fused_layout_transform         tvmgen_default_fused_layout_transform         0.963     0.306    (1, 1, 10, 10, 3)  1       1        [0.963]           
+    Total_time                                    -                                             314.541   -        -                  -       -        -                 
 
 
 
@@ -436,10 +436,10 @@ Timing the tuned program
     ########## Build with Autotuning ##########
     Node Name                                     Ops                                           Time(us)  Time(%)  Shape              Inputs  Outputs  Measurements(us)  
     ---------                                     ---                                           --------  -------  -----              ------  -------  ----------------  
-    tvmgen_default_fused_nn_contrib_conv2d_NCHWc  tvmgen_default_fused_nn_contrib_conv2d_NCHWc  103.6     97.527   (1, 6, 10, 10, 1)  2       1        [103.6]           
-    tvmgen_default_fused_layout_transform_1       tvmgen_default_fused_layout_transform_1       1.782     1.678    (1, 6, 10, 10)     1       1        [1.782]           
-    tvmgen_default_fused_layout_transform         tvmgen_default_fused_layout_transform         0.845     0.795    (1, 3, 10, 10, 1)  1       1        [0.845]           
-    Total_time                                    -                                             106.227   -        -                  -       -        -                 
+    tvmgen_default_fused_nn_contrib_conv2d_NCHWc  tvmgen_default_fused_nn_contrib_conv2d_NCHWc  100.3     97.314   (1, 6, 10, 10, 1)  2       1        [100.3]           
+    tvmgen_default_fused_layout_transform_1       tvmgen_default_fused_layout_transform_1       1.784     1.731    (1, 6, 10, 10)     1       1        [1.784]           
+    tvmgen_default_fused_layout_transform         tvmgen_default_fused_layout_transform         0.984     0.955    (1, 1, 10, 10, 3)  1       1        [0.984]           
+    Total_time                                    -                                             103.069   -        -                  -       -        -                 
 
 
 
diff --git a/docs/_sources/how_to/work_with_microtvm/micro_pytorch.rst.txt b/docs/_sources/how_to/work_with_microtvm/micro_pytorch.rst.txt
index a828942f53..204e6b7f4f 100644
--- a/docs/_sources/how_to/work_with_microtvm/micro_pytorch.rst.txt
+++ b/docs/_sources/how_to/work_with_microtvm/micro_pytorch.rst.txt
@@ -117,7 +117,7 @@ download a cat image and preprocess it to use as the model input.
     /venv/apache-tvm-py3.7/lib/python3.7/site-packages/torch/ao/quantization/utils.py:281: UserWarning: must run observer before calling calculate_qparams. Returning default values.
       "must run observer before calling calculate_qparams. " +
     Downloading: "https://download.pytorch.org/models/quantized/mobilenet_v2_qnnpack_37f702c5.pth" to /workspace/.cache/torch/hub/checkpoints/mobilenet_v2_qnnpack_37f702c5.pth
-
      0%|          | 0.00/3.42M [00:00<?, ?B/s]
    100%|##########| 3.42M/3.42M [00:00<00:00, 35.8MB/s]
    100%|##########| 3.42M/3.42M [00:00<00:00, 35.7MB/s]
+
      0%|          | 0.00/3.42M [00:00<?, ?B/s]
    100%|##########| 3.42M/3.42M [00:00<00:00, 37.9MB/s]
     /workspace/python/tvm/relay/frontend/pytorch_utils.py:47: DeprecationWarning: distutils Version classes are deprecated. Use packaging.version instead.
       return LooseVersion(torch_ver) > ver
     /venv/apache-tvm-py3.7/lib/python3.7/site-packages/setuptools/_distutils/version.py:346: DeprecationWarning: distutils Version classes are deprecated. Use packaging.version instead.
@@ -322,7 +322,7 @@ Look up prediction top 1 index in 1000 class synset.
 
 .. rst-class:: sphx-glr-timing
 
-   **Total running time of the script:** ( 1 minutes  5.814 seconds)
+   **Total running time of the script:** ( 1 minutes  3.361 seconds)
 
 
 .. _sphx_glr_download_how_to_work_with_microtvm_micro_pytorch.py:
diff --git a/docs/_sources/how_to/work_with_microtvm/micro_train.rst.txt b/docs/_sources/how_to/work_with_microtvm/micro_train.rst.txt
index aa04d25fe9..a6531ad9f2 100644
--- a/docs/_sources/how_to/work_with_microtvm/micro_train.rst.txt
+++ b/docs/_sources/how_to/work_with_microtvm/micro_train.rst.txt
@@ -218,7 +218,7 @@ take about **2 minutes** to download the Stanford Cars, while COCO 2017 validati
  .. code-block:: none
 
 
-    '/tmp/tmpk1esgsax/images/random'
+    '/tmp/tmp1fwqht9d/images/random'
 
 
 
@@ -309,7 +309,7 @@ objects to other stuff? We can display some examples from our datasets using ``m
 
 
 .. image-sg:: /how_to/work_with_microtvm/images/sphx_glr_micro_train_001.png
-   :alt: [0.0, 1.0], [0.0, 1.0], [1.0, 0.0], [0.0, 1.0], [0.0, 1.0], [1.0, 0.0], [1.0, 0.0], [1.0, 0.0], [1.0, 0.0], [0.0, 1.0]
+   :alt: [0.0, 1.0], [0.0, 1.0], [0.0, 1.0], [1.0, 0.0], [0.0, 1.0], [0.0, 1.0], [0.0, 1.0], [0.0, 1.0], [0.0, 1.0], [0.0, 1.0]
    :srcset: /how_to/work_with_microtvm/images/sphx_glr_micro_train_001.png
    :class: sphx-glr-single-img
 
@@ -318,8 +318,8 @@ objects to other stuff? We can display some examples from our datasets using ``m
 
  .. code-block:: none
 
-    /tmp/tmpk1esgsax/images/target contains 8144 images
-    /tmp/tmpk1esgsax/images/random contains 5000 images
+    /tmp/tmp1fwqht9d/images/target contains 8144 images
+    /tmp/tmp1fwqht9d/images/random contains 5000 images
 
 
 
@@ -494,13 +494,13 @@ the time on our validation set).
  .. code-block:: none
 
     Epoch 1/3
-    328/328 - 47s - loss: 0.2310 - accuracy: 0.9208 - val_loss: 0.1758 - val_accuracy: 0.9411 - 47s/epoch - 145ms/step
+    328/328 - 46s - loss: 0.2092 - accuracy: 0.9272 - val_loss: 0.1058 - val_accuracy: 0.9664 - 46s/epoch - 142ms/step
     Epoch 2/3
-    328/328 - 44s - loss: 0.1006 - accuracy: 0.9616 - val_loss: 0.1315 - val_accuracy: 0.9585 - 44s/epoch - 133ms/step
+    328/328 - 43s - loss: 0.0983 - accuracy: 0.9609 - val_loss: 0.1074 - val_accuracy: 0.9683 - 43s/epoch - 131ms/step
     Epoch 3/3
-    328/328 - 43s - loss: 0.0624 - accuracy: 0.9766 - val_loss: 0.1429 - val_accuracy: 0.9588 - 43s/epoch - 132ms/step
+    328/328 - 43s - loss: 0.0669 - accuracy: 0.9746 - val_loss: 0.1026 - val_accuracy: 0.9630 - 43s/epoch - 131ms/step
 
-    <keras.callbacks.History object at 0x7ff9106419d0>
+    <keras.callbacks.History object at 0x7f2b89178b10>
 
 
 
@@ -857,7 +857,7 @@ Arduino tutorial for how to do that `on GitHub <https://github.com/guberti/tvm-a
 
 .. rst-class:: sphx-glr-timing
 
-   **Total running time of the script:** ( 4 minutes  25.812 seconds)
+   **Total running time of the script:** ( 4 minutes  21.961 seconds)
 
 
 .. _sphx_glr_download_how_to_work_with_microtvm_micro_train.py:
diff --git a/docs/_sources/how_to/work_with_microtvm/sg_execution_times.rst.txt b/docs/_sources/how_to/work_with_microtvm/sg_execution_times.rst.txt
index 3557556277..286a2c0c8c 100644
--- a/docs/_sources/how_to/work_with_microtvm/sg_execution_times.rst.txt
+++ b/docs/_sources/how_to/work_with_microtvm/sg_execution_times.rst.txt
@@ -5,18 +5,18 @@
 
 Computation times
 =================
-**06:37.338** total execution time for **how_to_work_with_microtvm** files:
+**06:28.725** total execution time for **how_to_work_with_microtvm** files:
 
 +---------------------------------------------------------------------------------------------+-----------+--------+
-| :ref:`sphx_glr_how_to_work_with_microtvm_micro_train.py` (``micro_train.py``)               | 04:25.812 | 0.0 MB |
+| :ref:`sphx_glr_how_to_work_with_microtvm_micro_train.py` (``micro_train.py``)               | 04:21.961 | 0.0 MB |
 +---------------------------------------------------------------------------------------------+-----------+--------+
-| :ref:`sphx_glr_how_to_work_with_microtvm_micro_pytorch.py` (``micro_pytorch.py``)           | 01:05.814 | 0.0 MB |
+| :ref:`sphx_glr_how_to_work_with_microtvm_micro_pytorch.py` (``micro_pytorch.py``)           | 01:03.361 | 0.0 MB |
 +---------------------------------------------------------------------------------------------+-----------+--------+
-| :ref:`sphx_glr_how_to_work_with_microtvm_micro_autotune.py` (``micro_autotune.py``)         | 00:53.640 | 0.0 MB |
+| :ref:`sphx_glr_how_to_work_with_microtvm_micro_autotune.py` (``micro_autotune.py``)         | 00:51.580 | 0.0 MB |
 +---------------------------------------------------------------------------------------------+-----------+--------+
-| :ref:`sphx_glr_how_to_work_with_microtvm_micro_aot.py` (``micro_aot.py``)                   | 00:08.077 | 0.0 MB |
+| :ref:`sphx_glr_how_to_work_with_microtvm_micro_aot.py` (``micro_aot.py``)                   | 00:07.975 | 0.0 MB |
 +---------------------------------------------------------------------------------------------+-----------+--------+
-| :ref:`sphx_glr_how_to_work_with_microtvm_micro_tflite.py` (``micro_tflite.py``)             | 00:03.992 | 0.0 MB |
+| :ref:`sphx_glr_how_to_work_with_microtvm_micro_tflite.py` (``micro_tflite.py``)             | 00:03.846 | 0.0 MB |
 +---------------------------------------------------------------------------------------------+-----------+--------+
 | :ref:`sphx_glr_how_to_work_with_microtvm_micro_reference_vm.py` (``micro_reference_vm.py``) | 00:00.001 | 0.0 MB |
 +---------------------------------------------------------------------------------------------+-----------+--------+
diff --git a/docs/_sources/how_to/work_with_relay/sg_execution_times.rst.txt b/docs/_sources/how_to/work_with_relay/sg_execution_times.rst.txt
index 5e7e7cc270..c95746e06f 100644
--- a/docs/_sources/how_to/work_with_relay/sg_execution_times.rst.txt
+++ b/docs/_sources/how_to/work_with_relay/sg_execution_times.rst.txt
@@ -5,14 +5,14 @@
 
 Computation times
 =================
-**00:45.335** total execution time for **how_to_work_with_relay** files:
+**00:44.444** total execution time for **how_to_work_with_relay** files:
 
 +----------------------------------------------------------------------------------------------------+-----------+--------+
-| :ref:`sphx_glr_how_to_work_with_relay_using_pipeline_executor.py` (``using_pipeline_executor.py``) | 00:33.475 | 0.0 MB |
+| :ref:`sphx_glr_how_to_work_with_relay_using_pipeline_executor.py` (``using_pipeline_executor.py``) | 00:32.558 | 0.0 MB |
 +----------------------------------------------------------------------------------------------------+-----------+--------+
-| :ref:`sphx_glr_how_to_work_with_relay_using_external_lib.py` (``using_external_lib.py``)           | 00:10.314 | 0.0 MB |
+| :ref:`sphx_glr_how_to_work_with_relay_using_external_lib.py` (``using_external_lib.py``)           | 00:10.293 | 0.0 MB |
 +----------------------------------------------------------------------------------------------------+-----------+--------+
-| :ref:`sphx_glr_how_to_work_with_relay_build_gcn.py` (``build_gcn.py``)                             | 00:01.538 | 0.0 MB |
+| :ref:`sphx_glr_how_to_work_with_relay_build_gcn.py` (``build_gcn.py``)                             | 00:01.586 | 0.0 MB |
 +----------------------------------------------------------------------------------------------------+-----------+--------+
 | :ref:`sphx_glr_how_to_work_with_relay_using_relay_viz.py` (``using_relay_viz.py``)                 | 00:00.007 | 0.0 MB |
 +----------------------------------------------------------------------------------------------------+-----------+--------+
diff --git a/docs/_sources/how_to/work_with_schedules/intrin_math.rst.txt b/docs/_sources/how_to/work_with_schedules/intrin_math.rst.txt
index 19b2e3cb94..fdfe7ef310 100644
--- a/docs/_sources/how_to/work_with_schedules/intrin_math.rst.txt
+++ b/docs/_sources/how_to/work_with_schedules/intrin_math.rst.txt
@@ -265,7 +265,7 @@ The following example customizes CUDA lowering rule for :code:`exp`.
  .. code-block:: none
 
 
-    <function my_cuda_math_rule at 0x7ff985e1c5f0>
+    <function my_cuda_math_rule at 0x7f2bd76137a0>
 
 
 
diff --git a/docs/_sources/how_to/work_with_schedules/sg_execution_times.rst.txt b/docs/_sources/how_to/work_with_schedules/sg_execution_times.rst.txt
index 3c0242e469..2b26872514 100644
--- a/docs/_sources/how_to/work_with_schedules/sg_execution_times.rst.txt
+++ b/docs/_sources/how_to/work_with_schedules/sg_execution_times.rst.txt
@@ -5,20 +5,20 @@
 
 Computation times
 =================
-**00:07.474** total execution time for **how_to_work_with_schedules** files:
+**00:07.328** total execution time for **how_to_work_with_schedules** files:
 
 +------------------------------------------------------------------------------------------------+-----------+--------+
-| :ref:`sphx_glr_how_to_work_with_schedules_intrin_math.py` (``intrin_math.py``)                 | 00:04.850 | 0.0 MB |
+| :ref:`sphx_glr_how_to_work_with_schedules_intrin_math.py` (``intrin_math.py``)                 | 00:04.772 | 0.0 MB |
 +------------------------------------------------------------------------------------------------+-----------+--------+
-| :ref:`sphx_glr_how_to_work_with_schedules_tensorize.py` (``tensorize.py``)                     | 00:01.194 | 0.0 MB |
+| :ref:`sphx_glr_how_to_work_with_schedules_tensorize.py` (``tensorize.py``)                     | 00:01.193 | 0.0 MB |
 +------------------------------------------------------------------------------------------------+-----------+--------+
-| :ref:`sphx_glr_how_to_work_with_schedules_reduction.py` (``reduction.py``)                     | 00:00.618 | 0.0 MB |
+| :ref:`sphx_glr_how_to_work_with_schedules_reduction.py` (``reduction.py``)                     | 00:00.584 | 0.0 MB |
 +------------------------------------------------------------------------------------------------+-----------+--------+
-| :ref:`sphx_glr_how_to_work_with_schedules_scan.py` (``scan.py``)                               | 00:00.594 | 0.0 MB |
+| :ref:`sphx_glr_how_to_work_with_schedules_scan.py` (``scan.py``)                               | 00:00.561 | 0.0 MB |
 +------------------------------------------------------------------------------------------------+-----------+--------+
-| :ref:`sphx_glr_how_to_work_with_schedules_extern_op.py` (``extern_op.py``)                     | 00:00.115 | 0.0 MB |
+| :ref:`sphx_glr_how_to_work_with_schedules_extern_op.py` (``extern_op.py``)                     | 00:00.117 | 0.0 MB |
 +------------------------------------------------------------------------------------------------+-----------+--------+
-| :ref:`sphx_glr_how_to_work_with_schedules_schedule_primitives.py` (``schedule_primitives.py``) | 00:00.051 | 0.0 MB |
+| :ref:`sphx_glr_how_to_work_with_schedules_schedule_primitives.py` (``schedule_primitives.py``) | 00:00.049 | 0.0 MB |
 +------------------------------------------------------------------------------------------------+-----------+--------+
 | :ref:`sphx_glr_how_to_work_with_schedules_tedd.py` (``tedd.py``)                               | 00:00.029 | 0.0 MB |
 +------------------------------------------------------------------------------------------------+-----------+--------+
diff --git a/docs/_sources/how_to/work_with_schedules/tensorize.rst.txt b/docs/_sources/how_to/work_with_schedules/tensorize.rst.txt
index b6684fe3fc..57cbe91ddb 100644
--- a/docs/_sources/how_to/work_with_schedules/tensorize.rst.txt
+++ b/docs/_sources/how_to/work_with_schedules/tensorize.rst.txt
@@ -347,7 +347,7 @@ The importing needs to happen before the tensorized GEMV being executed.
                  B: Buffer(B_2: Pointer(float32), float32, [512, 64], []),
                  C: Buffer(C_2: Pointer(float32), float32, [1024, 512], [])}
       buffer_map = {A_1: A, B_1: B, C_1: C} {
-      attr [IterVar(i: int32, (nullptr), "DataPar", "")] "pragma_import_llvm" = "; ModuleID = '/tmp/tmpyjqxigcy/input0.cc'\nsource_filename = \"/tmp/tmpyjqxigcy/input0.cc\"\ntarget datalayout = \"e-m:e-i64:64-f80:128-n8:16:32:64-S128\"\ntarget triple = \"x86_64-pc-linux-gnu\"\n\n; Function Attrs: noinline nounwind optnone uwtable\ndefine dso_local i32 @gemv_update(float*, float*, float*, i32, i32, i32) #0 {\n  %7 = alloca float*, align 8\n  %8 = alloca float*, align 8\n  %9 = alloca floa [...]
+      attr [IterVar(i: int32, (nullptr), "DataPar", "")] "pragma_import_llvm" = "; ModuleID = '/tmp/tmp_p3by85r/input0.cc'\nsource_filename = \"/tmp/tmp_p3by85r/input0.cc\"\ntarget datalayout = \"e-m:e-i64:64-f80:128-n8:16:32:64-S128\"\ntarget triple = \"x86_64-pc-linux-gnu\"\n\n; Function Attrs: noinline nounwind optnone uwtable\ndefine dso_local i32 @gemv_update(float*, float*, float*, i32, i32, i32) #0 {\n  %7 = alloca float*, align 8\n  %8 = alloca float*, align 8\n  %9 = alloca floa [...]
       for (i, 0, 1024) {
         for (j.outer: int32, 0, 32) {
           @tir.call_extern("gemv_update", @tir.tvm_access_ptr(@tir.type_annotation(, dtype=float32), C_2, ((i*512) + (j.outer*16)), 16, 2, dtype=handle), @tir.tvm_access_ptr(@tir.type_annotation(, dtype=float32), A_2, (i*64), 64, 1, dtype=handle), @tir.tvm_access_ptr(@tir.type_annotation(, dtype=float32), B_2, (j.outer*1024), 1024, 1, dtype=handle), 16, 64, 64, dtype=int32)
diff --git a/docs/_sources/topic/vta/tutorials/autotvm/sg_execution_times.rst.txt b/docs/_sources/topic/vta/tutorials/autotvm/sg_execution_times.rst.txt
index 0e26a51041..15bc52c5a9 100644
--- a/docs/_sources/topic/vta/tutorials/autotvm/sg_execution_times.rst.txt
+++ b/docs/_sources/topic/vta/tutorials/autotvm/sg_execution_times.rst.txt
@@ -5,10 +5,10 @@
 
 Computation times
 =================
-**00:26.954** total execution time for **topic_vta_tutorials_autotvm** files:
+**00:26.421** total execution time for **topic_vta_tutorials_autotvm** files:
 
 +---------------------------------------------------------------------------------------+-----------+--------+
-| :ref:`sphx_glr_topic_vta_tutorials_autotvm_tune_relay_vta.py` (``tune_relay_vta.py``) | 00:26.948 | 0.0 MB |
+| :ref:`sphx_glr_topic_vta_tutorials_autotvm_tune_relay_vta.py` (``tune_relay_vta.py``) | 00:26.415 | 0.0 MB |
 +---------------------------------------------------------------------------------------+-----------+--------+
 | :ref:`sphx_glr_topic_vta_tutorials_autotvm_tune_alu_vta.py` (``tune_alu_vta.py``)     | 00:00.006 | 0.0 MB |
 +---------------------------------------------------------------------------------------+-----------+--------+
diff --git a/docs/_sources/topic/vta/tutorials/frontend/deploy_classification.rst.txt b/docs/_sources/topic/vta/tutorials/frontend/deploy_classification.rst.txt
index 17ab36bf11..66a4ec453f 100644
--- a/docs/_sources/topic/vta/tutorials/frontend/deploy_classification.rst.txt
+++ b/docs/_sources/topic/vta/tutorials/frontend/deploy_classification.rst.txt
@@ -293,7 +293,7 @@ The compilation steps are:
       DeprecationWarning,
     /workspace/vta/tutorials/frontend/deploy_classification.py:213: DeprecationWarning: legacy graph executor behavior of producing json / lib / params will be removed in the next release. Please see documents of tvm.contrib.graph_executor.GraphModule for the  new recommended usage.
       relay_prog, target=tvm.target.Target(target, host=env.target_host), params=params
-    resnet18_v1 inference graph built in 30.13s!
+    resnet18_v1 inference graph built in 28.69s!
 
 
 
diff --git a/docs/_sources/topic/vta/tutorials/frontend/deploy_detection.rst.txt b/docs/_sources/topic/vta/tutorials/frontend/deploy_detection.rst.txt
index 2522d95550..ffead96a7f 100644
--- a/docs/_sources/topic/vta/tutorials/frontend/deploy_detection.rst.txt
+++ b/docs/_sources/topic/vta/tutorials/frontend/deploy_detection.rst.txt
@@ -337,7 +337,7 @@ The compilation steps are:
 
     /workspace/python/tvm/relay/build_module.py:348: DeprecationWarning: Please use input parameter mod (tvm.IRModule) instead of deprecated parameter mod (tvm.relay.function.Function)
       DeprecationWarning,
-    yolov3-tiny inference graph built in 20.12s!
+    yolov3-tiny inference graph built in 19.51s!
 
 
 
diff --git a/docs/_sources/topic/vta/tutorials/frontend/sg_execution_times.rst.txt b/docs/_sources/topic/vta/tutorials/frontend/sg_execution_times.rst.txt
index d4ae5a8baa..052b90000f 100644
--- a/docs/_sources/topic/vta/tutorials/frontend/sg_execution_times.rst.txt
+++ b/docs/_sources/topic/vta/tutorials/frontend/sg_execution_times.rst.txt
@@ -5,10 +5,10 @@
 
 Computation times
 =================
-**01:34.127** total execution time for **topic_vta_tutorials_frontend** files:
+**01:31.917** total execution time for **topic_vta_tutorials_frontend** files:
 
 +------------------------------------------------------------------------------------------------------+-----------+--------+
-| :ref:`sphx_glr_topic_vta_tutorials_frontend_deploy_classification.py` (``deploy_classification.py``) | 00:47.091 | 0.0 MB |
+| :ref:`sphx_glr_topic_vta_tutorials_frontend_deploy_detection.py` (``deploy_detection.py``)           | 00:46.415 | 0.0 MB |
 +------------------------------------------------------------------------------------------------------+-----------+--------+
-| :ref:`sphx_glr_topic_vta_tutorials_frontend_deploy_detection.py` (``deploy_detection.py``)           | 00:47.035 | 0.0 MB |
+| :ref:`sphx_glr_topic_vta_tutorials_frontend_deploy_classification.py` (``deploy_classification.py``) | 00:45.502 | 0.0 MB |
 +------------------------------------------------------------------------------------------------------+-----------+--------+
diff --git a/docs/_sources/topic/vta/tutorials/optimize/sg_execution_times.rst.txt b/docs/_sources/topic/vta/tutorials/optimize/sg_execution_times.rst.txt
index ce37a0f0f6..d80b986df0 100644
--- a/docs/_sources/topic/vta/tutorials/optimize/sg_execution_times.rst.txt
+++ b/docs/_sources/topic/vta/tutorials/optimize/sg_execution_times.rst.txt
@@ -5,10 +5,10 @@
 
 Computation times
 =================
-**00:03.267** total execution time for **topic_vta_tutorials_optimize** files:
+**00:03.156** total execution time for **topic_vta_tutorials_optimize** files:
 
 +--------------------------------------------------------------------------------------------------+-----------+--------+
-| :ref:`sphx_glr_topic_vta_tutorials_optimize_convolution_opt.py` (``convolution_opt.py``)         | 00:02.768 | 0.0 MB |
+| :ref:`sphx_glr_topic_vta_tutorials_optimize_convolution_opt.py` (``convolution_opt.py``)         | 00:02.691 | 0.0 MB |
 +--------------------------------------------------------------------------------------------------+-----------+--------+
-| :ref:`sphx_glr_topic_vta_tutorials_optimize_matrix_multiply_opt.py` (``matrix_multiply_opt.py``) | 00:00.499 | 0.0 MB |
+| :ref:`sphx_glr_topic_vta_tutorials_optimize_matrix_multiply_opt.py` (``matrix_multiply_opt.py``) | 00:00.465 | 0.0 MB |
 +--------------------------------------------------------------------------------------------------+-----------+--------+
diff --git a/docs/_sources/topic/vta/tutorials/sg_execution_times.rst.txt b/docs/_sources/topic/vta/tutorials/sg_execution_times.rst.txt
index c8ac76ef46..2ab52f1245 100644
--- a/docs/_sources/topic/vta/tutorials/sg_execution_times.rst.txt
+++ b/docs/_sources/topic/vta/tutorials/sg_execution_times.rst.txt
@@ -5,10 +5,10 @@
 
 Computation times
 =================
-**00:00.873** total execution time for **topic_vta_tutorials** files:
+**00:00.832** total execution time for **topic_vta_tutorials** files:
 
 +---------------------------------------------------------------------------------+-----------+--------+
-| :ref:`sphx_glr_topic_vta_tutorials_matrix_multiply.py` (``matrix_multiply.py``) | 00:00.452 | 0.0 MB |
+| :ref:`sphx_glr_topic_vta_tutorials_matrix_multiply.py` (``matrix_multiply.py``) | 00:00.449 | 0.0 MB |
 +---------------------------------------------------------------------------------+-----------+--------+
-| :ref:`sphx_glr_topic_vta_tutorials_vta_get_started.py` (``vta_get_started.py``) | 00:00.420 | 0.0 MB |
+| :ref:`sphx_glr_topic_vta_tutorials_vta_get_started.py` (``vta_get_started.py``) | 00:00.382 | 0.0 MB |
 +---------------------------------------------------------------------------------+-----------+--------+
diff --git a/docs/_sources/tutorial/auto_scheduler_matmul_x86.rst.txt b/docs/_sources/tutorial/auto_scheduler_matmul_x86.rst.txt
index d4ec5f646e..9295db4bd7 100644
--- a/docs/_sources/tutorial/auto_scheduler_matmul_x86.rst.txt
+++ b/docs/_sources/tutorial/auto_scheduler_matmul_x86.rst.txt
@@ -211,12 +211,12 @@ trials, we can load the best schedule from the log file and apply it.
 
  .. code-block:: none
 
-
     *E
 
 
 
 
+
 .. GENERATED FROM PYTHON SOURCE LINES 138-144
 
 Inspecting the Optimized Schedule
@@ -336,7 +336,7 @@ We build the binary and check its correctness and performance.
 
  .. code-block:: none
 
-    Execution time of this operator: 95.175 ms
+    Execution time of this operator: 95.293 ms
 
 
 
@@ -454,7 +454,7 @@ operations.
 
 .. rst-class:: sphx-glr-timing
 
-   **Total running time of the script:** ( 1 minutes  30.687 seconds)
+   **Total running time of the script:** ( 1 minutes  34.758 seconds)
 
 
 .. _sphx_glr_download_tutorial_auto_scheduler_matmul_x86.py:
diff --git a/docs/_sources/tutorial/autotvm_matmul_x86.rst.txt b/docs/_sources/tutorial/autotvm_matmul_x86.rst.txt
index 46f8586bd8..33e62c2327 100644
--- a/docs/_sources/tutorial/autotvm_matmul_x86.rst.txt
+++ b/docs/_sources/tutorial/autotvm_matmul_x86.rst.txt
@@ -454,16 +454,16 @@ reduce variance, we take 5 measurements and average them.
     waiting for device...
     device available
     Get devices for measurement successfully!
-    No: 1   GFLOPS: 6.47/6.47       result: MeasureResult(costs=(0.041515294599999995,), error_no=MeasureErrorNo.NO_ERROR, all_cost=0.8913285732269287, timestamp=1673391404.9843054)       [('tile_y', [-1, 1]), ('tile_x', [-1, 32])],None,50
-    No: 2   GFLOPS: 2.98/6.47       result: MeasureResult(costs=(0.090220976,), error_no=MeasureErrorNo.NO_ERROR, all_cost=1.700326919555664, timestamp=1673391407.456227)  [('tile_y', [-1, 2]), ('tile_x', [-1, 16])],None,41
-    No: 3   GFLOPS: 8.93/8.93       result: MeasureResult(costs=(0.030065189800000003,), error_no=MeasureErrorNo.NO_ERROR, all_cost=0.7250089645385742, timestamp=1673391408.9627564)       [('tile_y', [-1, 4]), ('tile_x', [-1, 32])],None,52
-    No: 4   GFLOPS: 0.48/8.93       result: MeasureResult(costs=(0.553561442,), error_no=MeasureErrorNo.NO_ERROR, all_cost=9.072837591171265, timestamp=1673391418.0785744) [('tile_y', [-1, 512]), ('tile_x', [-1, 1])],None,9
-    No: 5   GFLOPS: 2.77/8.93       result: MeasureResult(costs=(0.0969489894,), error_no=MeasureErrorNo.NO_ERROR, all_cost=1.786363124847412, timestamp=1673391419.9909449)        [('tile_y', [-1, 4]), ('tile_x', [-1, 4])],None,22
-    No: 6   GFLOPS: 1.59/8.93       result: MeasureResult(costs=(0.16925801759999998,), error_no=MeasureErrorNo.NO_ERROR, all_cost=2.9514756202697754, timestamp=1673391422.9553232)        [('tile_y', [-1, 64]), ('tile_x', [-1, 4])],None,26
-    No: 7   GFLOPS: 12.62/12.62     result: MeasureResult(costs=(0.0212633308,), error_no=MeasureErrorNo.NO_ERROR, all_cost=0.5868525505065918, timestamp=1673391424.3278058)       [('tile_y', [-1, 256]), ('tile_x', [-1, 256])],None,88
-    No: 8   GFLOPS: 1.59/12.62      result: MeasureResult(costs=(0.16891188699999998,), error_no=MeasureErrorNo.NO_ERROR, all_cost=2.933110475540161, timestamp=1673391427.2752721) [('tile_y', [-1, 32]), ('tile_x', [-1, 4])],None,25
-    No: 9   GFLOPS: 10.08/12.62     result: MeasureResult(costs=(0.026640367600000003,), error_no=MeasureErrorNo.NO_ERROR, all_cost=0.6851820945739746, timestamp=1673391428.0736556)       [('tile_y', [-1, 512]), ('tile_x', [-1, 64])],None,69
-    No: 10  GFLOPS: 4.39/12.62      result: MeasureResult(costs=(0.061145176,), error_no=MeasureErrorNo.NO_ERROR, all_cost=1.1997015476226807, timestamp=1673391429.3030593)        [('tile_y', [-1, 8]), ('tile_x', [-1, 16])],None,43
+    No: 1   GFLOPS: 1.55/1.55       result: MeasureResult(costs=(0.1727136912,), error_no=MeasureErrorNo.NO_ERROR, all_cost=3.0090889930725098, timestamp=1673399082.22959) [('tile_y', [-1, 64]), ('tile_x', [-1, 4])],None,26
+    No: 2   GFLOPS: 2.06/2.06       result: MeasureResult(costs=(0.1304848428,), error_no=MeasureErrorNo.NO_ERROR, all_cost=2.342974901199341, timestamp=1673399084.5735672)        [('tile_y', [-1, 128]), ('tile_x', [-1, 4])],None,27
+    No: 3   GFLOPS: 1.57/2.06       result: MeasureResult(costs=(0.1712796658,), error_no=MeasureErrorNo.NO_ERROR, all_cost=2.9583277702331543, timestamp=1673399088.3322468)       [('tile_y', [-1, 8]), ('tile_x', [-1, 1])],None,3
+    No: 4   GFLOPS: 1.31/2.06       result: MeasureResult(costs=(0.2052259786,), error_no=MeasureErrorNo.NO_ERROR, all_cost=3.5049915313720703, timestamp=1673399092.6279268)       [('tile_y', [-1, 2]), ('tile_x', [-1, 1])],None,1
+    No: 5   GFLOPS: 10.67/10.67     result: MeasureResult(costs=(0.0251632978,), error_no=MeasureErrorNo.NO_ERROR, all_cost=0.6383941173553467, timestamp=1673399094.1546712)       [('tile_y', [-1, 512]), ('tile_x', [-1, 256])],None,89
+    No: 6   GFLOPS: 12.33/12.33     result: MeasureResult(costs=(0.0217639196,), error_no=MeasureErrorNo.NO_ERROR, all_cost=0.5869083404541016, timestamp=1673399094.7599728)       [('tile_y', [-1, 256]), ('tile_x', [-1, 256])],None,88
+    No: 7   GFLOPS: 1.63/12.33      result: MeasureResult(costs=(0.1650163328,), error_no=MeasureErrorNo.NO_ERROR, all_cost=2.8642807006835938, timestamp=1673399097.6541824)       [('tile_y', [-1, 1]), ('tile_x', [-1, 4])],None,20
+    No: 8   GFLOPS: 1.64/12.33      result: MeasureResult(costs=(0.1634693014,), error_no=MeasureErrorNo.NO_ERROR, all_cost=2.83781099319458, timestamp=1673399100.5211616) [('tile_y', [-1, 2]), ('tile_x', [-1, 2])],None,11
+    No: 9   GFLOPS: 13.71/13.71     result: MeasureResult(costs=(0.019579842,), error_no=MeasureErrorNo.NO_ERROR, all_cost=0.5451052188873291, timestamp=1673399101.2340999)        [('tile_y', [-1, 16]), ('tile_x', [-1, 512])],None,94
+    No: 10  GFLOPS: 12.40/13.71     result: MeasureResult(costs=(0.0216520308,), error_no=MeasureErrorNo.NO_ERROR, all_cost=0.6318655014038086, timestamp=1673399101.8354917)       [('tile_y', [-1, 1]), ('tile_x', [-1, 64])],None,60
 
 
 
diff --git a/docs/_sources/tutorial/autotvm_relay_x86.rst.txt b/docs/_sources/tutorial/autotvm_relay_x86.rst.txt
index 63eca044a1..e02a4b08b3 100644
--- a/docs/_sources/tutorial/autotvm_relay_x86.rst.txt
+++ b/docs/_sources/tutorial/autotvm_relay_x86.rst.txt
@@ -324,7 +324,7 @@ standard deviation.
 
  .. code-block:: none
 
-    {'mean': 519.0110534800078, 'median': 518.6799677499948, 'std': 1.8895944523185717}
+    {'mean': 517.1232215099917, 'median': 517.1692200999587, 'std': 2.4466269662533757}
 
 
 
@@ -558,31 +558,30 @@ the tuning data to.
 
  .. code-block:: none
 
-
    [Task  1/25]  Current/Best:    0.00/   0.00 GFLOPS | Progress: (0/20) | 0.00 s
    [Task  1/25]  Current/Best:    8.56/  11.51 GFLOPS | Progress: (4/20) | 8.42 s
    [Task  1/25]  Current/Best:   23.45/  23.45 GFLOPS | Progress: (8/20) | 11.77 s
    [Task  1/25]  Current/Best:    7.17/  23.45 GFLOPS | Progress: (12/20) | 14.62 s
    [Task  1/25]  Current/Best:   11.05/  23.45 GFLOPS | Progress: (16/20) | 19.93 s
    [Task  1/25]  Current/Best:    1.92/  23.45 GFLOPS | Progress: (20/20) | 24.69 s Done.
-
    [Task  2/25]  Current/Best:    0.00/   0.00 GFLOPS | Progress: (0/20) | 0.00 s
    [Task  2/25]  Current/Best:    5.19/  18.70 GFLOPS | Progress: (4/20) | 3.64 s
    [Task  2/25]  Current/Best:   12.61/  18.70 GFLOPS | Progress: (8/20) | 6.72 s
    [Task  2/25]  Current/Best:    9.56/  20.84 GFLOPS | Progress: (12/20) | 8.15 s
    [Task  2/25]  Current/Best:   11.73/  20.84 GFLOPS | Progress: (16/20) | 10.41 s
    [Task  2/25]  Current/Best:    9.74/  21.97 GFLOPS | Progress: (20/20) | 12.35 s Done.
-
    [Task  3/25]  Current/Best:    0.00/   0.00 GFLOPS | Progress: (0/20) | 0.00 s
    [Task  3/25]  Current/Best:    1.62/  23.50 GFLOPS | Progress: (4/20) | 5.70 s
    [Task  3/25]  Current/Best:   10.42/  23.50 GFLOPS | Progress: (8/20) | 8.61 s
    [Task  3/25]  Current/Best:   12.71/  23.63 GFLOPS | Progress: (12/20) | 11.35 s
    [Task  3/25]  Current/Best:   12.00/  23.63 GFLOPS | Progress: (16/20) | 14.30 s
    [Task  3/25]  Current/Best:   15.71/  23.63 GFLOPS | Progress: (20/20) | 16.90 s Done.
-
    [Task  4/25]  Current/Best:    0.00/   0.00 GFLOPS | Progress: (0/20) | 0.00 s
    [Task  4/25]  Current/Best:   10.35/  16.16 GFLOPS | Progress: (4/20) | 3.75 s
    [Task  4/25]  Current/Best:   10.63/  19.41 GFLOPS | Progress: (8/20) | 9.85 s
    [Task  4/25]  Current/Best:   16.12/  19.41 GFLOPS | Progress: (12/20) | 12.24 s
    [Task  4/25]  Current/Best:    9.01/  19.41 GFLOPS | Progress: (16/20) | 21.34 s
    [Task  4/25]  Current/Best:   20.94/  20.94 GFLOPS | Progress: (20/20) | 23.46 s Done.
-
    [Task  5/25]  Current/Best:    0.00/   0.00 GFLOPS | Progress: (0/20) | 0.00 s
    [Task  5/25]  Current/Best:   17.88/  17.88 GFLOPS | Progress: (4/20) | 4.03 s
    [Task  5/25]  Current/Best:   14.89/  18.22 GFLOPS | Progress: (8/20) | 6.29 s
    [Task  5/25]  Current/Best:   13.42/  18.22 GFLOPS | Progress: (12/20) | 8.29 s
    [Task  5/25]  Current/Best:   14.67/  18.22 GFLOPS | Progress: (16/20) | 10.76 s
    [Task  5/25]  Current/Best:   16.29/  21.41 GFLOPS | Progress: (20/20) | 12.34 s Done.
-
    [Task  6/25]  Current/Best:    0.00/   0.00 GFLOPS | Progress: (0/20) | 0.00 s
    [Task  6/25]  Current/Best:   13.10/  13.10 GFLOPS | Progress: (4/20) | 5.16 s
    [Task  6/25]  Current/Best:   14.75/  14.75 GFLOPS | Progress: (8/20) | 9.16 s
    [Task  6/25]  Current/Best:    8.26/  18.95 GFLOPS | Progress: (12/20) | 11.62 s
    [Task  6/25]  Current/Best:    5.95/  18.95 GFLOPS | Progress: (16/20) | 14.62 s
    [Task  6/25]  Current/Best:   11.34/  23.11 GFLOPS | Progress: (20/20) | 16.99 s Done.
-
    [Task  7/25]  Current/Best:    0.00/   0.00 GFLOPS | Progress: (0/20) | 0.00 s
    [Task  7/25]  Current/Best:   11.46/  20.36 GFLOPS | Progress: (4/20) | 3.80 s
    [Task  7/25]  Current/Best:   22.66/  22.66 GFLOPS | Progress: (8/20) | 6.02 s
    [Task  7/25]  Current/Best:   17.71/  22.66 GFLOPS | Progress: (12/20) | 8.74 s
    [Task  7/25]  Current/Best:   16.55/  22.66 GFLOPS | Progress: (16/20) | 10.90 s
    [Task  7/25]  Current/Best:    7.90/  22.66 GFLOPS | Progress: (20/20) | 14.07 s Done.
-
    [Task  8/25]  Current/Best:    0.00/   0.00 GFLOPS | Progress: (0/20) | 0.00 s
    [Task  8/25]  Current/Best:   13.82/  13.82 GFLOPS | Progress: (4/20) | 7.47 s
    [Task  8/25]  Current/Best:    9.69/  13.82 GFLOPS | Progress: (8/20) | 14.71 s
    [Task  8/25]  Current/Best:    7.60/  17.60 GFLOPS | Progress: (12/20) | 21.41 s
    [Task  8/25]  Current/Best:    3.73/  17.60 GFLOPS | Progress: (16/20) | 25.89 s
    [Task  8/25]  Current/Best:   18.54/  18.54 GFLOPS | Progress: (20/20) | 37.37 s
    [Task  9/25]  Current/Best:    0.00/   0.00 GFLOPS | Progress: (0/20) | 0.00 s
    [Task  9/25]  Current/Best:   22.94/  22.94 GFLOPS | Progress: (4/20) | 4.55 s
    [Task  9/25]  Current/Best:   20.51/  22.94 GFLOPS | Progress: (8/20) | 10.28 s
    [Task  9/25]  Current/Best:   13.92/  22.94 GFLOPS | Progress: (12/20) | 14.13 s
    [Task  9/25]  Current/Best:    5.36/  22.94 GFLOPS | Progress: (16/20) | 16.07 s
    [Task  9/25]  Current/Best:   17.28/  22.94 GFLOPS | Progress: (20/
 20) | 17.80 s Done.
-
    [Task 10/25]  Current/Best:    0.00/   0.00 GFLOPS | Progress: (0/20) | 0.00 s
    [Task 10/25]  Current/Best:   13.02/  18.32 GFLOPS | Progress: (4/20) | 4.05 s
    [Task 10/25]  Current/Best:   19.92/  20.92 GFLOPS | Progress: (8/20) | 6.41 s
    [Task 10/25]  Current/Best:   12.10/  20.92 GFLOPS | Progress: (12/20) | 8.17 s
    [Task 10/25]  Current/Best:    3.05/  20.92 GFLOPS | Progress: (16/20) | 10.20 s
    [Task 10/25]  Current/Best:    9.41/  20.92 GFLOPS | Progress: (20/20) | 11.94 s Done.
-
    [Task 11/25]  Current/Best:    0.00/   0.00 GFLOPS | Progress: (0/20) | 0.00 s
    [Task 11/25]  Current/Best:   11.24/  11.66 GFLOPS | Progress: (4/20) | 5.45 s
    [Task 11/25]  Current/Best:    8.75/  17.40 GFLOPS | Progress: (8/20) | 8.25 s
    [Task 11/25]  Current/Best:   15.10/  17.40 GFLOPS | Progress: (12/20) | 10.94 s
    [Task 11/25]  Current/Best:   21.68/  21.68 GFLOPS | Progress: (16/20) | 14.79 s
    [Task 11/25]  Current/Best:    6.23/  21.68 GFLOPS | Progress: (20/20) | 17.20 s Done.
-
    [Task 12/25]  Current/Best:    0.00/   0.00 GFLOPS | Progress: (0/20) | 0.00 s
    [Task 12/25]  Current/Best:   18.35/  18.35 GFLOPS | Progress: (4/20) | 4.62 s
    [Task 12/25]  Current/Best:   11.69/  18.35 GFLOPS | Progress: (8/20) | 11.16 s
    [Task 12/25]  Current/Best:   15.20/  18.35 GFLOPS | Progress: (12/20) | 13.35 s
    [Task 12/25]  Current/Best:   12.01/  18.35 GFLOPS | Progress: (16/20) | 17.28 s
    [Task 12/25]  Current/Best:    8.18/  18.83 GFLOPS | Progress: (20/20) | 19.32 s Done.
-
    [Task 13/25]  Current/Best:    0.00/   0.00 GFLOPS | Progress: (0/20) | 0.00 s
    [Task 13/25]  Current/Best:   17.28/  17.28 GFLOPS | Progress: (4/20) | 5.46 s
    [Task 13/25]  Current/Best:   14.19/  20.43 GFLOPS | Progress: (8/20) | 8.68 s
    [Task 13/25]  Current/Best:   11.37/  20.43 GFLOPS | Progress: (12/20) | 12.21 s
    [Task 13/25]  Current/Best:   16.90/  20.43 GFLOPS | Progress: (16/20) | 15.80 s
    [Task 13/25]  Current/Best:    5.10/  20.43 GFLOPS | Progress: (20/20) | 18.47 s Done.
-
    [Task 14/25]  Current/Best:    0.00/   0.00 GFLOPS | Progress: (0/20) | 0.00 s
    [Task 14/25]  Current/Best:   12.77/  12.77 GFLOPS | Progress: (4/20) | 4.27 s
    [Task 14/25]  Current/Best:   16.89/  16.89 GFLOPS | Progress: (8/20) | 7.35 s
    [Task 14/25]  Current/Best:   11.09/  16.89 GFLOPS | Progress: (12/20) | 13.60 s
    [Task 14/25]  Current/Best:   13.58/  16.89 GFLOPS | Progress: (16/20) | 21.06 s
    [Task 14/25]  Current/Best:   12.41/  16.89 GFLOPS | Progress: (20/20) | 24.18 s
    [Task 15/25]  Current/Best:    0.00/   0.00 GFLOPS | Progress: (0/20) | 0.00 s
    [Task 15/25]  Current/Best:   15.61/  15.61 GFLOPS | Progress: (4/20) | 4.63 s
    [Task 15/25]  Current/Best:   18.64/  18.64 GFLOPS | Progress: (8/20) | 10.47 s
    [Task 15/25]  Current/Best:   21.72/  21.72 GFLOPS | Progress: (12/20) | 12.45 s
    [Task 15/25]  Current/Best:   17.01/  21.72 GFLOPS | Progress: (16/20) | 14.67 s Done.
+
    [Task  1/25]  Current/Best:    0.00/   0.00 GFLOPS | Progress: (0/20) | 0.00 s
    [Task  1/25]  Current/Best:    8.48/  12.88 GFLOPS | Progress: (4/20) | 7.97 s
    [Task  1/25]  Current/Best:    5.59/  14.32 GFLOPS | Progress: (8/20) | 11.78 s
    [Task  1/25]  Current/Best:    5.67/  14.96 GFLOPS | Progress: (12/20) | 14.69 s
    [Task  1/25]  Current/Best:    9.47/  16.74 GFLOPS | Progress: (16/20) | 20.10 s
    [Task  1/25]  Current/Best:   19.17/  19.17 GFLOPS | Progress: (20/20) | 22.78 s Done.
+
    [Task  2/25]  Current/Best:    0.00/   0.00 GFLOPS | Progress: (0/20) | 0.00 s
    [Task  2/25]  Current/Best:   17.70/  17.70 GFLOPS | Progress: (4/20) | 4.98 s
    [Task  2/25]  Current/Best:    6.56/  17.70 GFLOPS | Progress: (8/20) | 8.18 s
    [Task  2/25]  Current/Best:   13.93/  21.13 GFLOPS | Progress: (12/20) | 10.09 s
    [Task  2/25]  Current/Best:   14.14/  21.13 GFLOPS | Progress: (16/20) | 12.04 s
    [Task  2/25]  Current/Best:   15.31/  21.13 GFLOPS | Progress: (20/20) | 14.13 s Done.
+
    [Task  3/25]  Current/Best:    0.00/   0.00 GFLOPS | Progress: (0/20) | 0.00 s
    [Task  3/25]  Current/Best:   15.96/  16.14 GFLOPS | Progress: (4/20) | 4.06 s
    [Task  3/25]  Current/Best:   11.86/  19.51 GFLOPS | Progress: (8/20) | 6.26 s
    [Task  3/25]  Current/Best:   18.90/  19.51 GFLOPS | Progress: (12/20) | 8.46 s
    [Task  3/25]  Current/Best:   17.83/  19.51 GFLOPS | Progress: (16/20) | 10.74 s
    [Task  3/25]  Current/Best:    5.42/  24.03 GFLOPS | Progress: (20/20) | 13.18 s Done.
+
    [Task  4/25]  Current/Best:    0.00/   0.00 GFLOPS | Progress: (0/20) | 0.00 s
    [Task  4/25]  Current/Best:   13.04/  17.45 GFLOPS | Progress: (4/20) | 4.54 s
    [Task  4/25]  Current/Best:   18.17/  18.17 GFLOPS | Progress: (8/20) | 6.35 s
    [Task  4/25]  Current/Best:   10.02/  18.17 GFLOPS | Progress: (12/20) | 13.89 s
    [Task  4/25]  Current/Best:   13.94/  18.17 GFLOPS | Progress: (16/20) | 16.23 s
    [Task  4/25]  Current/Best:   16.27/  18.62 GFLOPS | Progress: (20/20) | 18.29 s Done.
+
    [Task  5/25]  Current/Best:    0.00/   0.00 GFLOPS | Progress: (0/20) | 0.00 s
    [Task  5/25]  Current/Best:   13.97/  15.94 GFLOPS | Progress: (4/20) | 4.01 s
    [Task  5/25]  Current/Best:   11.47/  15.94 GFLOPS | Progress: (8/20) | 5.93 s
    [Task  5/25]  Current/Best:   12.14/  18.03 GFLOPS | Progress: (12/20) | 8.00 s
    [Task  5/25]  Current/Best:   10.04/  18.03 GFLOPS | Progress: (16/20) | 10.28 s
    [Task  5/25]  Current/Best:   14.13/  22.49 GFLOPS | Progress: (20/20) | 12.08 s Done.
+
    [Task  6/25]  Current/Best:    0.00/   0.00 GFLOPS | Progress: (0/20) | 0.00 s
    [Task  6/25]  Current/Best:   12.56/  14.80 GFLOPS | Progress: (4/20) | 5.29 s
    [Task  6/25]  Current/Best:   14.65/  14.80 GFLOPS | Progress: (8/20) | 8.28 s
    [Task  6/25]  Current/Best:   22.00/  22.00 GFLOPS | Progress: (12/20) | 11.91 s
    [Task  6/25]  Current/Best:   16.47/  22.00 GFLOPS | Progress: (16/20) | 14.04 s
    [Task  6/25]  Current/Best:   11.73/  22.00 GFLOPS | Progress: (20/20) | 16.94 s Done.
+
    [Task  7/25]  Current/Best:    0.00/   0.00 GFLOPS | Progress: (0/20) | 0.00 s
    [Task  7/25]  Current/Best:   14.07/  16.45 GFLOPS | Progress: (4/20) | 4.13 s
    [Task  7/25]  Current/Best:    8.76/  16.45 GFLOPS | Progress: (8/20) | 7.27 s
    [Task  7/25]  Current/Best:   11.86/  16.45 GFLOPS | Progress: (12/20) | 10.10 s
    [Task  7/25]  Current/Best:    6.82/  19.62 GFLOPS | Progress: (16/20) | 12.91 s
    [Task  7/25]  Current/Best:   20.55/  20.55 GFLOPS | Progress: (20/20) | 14.78 s Done.
+
    [Task  8/25]  Current/Best:    0.00/   0.00 GFLOPS | Progress: (0/20) | 0.00 s
    [Task  8/25]  Current/Best:   13.37/  13.37 GFLOPS | Progress: (4/20) | 8.86 s
    [Task  8/25]  Current/Best:    6.11/  18.63 GFLOPS | Progress: (8/20) | 17.89 s
    [Task  8/25]  Current/Best:   14.23/  18.63 GFLOPS | Progress: (12/20) | 20.61 s
    [Task  8/25]  Current/Best:   11.87/  21.10 GFLOPS | Progress: (16/20) | 28.03 s
    [Task  8/25]  Current/Best:   11.30/  21.10 GFLOPS | Progress: (20/20) | 39.48 s
    [Task  9/25]  Current/Best:    0.00/   0.00 GFLOPS | Progress: (0/20) | 0.00 s
    [Task  9/25]  Current/Best:    9.74/  16.50 GFLOPS | Progress: (4/20) | 4.23 s
    [Task  9/25]  Current/Best:   13.96/  16.50 GFLOPS | Progress: (8/20) | 6.16 s
    [Task  9/25]  Current/Best:   14.94/  17.81 GFLOPS | Progress: (12/20) | 8.00 s
    [Task  9/25]  Current/Best:   21.47/  21.47 GFLOPS | Progress: (16/20) | 9.70 s
    [Task  9/25]  Current/Best:   14.49/  21.47 GFLOPS | Progress: (20/20)
  | 20.80 s
    [Task 10/25]  Current/Best:    0.00/   0.00 GFLOPS | Progress: (0/20) | 0.00 s
    [Task 10/25]  Current/Best:    5.45/  19.21 GFLOPS | Progress: (4/20) | 4.27 s
    [Task 10/25]  Current/Best:   18.17/  19.21 GFLOPS | Progress: (8/20) | 6.64 s
    [Task 10/25]  Current/Best:   15.65/  19.21 GFLOPS | Progress: (12/20) | 9.02 s
    [Task 10/25]  Current/Best:   18.36/  19.21 GFLOPS | Progress: (16/20) | 12.02 s
    [Task 10/25]  Current/Best:   13.11/  19.21 GFLOPS | Progress: (20/20) | 14.06 s Done.
+
    [Task 11/25]  Current/Best:    0.00/   0.00 GFLOPS | Progress: (0/20) | 0.00 s
    [Task 11/25]  Current/Best:   15.71/  15.99 GFLOPS | Progress: (4/20) | 4.44 s
    [Task 11/25]  Current/Best:   10.91/  15.99 GFLOPS | Progress: (8/20) | 7.34 s
    [Task 11/25]  Current/Best:   17.53/  17.53 GFLOPS | Progress: (12/20) | 11.04 s
    [Task 11/25]  Current/Best:    3.11/  22.12 GFLOPS | Progress: (16/20) | 13.68 s
    [Task 11/25]  Current/Best:   22.94/  22.94 GFLOPS | Progress: (20/20) | 16.08 s Done.
+
    [Task 12/25]  Current/Best:    0.00/   0.00 GFLOPS | Progress: (0/20) | 0.00 s
    [Task 12/25]  Current/Best:   11.18/  16.22 GFLOPS | Progress: (4/20) | 8.08 s
    [Task 12/25]  Current/Best:   18.19/  18.19 GFLOPS | Progress: (8/20) | 11.61 s
    [Task 12/25]  Current/Best:    9.32/  18.50 GFLOPS | Progress: (12/20) | 15.91 s
    [Task 12/25]  Current/Best:   19.61/  19.61 GFLOPS | Progress: (16/20) | 17.92 s
    [Task 12/25]  Current/Best:   19.31/  19.61 GFLOPS | Progress: (20/20) | 20.18 s Done.
+
    [Task 13/25]  Current/Best:    0.00/   0.00 GFLOPS | Progress: (0/20) | 0.00 s
    [Task 13/25]  Current/Best:    3.10/  11.11 GFLOPS | Progress: (4/20) | 6.28 s
    [Task 13/25]  Current/Best:    3.05/  22.72 GFLOPS | Progress: (8/20) | 9.31 s
    [Task 13/25]  Current/Best:   16.31/  22.72 GFLOPS | Progress: (12/20) | 11.71 s
    [Task 13/25]  Current/Best:   18.22/  22.72 GFLOPS | Progress: (16/20) | 15.24 s
    [Task 13/25]  Current/Best:   10.33/  22.72 GFLOPS | Progress: (20/20) | 18.77 s Done.
+
    [Task 14/25]  Current/Best:    0.00/   0.00 GFLOPS | Progress: (0/20) | 0.00 s
    [Task 14/25]  Current/Best:   15.92/  15.92 GFLOPS | Progress: (4/20) | 8.54 s
    [Task 14/25]  Current/Best:   12.81/  15.92 GFLOPS | Progress: (8/20) | 14.16 s
    [Task 14/25]  Current/Best:   13.85/  16.22 GFLOPS | Progress: (12/20) | 17.44 s
    [Task 14/25]  Current/Best:    9.98/  17.20 GFLOPS | Progress: (16/20) | 23.68 s
    [Task 14/25]  Current/Best:    8.42/  17.20 GFLOPS | Progress: (20/20) | 31.06 s Done.
+
    [Task 15/25]  Current/Best:    0.00/   0.00 GFLOPS | Progress: (0/20) | 0.00 s
    [Task 15/25]  Current/Best:   17.32/  19.51 GFLOPS | Progress: (4/20) | 3.23 s Done.
      Done.
-
    [Task 15/25]  Current/Best:   14.09/  21.72 GFLOPS | Progress: (20/20) | 17.14 s Done.
-
    [Task 16/25]  Current/Best:    0.00/   0.00 GFLOPS | Progress: (0/20) | 0.00 s
    [Task 16/25]  Current/Best:    3.12/  16.01 GFLOPS | Progress: (4/20) | 4.06 s
    [Task 16/25]  Current/Best:   10.60/  17.31 GFLOPS | Progress: (8/20) | 5.81 s
    [Task 16/25]  Current/Best:   14.80/  19.13 GFLOPS | Progress: (12/20) | 7.45 s
    [Task 16/25]  Current/Best:   18.30/  19.13 GFLOPS | Progress: (16/20) | 9.63 s
    [Task 16/25]  Current/Best:   10.65/  19.13 GFLOPS | Progress: (20/20) | 12.28 s Done.
-
    [Task 17/25]  Current/Best:    0.00/   0.00 GFLOPS | Progress: (0/20) | 0.00 s
    [Task 17/25]  Current/Best:    8.49/  19.43 GFLOPS | Progress: (4/20) | 4.56 s
    [Task 17/25]  Current/Best:   13.81/  19.43 GFLOPS | Progress: (8/20) | 6.87 s
    [Task 17/25]  Current/Best:    7.69/  21.20 GFLOPS | Progress: (12/20) | 9.47 s
    [Task 17/25]  Current/Best:   18.36/  21.20 GFLOPS | Progress: (16/20) | 12.16 s
    [Task 17/25]  Current/Best:   17.57/  22.27 GFLOPS | Progress: (20/20) | 14.99 s Done.
-
    [Task 18/25]  Current/Best:    0.00/   0.00 GFLOPS | Progress: (0/20) | 0.00 s
    [Task 18/25]  Current/Best:   16.51/  18.80 GFLOPS | Progress: (4/20) | 5.99 s
    [Task 18/25]  Current/Best:    7.62/  21.69 GFLOPS | Progress: (8/20) | 9.09 s
    [Task 18/25]  Current/Best:   14.99/  21.69 GFLOPS | Progress: (12/20) | 17.82 s
    [Task 18/25]  Current/Best:   20.07/  21.69 GFLOPS | Progress: (16/20) | 21.99 s
    [Task 18/25]  Current/Best:   15.49/  21.69 GFLOPS | Progress: (20/20) | 26.70 s Done.
-
    [Task 19/25]  Current/Best:    0.00/   0.00 GFLOPS | Progress: (0/20) | 0.00 s
    [Task 19/25]  Current/Best:   17.83/  19.45 GFLOPS | Progress: (4/20) | 4.26 s
    [Task 19/25]  Current/Best:   21.40/  21.40 GFLOPS | Progress: (8/20) | 7.74 s
    [Task 19/25]  Current/Best:   11.94/  21.40 GFLOPS | Progress: (12/20) | 10.58 s
    [Task 19/25]  Current/Best:   12.06/  21.40 GFLOPS | Progress: (16/20) | 16.26 s
    [Task 19/25]  Current/Best:   14.47/  21.40 GFLOPS | Progress: (20/20) | 20.03 s Done.
-
    [Task 20/25]  Current/Best:    0.00/   0.00 GFLOPS | Progress: (0/20) | 0.00 s
    [Task 20/25]  Current/Best:   16.03/  16.03 GFLOPS | Progress: (4/20) | 3.71 s
    [Task 20/25]  Current/Best:   18.05/  18.05 GFLOPS | Progress: (8/20) | 6.63 s
    [Task 20/25]  Current/Best:   15.59/  18.05 GFLOPS | Progress: (12/20) | 9.91 s
    [Task 20/25]  Current/Best:   11.36/  18.54 GFLOPS | Progress: (16/20) | 14.21 s
    [Task 20/25]  Current/Best:   10.17/  18.54 GFLOPS | Progress: (20/20) | 16.54 s
    [Task 21/25]  Current/Best:    0.00/   0.00 GFLOPS | Progress: (0/20) | 0.00 s
    [Task 21/25]  Current/Best:   14.56/  14.56 GFLOPS | Progress: (4/20) | 4.98 s
    [Task 21/25]  Current/Best:   14.38/  14.56 GFLOPS | Progress: (8/20) | 6.74 s
    [Task 21/25]  Current/Best:   12.90/  18.99 GFLOPS | Progress: (12/20) | 8.39 s Done.
-
    [Task 21/25]  Current/Best:   12.76/  20.09 GFLOPS | Progress: (16/20) | 11.15 s
    [Task 21/25]  Current/Best:   18.24/  20.09 GFLOPS | Progress: (20/20) | 13.68 s Done.
-
    [Task 22/25]  Current/Best:    0.00/   0.00 GFLOPS | Progress: (0/20) | 0.00 s
    [Task 22/25]  Current/Best:    5.07/  19.37 GFLOPS | Progress: (4/20) | 4.42 s
    [Task 22/25]  Current/Best:   13.87/  19.37 GFLOPS | Progress: (8/20) | 6.45 s
    [Task 22/25]  Current/Best:    6.54/  19.37 GFLOPS | Progress: (12/20) | 8.31 s
    [Task 22/25]  Current/Best:   12.04/  19.37 GFLOPS | Progress: (16/20) | 11.79 s
    [Task 22/25]  Current/Best:   11.74/  19.37 GFLOPS | Progress: (20/20) | 14.13 s Done.
-
    [Task 23/25]  Current/Best:    0.00/   0.00 GFLOPS | Progress: (0/20) | 0.00 s
    [Task 23/25]  Current/Best:   13.19/  20.01 GFLOPS | Progress: (4/20) | 4.36 s
    [Task 23/25]  Current/Best:   10.89/  23.17 GFLOPS | Progress: (8/20) | 7.99 s
    [Task 23/25]  Current/Best:   19.97/  23.17 GFLOPS | Progress: (12/20) | 11.46 s
    [Task 23/25]  Current/Best:   13.22/  23.17 GFLOPS | Progress: (16/20) | 16.06 s
    [Task 23/25]  Current/Best:    1.55/  23.17 GFLOPS | Progress: (20/20) | 19.92 s Done.
-
    [Task 24/25]  Current/Best:    0.00/   0.00 GFLOPS | Progress: (0/20) | 0.00 s
    [Task 24/25]  Current/Best:    4.23/   6.66 GFLOPS | Progress: (4/20) | 12.78 s
    [Task 24/25]  Current/Best:   10.24/  10.24 GFLOPS | Progress: (8/20) | 23.72 s
    [Task 24/25]  Current/Best:    0.90/  10.24 GFLOPS | Progress: (12/20) | 34.72 s
    [Task 24/25]  Current/Best:    3.34/  10.24 GFLOPS | Progress: (16/20) | 46.28 s
    [Task 24/25]  Current/Best:    1.22/  10.24 GFLOPS | Progress: (20/20) | 57.40 s
    [Task 25/25]  Current/Best:    0.00/   0.00 GFLOPS | Progress: (0/20) | 0.00 s Done.
-
    [Task 25/25]  Current/Best:    8.26/   8.80 GFLOPS | Progress: (4/20) | 12.76 s
    [Task 25/25]  Current/Best:    8.50/   8.80 GFLOPS | Progress: (8/20) | 23.70 s
    [Task 25/25]  Current/Best:    2.57/   8.80 GFLOPS | Progress: (12/20) | 34.67 s
    [Task 25/25]  Current/Best:    1.54/   8.80 GFLOPS | Progress: (16/20) | 36.88 s
    [Task 25/25]  Current/Best:    8.07/   8.80 GFLOPS | Progress: (20/20) | 38.85 s
+
    [Task 15/25]  Current/Best:   12.04/  19.51 GFLOPS | Progress: (8/20) | 5.49 s
    [Task 15/25]  Current/Best:   13.96/  19.51 GFLOPS | Progress: (12/20) | 6.79 s
    [Task 15/25]  Current/Best:   10.77/  19.51 GFLOPS | Progress: (16/20) | 10.77 s
    [Task 15/25]  Current/Best:    5.07/  19.51 GFLOPS | Progress: (20/20) | 15.41 s
    [Task 16/25]  Current/Best:    0.00/   0.00 GFLOPS | Progress: (0/20) | 0.00 s
    [Task 16/25]  Current/Best:   14.23/  16.03 GFLOPS | Progress: (4/20) | 4.31 s
    [Task 16/25]  Current/Best:   16.12/  16.12 GFLOPS | Progress: (8/20) | 6.07 s
    [Task 16/25]  Current/Best:   14.69/  17.85 GFLOPS | Progress: (12/20) | 7.70 s
    [Task 16/25]  Current/Best:   15.07/  18.91 GFLOPS | Progress: (16/20) | 11.04 s
    [Task 16/25]  Current/Best:   11.86/  18.91 GFLOPS | Progress: (20/20) | 14.96 s Done.
+
    [Task 17/25]  Current/Best:    0.00/   0.00 GFLOPS | Progress: (0/20) | 0.00 s
    [Task 17/25]  Current/Best:   12.29/  18.66 GFLOPS | Progress: (4/20) | 5.99 s
    [Task 17/25]  Current/Best:   17.30/  18.66 GFLOPS | Progress: (8/20) | 8.55 s
    [Task 17/25]  Current/Best:   17.17/  18.66 GFLOPS | Progress: (12/20) | 10.75 s
    [Task 17/25]  Current/Best:   18.39/  20.70 GFLOPS | Progress: (16/20) | 13.03 s
    [Task 17/25]  Current/Best:   20.11/  20.70 GFLOPS | Progress: (20/20) | 15.53 s Done.
+
    [Task 18/25]  Current/Best:    0.00/   0.00 GFLOPS | Progress: (0/20) | 0.00 s
    [Task 18/25]  Current/Best:   12.19/  13.64 GFLOPS | Progress: (4/20) | 5.13 s
    [Task 18/25]  Current/Best:    3.08/  19.29 GFLOPS | Progress: (8/20) | 7.51 s
    [Task 18/25]  Current/Best:   10.35/  19.29 GFLOPS | Progress: (12/20) | 10.91 s
    [Task 18/25]  Current/Best:   18.38/  19.54 GFLOPS | Progress: (16/20) | 12.82 s
    [Task 18/25]  Current/Best:   10.40/  19.54 GFLOPS | Progress: (20/20) | 15.82 s Done.
+
    [Task 19/25]  Current/Best:    0.00/   0.00 GFLOPS | Progress: (0/20) | 0.00 s
    [Task 19/25]  Current/Best:    4.41/  12.04 GFLOPS | Progress: (4/20) | 5.40 s
    [Task 19/25]  Current/Best:   19.65/  21.16 GFLOPS | Progress: (8/20) | 8.05 s
    [Task 19/25]  Current/Best:    2.66/  21.16 GFLOPS | Progress: (12/20) | 12.11 s
    [Task 19/25]  Current/Best:    1.55/  21.16 GFLOPS | Progress: (16/20) | 15.94 s
    [Task 19/25]  Current/Best:    9.95/  21.16 GFLOPS | Progress: (20/20) | 19.17 s Done.
+
    [Task 20/25]  Current/Best:    0.00/   0.00 GFLOPS | Progress: (0/20) | 0.00 s
    [Task 20/25]  Current/Best:   16.04/  16.04 GFLOPS | Progress: (4/20) | 4.16 s
    [Task 20/25]  Current/Best:   15.40/  19.88 GFLOPS | Progress: (8/20) | 6.93 s
    [Task 20/25]  Current/Best:   10.59/  19.88 GFLOPS | Progress: (12/20) | 9.27 s
    [Task 20/25]  Current/Best:   13.18/  19.88 GFLOPS | Progress: (16/20) | 12.13 s
    [Task 20/25]  Current/Best:   10.44/  19.88 GFLOPS | Progress: (20/20) | 16.30 s
    [Task 21/25]  Current/Best:    0.00/   0.00 GFLOPS | Progress: (0/20) | 0.00 s
    [Task 21/25]  Current/Best:    9.46/  22.58 GFLOPS | Progress: (4/20) | 5.43 s
    [Task 21/25]  Current/Best:   17.26/  22.58 GFLOPS | Progress: (8/20) | 6.96 s Done.
+     Done.
+
    [Task 21/25]  Current/Best:   16.74/  22.58 GFLOPS | Progress: (12/20) | 9.03 s
    [Task 21/25]  Current/Best:   17.41/  22.58 GFLOPS | Progress: (16/20) | 12.37 s
    [Task 21/25]  Current/Best:    9.06/  22.58 GFLOPS | Progress: (20/20) | 15.05 s
    [Task 22/25]  Current/Best:    0.00/   0.00 GFLOPS | Progress: (0/20) | 0.00 s
    [Task 22/25]  Current/Best:    5.35/  18.24 GFLOPS | Progress: (4/20) | 5.21 s
    [Task 22/25]  Current/Best:   15.02/  18.24 GFLOPS | Progress: (8/20) | 6.99 s
    [Task 22/25]  Current/Best:   17.94/  18.24 GFLOPS | Progress: (12/20) | 8.62 s
    [Task 22/25]  Current/Best:   15.39/  18.92 GFLOPS | Progress: (16/20) | 10.19 s
    [Task 22/25]  Current/Best:   13.34/  21.41 GFLOPS | Progress: (20/20) | 13.10 s Done.
+
    [Task 23/25]  Current/Best:    0.00/   0.00 GFLOPS | Progress: (0/20) | 0.00 s
    [Task 23/25]  Current/Best:   10.36/  18.27 GFLOPS | Progress: (4/20) | 5.16 s
    [Task 23/25]  Current/Best:   10.63/  20.21 GFLOPS | Progress: (8/20) | 7.80 s
    [Task 23/25]  Current/Best:    9.91/  23.43 GFLOPS | Progress: (12/20) | 10.07 s
    [Task 23/25]  Current/Best:   21.07/  23.43 GFLOPS | Progress: (16/20) | 13.95 s
    [Task 23/25]  Current/Best:   16.37/  23.43 GFLOPS | Progress: (20/20) | 16.42 s Done.
+
    [Task 24/25]  Current/Best:    0.00/   0.00 GFLOPS | Progress: (0/20) | 0.00 s
    [Task 24/25]  Current/Best:    3.00/   8.72 GFLOPS | Progress: (4/20) | 12.74 s
    [Task 24/25]  Current/Best:    8.98/   8.98 GFLOPS | Progress: (8/20) | 24.23 s
    [Task 24/25]  Current/Best:    3.89/   8.98 GFLOPS | Progress: (12/20) | 36.16 s
    [Task 24/25]  Current/Best:    1.55/   8.98 GFLOPS | Progress: (16/20) | 43.41 s
    [Task 24/25]  Current/Best:    3.59/   9.40 GFLOPS | Progress: (20/20) | 52.81 s
    [Task 25/25]  Current/Best:    0.00/   0.00 GFLOPS | Progress: (0/20) | 0.00 s Done.
+
    [Task 25/25]  Current/Best:    1.55/   9.39 GFLOPS | Progress: (4/20) | 4.36 s
    [Task 25/25]  Current/Best:    5.78/   9.39 GFLOPS | Progress: (8/20) | 15.32 s
    [Task 25/25]  Current/Best:    7.00/   9.39 GFLOPS | Progress: (12/20) | 26.23 s
    [Task 25/25]  Current/Best:    9.47/   9.47 GFLOPS | Progress: (16/20) | 38.00 s
    [Task 25/25]  Current/Best:    3.94/   9.47 GFLOPS | Progress: (20/20) | 48.94 s
 
 
 
@@ -678,7 +677,7 @@ Verify that the optimized model runs and produces the same results:
 
  .. code-block:: none
 
-    class='n02123045 tabby, tabby cat' with probability=0.621102
+    class='n02123045 tabby, tabby cat' with probability=0.621104
     class='n02123159 tiger cat' with probability=0.356379
     class='n02124075 Egyptian cat' with probability=0.019712
     class='n02129604 tiger, Panthera tigris' with probability=0.001215
@@ -736,8 +735,8 @@ improvement in comparing the optimized model to the unoptimized model.
 
  .. code-block:: none
 
-    optimized: {'mean': 413.3896778399685, 'median': 413.6363552000148, 'std': 0.8094517843820932}
-    unoptimized: {'mean': 519.0110534800078, 'median': 518.6799677499948, 'std': 1.8895944523185717}
+    optimized: {'mean': 423.0049592999785, 'median': 422.38120119995983, 'std': 1.7630251695608055}
+    unoptimized: {'mean': 517.1232215099917, 'median': 517.1692200999587, 'std': 2.4466269662533757}
 
 
 
@@ -760,7 +759,7 @@ profiling/benchmarking.
 
 .. rst-class:: sphx-glr-timing
 
-   **Total running time of the script:** ( 12 minutes  11.772 seconds)
+   **Total running time of the script:** ( 12 minutes  6.783 seconds)
 
 
 .. _sphx_glr_download_tutorial_autotvm_relay_x86.py:
diff --git a/docs/_sources/tutorial/cross_compilation_and_rpc.rst.txt b/docs/_sources/tutorial/cross_compilation_and_rpc.rst.txt
index b19335a7c2..114fb643ad 100644
--- a/docs/_sources/tutorial/cross_compilation_and_rpc.rst.txt
+++ b/docs/_sources/tutorial/cross_compilation_and_rpc.rst.txt
@@ -274,7 +274,7 @@ device and returns the measured cost. Network overhead is excluded.
 
  .. code-block:: none
 
-    1.218e-07 secs/op
+    1.252e-07 secs/op
 
 
 
diff --git a/docs/_sources/tutorial/intro_topi.rst.txt b/docs/_sources/tutorial/intro_topi.rst.txt
index ad7920dcdb..8a6e90c297 100644
--- a/docs/_sources/tutorial/intro_topi.rst.txt
+++ b/docs/_sources/tutorial/intro_topi.rst.txt
@@ -264,7 +264,7 @@ As you can see, scheduled stages of computation have been accumulated and we can
 
  .. code-block:: none
 
-    [stage(a, placeholder(a, 0x20b11ef0)), stage(b, placeholder(b, 0xc6b84b0)), stage(T_add, compute(T_add, body=[(a[ax0, ax1, ax2] + b[ax1, ax2])], axis=[iter_var(ax0, range(min=0, ext=100)), iter_var(ax1, range(min=0, ext=10)), iter_var(ax2, range(min=0, ext=10))], reduce_axis=[], tag=broadcast, attrs={})), stage(T_multiply, compute(T_multiply, body=[(a[ax0, ax1, ax2]*b[ax1, ax2])], axis=[iter_var(ax0, range(min=0, ext=100)), iter_var(ax1, range(min=0, ext=10)), iter_var(ax2, range(min [...]
+    [stage(a, placeholder(a, 0x53510c0)), stage(b, placeholder(b, 0x9d4ad90)), stage(T_add, compute(T_add, body=[(a[ax0, ax1, ax2] + b[ax1, ax2])], axis=[iter_var(ax0, range(min=0, ext=100)), iter_var(ax1, range(min=0, ext=10)), iter_var(ax2, range(min=0, ext=10))], reduce_axis=[], tag=broadcast, attrs={})), stage(T_multiply, compute(T_multiply, body=[(a[ax0, ax1, ax2]*b[ax1, ax2])], axis=[iter_var(ax0, range(min=0, ext=100)), iter_var(ax1, range(min=0, ext=10)), iter_var(ax2, range(min= [...]
 
 
 
diff --git a/docs/_sources/tutorial/sg_execution_times.rst.txt b/docs/_sources/tutorial/sg_execution_times.rst.txt
index 0d075c66df..7695e60348 100644
--- a/docs/_sources/tutorial/sg_execution_times.rst.txt
+++ b/docs/_sources/tutorial/sg_execution_times.rst.txt
@@ -5,28 +5,28 @@
 
 Computation times
 =================
-**15:48.372** total execution time for **tutorial** files:
+**15:42.459** total execution time for **tutorial** files:
 
 +------------------------------------------------------------------------------------------+-----------+--------+
-| :ref:`sphx_glr_tutorial_autotvm_relay_x86.py` (``autotvm_relay_x86.py``)                 | 12:11.772 | 0.0 MB |
+| :ref:`sphx_glr_tutorial_autotvm_relay_x86.py` (``autotvm_relay_x86.py``)                 | 12:06.783 | 0.0 MB |
 +------------------------------------------------------------------------------------------+-----------+--------+
-| :ref:`sphx_glr_tutorial_auto_scheduler_matmul_x86.py` (``auto_scheduler_matmul_x86.py``) | 01:30.687 | 0.0 MB |
+| :ref:`sphx_glr_tutorial_auto_scheduler_matmul_x86.py` (``auto_scheduler_matmul_x86.py``) | 01:34.758 | 0.0 MB |
 +------------------------------------------------------------------------------------------+-----------+--------+
-| :ref:`sphx_glr_tutorial_tensor_expr_get_started.py` (``tensor_expr_get_started.py``)     | 01:00.573 | 0.0 MB |
+| :ref:`sphx_glr_tutorial_tensor_expr_get_started.py` (``tensor_expr_get_started.py``)     | 00:58.939 | 0.0 MB |
 +------------------------------------------------------------------------------------------+-----------+--------+
-| :ref:`sphx_glr_tutorial_relay_quick_start.py` (``relay_quick_start.py``)                 | 00:34.158 | 0.0 MB |
+| :ref:`sphx_glr_tutorial_relay_quick_start.py` (``relay_quick_start.py``)                 | 00:33.495 | 0.0 MB |
 +------------------------------------------------------------------------------------------+-----------+--------+
-| :ref:`sphx_glr_tutorial_autotvm_matmul_x86.py` (``autotvm_matmul_x86.py``)               | 00:29.517 | 0.0 MB |
+| :ref:`sphx_glr_tutorial_autotvm_matmul_x86.py` (``autotvm_matmul_x86.py``)               | 00:26.879 | 0.0 MB |
 +------------------------------------------------------------------------------------------+-----------+--------+
-| :ref:`sphx_glr_tutorial_intro_topi.py` (``intro_topi.py``)                               | 00:00.843 | 0.0 MB |
+| :ref:`sphx_glr_tutorial_intro_topi.py` (``intro_topi.py``)                               | 00:00.827 | 0.0 MB |
 +------------------------------------------------------------------------------------------+-----------+--------+
-| :ref:`sphx_glr_tutorial_tensor_ir_blitz_course.py` (``tensor_ir_blitz_course.py``)       | 00:00.610 | 0.0 MB |
+| :ref:`sphx_glr_tutorial_tensor_ir_blitz_course.py` (``tensor_ir_blitz_course.py``)       | 00:00.602 | 0.0 MB |
 +------------------------------------------------------------------------------------------+-----------+--------+
-| :ref:`sphx_glr_tutorial_cross_compilation_and_rpc.py` (``cross_compilation_and_rpc.py``) | 00:00.203 | 0.0 MB |
+| :ref:`sphx_glr_tutorial_cross_compilation_and_rpc.py` (``cross_compilation_and_rpc.py``) | 00:00.166 | 0.0 MB |
 +------------------------------------------------------------------------------------------+-----------+--------+
 | :ref:`sphx_glr_tutorial_introduction.py` (``introduction.py``)                           | 00:00.006 | 0.0 MB |
 +------------------------------------------------------------------------------------------+-----------+--------+
-| :ref:`sphx_glr_tutorial_uma.py` (``uma.py``)                                             | 00:00.001 | 0.0 MB |
+| :ref:`sphx_glr_tutorial_uma.py` (``uma.py``)                                             | 00:00.002 | 0.0 MB |
 +------------------------------------------------------------------------------------------+-----------+--------+
 | :ref:`sphx_glr_tutorial_tvmc_command_line_driver.py` (``tvmc_command_line_driver.py``)   | 00:00.001 | 0.0 MB |
 +------------------------------------------------------------------------------------------+-----------+--------+
diff --git a/docs/_sources/tutorial/tensor_expr_get_started.rst.txt b/docs/_sources/tutorial/tensor_expr_get_started.rst.txt
index 1057ee0844..d7d13aa7c4 100644
--- a/docs/_sources/tutorial/tensor_expr_get_started.rst.txt
+++ b/docs/_sources/tutorial/tensor_expr_get_started.rst.txt
@@ -298,7 +298,7 @@ helper function to run a profile of the TVM generated code.
 
  .. code-block:: none
 
-    Numpy running time: 0.000007
+    Numpy running time: 0.000008
     naive: 0.000007
 
 
@@ -503,10 +503,10 @@ We can now compare the different schedules
  .. code-block:: none
 
                 Operator                  Timing             Performance
-                   numpy    7.4728899926412854e-06                   1.0
-                   naive    6.691700000000001e-06     0.8954634694996796
-                parallel              6.9606e-06      0.9314468708698043
-                  vector    2.4695499999999998e-05    3.3046786483299213
+                   numpy    7.738430003882967e-06                    1.0
+                   naive    6.7403000000000005e-06    0.8710164718964792
+                parallel              7.2974e-06      0.9430078189423864
+                  vector             2.45464e-05      3.1720129260952383
 
 
 
@@ -927,7 +927,7 @@ matrix multiplication.
 
  .. code-block:: none
 
-    Numpy running time: 0.018437
+    Numpy running time: 0.018564
 
 
 
@@ -985,7 +985,7 @@ optimizations.
 
  .. code-block:: none
 
-    none: 3.357307
+    none: 3.258042
 
 
 
@@ -1087,7 +1087,7 @@ schedule.
 
  .. code-block:: none
 
-    blocking: 0.304422
+    blocking: 0.299070
 
 
 
@@ -1182,7 +1182,7 @@ already cache friendly from our previous optimizations.
 
  .. code-block:: none
 
-    vectorization: 0.343061
+    vectorization: 0.336651
     @main = primfn(A_1: handle, B_1: handle, C_1: handle) -> ()
       attr = {"from_legacy_te_schedule": True, "global_symbol": "main", "tir.noalias": True}
       buffers = {A: Buffer(A_2: Pointer(float32), float32, [1024, 1024], []),
@@ -1255,7 +1255,7 @@ more cache friendly.
 
  .. code-block:: none
 
-    loop permutation: 0.120776
+    loop permutation: 0.116442
     @main = primfn(A_1: handle, B_1: handle, C_1: handle) -> ()
       attr = {"from_legacy_te_schedule": True, "global_symbol": "main", "tir.noalias": True}
       buffers = {A: Buffer(A_2: Pointer(float32), float32, [1024, 1024], []),
@@ -1353,7 +1353,7 @@ optimized schedule.
 
  .. code-block:: none
 
-    array packing: 0.110062
+    array packing: 0.108195
     @main = primfn(A_1: handle, B_1: handle, C_1: handle) -> ()
       attr = {"from_legacy_te_schedule": True, "global_symbol": "main", "tir.noalias": True}
       buffers = {A: Buffer(A_2: Pointer(float32), float32, [1024, 1024], []),
@@ -1445,7 +1445,7 @@ to `C` when all the block results are ready.
 
  .. code-block:: none
 
-    block caching: 0.111796
+    block caching: 0.110410
     @main = primfn(A_1: handle, B_1: handle, C_1: handle) -> ()
       attr = {"from_legacy_te_schedule": True, "global_symbol": "main", "tir.noalias": True}
       buffers = {A: Buffer(A_2: Pointer(float32), float32, [1024, 1024], []),
@@ -1530,7 +1530,7 @@ of thread-level parallelization.
 
  .. code-block:: none
 
-    parallelization: 0.146442
+    parallelization: 0.144161
     @main = primfn(A_1: handle, B_1: handle, C_1: handle) -> ()
       attr = {"from_legacy_te_schedule": True, "global_symbol": "main", "tir.noalias": True}
       buffers = {A: Buffer(A_2: Pointer(float32), float32, [1024, 1024], []),
@@ -1610,13 +1610,13 @@ working, we can compare the results.
  .. code-block:: none
 
                 Operator                  Timing             Performance
-                    none      3.3573074816000004                     1.0
-                blocking            0.3044220526     0.09067446287491095
-           vectorization     0.34306051649999997       0.102183228191094
-        loop permutation     0.12077605530000002     0.03597408219590345
-           array packing            0.1100624381     0.03278294844997256
-           block caching            0.1117963456    0.033299406209502426
-         parallelization            0.1464421032     0.04361891307322549
+                    none            3.2580415795                     1.0
+                blocking            0.2990700609     0.09179442729699515
+           vectorization            0.3366510253     0.10332925995120806
+        loop permutation            0.1164417592     0.03573980146007526
+           array packing     0.10819530380000002    0.033208693369899955
+           block caching            0.1104100152      0.0338884610603847
+         parallelization            0.1441612484     0.04424782338785373
 
 
 
@@ -1656,11 +1656,6 @@ operations with tunable parameters that allows you to automatically optimize
 the computation for specific platforms.
 
 
-.. rst-class:: sphx-glr-timing
-
-   **Total running time of the script:** ( 1 minutes  0.573 seconds)
-
-
 .. _sphx_glr_download_tutorial_tensor_expr_get_started.py:
 
 .. only:: html
diff --git a/docs/commit_hash b/docs/commit_hash
index e7f8587665..9c6a7af5da 100644
--- a/docs/commit_hash
+++ b/docs/commit_hash
@@ -1 +1 @@
-d979949993819953e6eb4c5c8e8183c41db0e258
+bf4fedd06e2e60ee4a947740738949e8acf64400
diff --git a/docs/how_to/compile_models/from_darknet.html b/docs/how_to/compile_models/from_darknet.html
index 9c74354785..2c000e073c 100644
--- a/docs/how_to/compile_models/from_darknet.html
+++ b/docs/how_to/compile_models/from_darknet.html
@@ -585,7 +585,7 @@ class:[&#39;truck 0.9266&#39;] left:471 top:83 right:689 bottom:169
 class:[&#39;bicycle 0.9984&#39;] left:111 top:113 right:577 bottom:447
 </pre></div>
 </div>
-<p class="sphx-glr-timing"><strong>Total running time of the script:</strong> ( 1 minutes  10.448 seconds)</p>
+<p class="sphx-glr-timing"><strong>Total running time of the script:</strong> ( 1 minutes  9.859 seconds)</p>
 <div class="sphx-glr-footer sphx-glr-footer-example docutils container" id="sphx-glr-download-how-to-compile-models-from-darknet-py">
 <div class="sphx-glr-download sphx-glr-download-python docutils container">
 <p><a class="reference download internal" download="" href="../../_downloads/7716f96385bd5abb6e822041e285be54/from_darknet.py"><code class="xref download docutils literal notranslate"><span class="pre">Download</span> <span class="pre">Python</span> <span class="pre">source</span> <span class="pre">code:</span> <span class="pre">from_darknet.py</span></code></a></p>
diff --git a/docs/how_to/compile_models/from_keras.html b/docs/how_to/compile_models/from_keras.html
index f994724f2b..384f35c1a1 100644
--- a/docs/how_to/compile_models/from_keras.html
+++ b/docs/how_to/compile_models/from_keras.html
@@ -506,7 +506,7 @@ Tensorflow is also required since it’s used as the default backend of keras.</
 <div class="sphx-glr-script-out highlight-none notranslate"><div class="highlight"><pre><span></span>Relay top-1 id: 285, class name: Egyptian cat
 
 1/1 [==============================] - ETA: 0s
-1/1 [==============================] - 1s 957ms/step
+1/1 [==============================] - 1s 923ms/step
 Keras top-1 id: 285, class name: Egyptian cat
 </pre></div>
 </div>
diff --git a/docs/how_to/compile_models/from_mxnet.html b/docs/how_to/compile_models/from_mxnet.html
index a109b12b16..320dab5854 100644
--- a/docs/how_to/compile_models/from_mxnet.html
+++ b/docs/how_to/compile_models/from_mxnet.html
@@ -439,7 +439,7 @@
 <span class="nb">print</span><span class="p">(</span><span class="s2">&quot;x&quot;</span><span class="p">,</span> <a href="https://docs.python.org/3/library/stdtypes.html#tuple" title="builtins.tuple" class="sphx-glr-backref-module-builtins sphx-glr-backref-type-py-class sphx-glr-backref-instance"><span class="n">x</span><span class="o">.</span><span class="n">shape</span></a><span class="p">)</span>
 </pre></div>
 </div>
-<img src="../../_images/sphx_glr_from_mxnet_001.png" srcset="../../_images/sphx_glr_from_mxnet_001.png" alt="from mxnet" class = "sphx-glr-single-img"/><div class="sphx-glr-script-out highlight-none notranslate"><div class="highlight"><pre><span></span>Downloading /workspace/.mxnet/models/resnet18_v1-a0666292.zipf8026450-0129-4148-8d3d-7308ae4026b4 from https://apache-mxnet.s3-accelerate.dualstack.amazonaws.com/gluon/models/resnet18_v1-a0666292.zip...
+<img src="../../_images/sphx_glr_from_mxnet_001.png" srcset="../../_images/sphx_glr_from_mxnet_001.png" alt="from mxnet" class = "sphx-glr-single-img"/><div class="sphx-glr-script-out highlight-none notranslate"><div class="highlight"><pre><span></span>Downloading /workspace/.mxnet/models/resnet18_v1-a0666292.zip7626fc43-7722-4bb2-b4c4-e3a0d2d4c363 from https://apache-mxnet.s3-accelerate.dualstack.amazonaws.com/gluon/models/resnet18_v1-a0666292.zip...
 x (1, 3, 224, 224)
 </pre></div>
 </div>
diff --git a/docs/how_to/compile_models/from_oneflow.html b/docs/how_to/compile_models/from_oneflow.html
index c40c5aa750..da0d2a2bf4 100644
--- a/docs/how_to/compile_models/from_oneflow.html
+++ b/docs/how_to/compile_models/from_oneflow.html
@@ -449,11 +449,12 @@ Deprecated in NumPy 1.20; for more details and guidance: https://numpy.org/devdo
 <div class="sphx-glr-script-out highlight-none notranslate"><div class="highlight"><pre><span></span>Downloading: &quot;https://oneflow-public.oss-cn-beijing.aliyuncs.com/model_zoo/flowvision/classification/ResNet/resnet18.zip&quot; to /workspace/.oneflow/flowvision_cache/resnet18.zip
 
   0%|          | 0.00/41.5M [00:00&lt;?, ?B/s]
- 21%|##        | 8.59M/41.5M [00:00&lt;00:00, 90.0MB/s]
- 41%|####1     | 17.2M/41.5M [00:00&lt;00:00, 46.2MB/s]
- 58%|#####7    | 24.0M/41.5M [00:00&lt;00:00, 44.9MB/s]
- 77%|#######7  | 32.0M/41.5M [00:00&lt;00:00, 51.3MB/s]
-100%|##########| 41.5M/41.5M [00:00&lt;00:00, 60.8MB/s]
+ 19%|#9        | 7.99M/41.5M [00:00&lt;00:00, 45.6MB/s]
+ 35%|###4      | 14.3M/41.5M [00:00&lt;00:00, 41.6MB/s]
+ 53%|#####3    | 22.1M/41.5M [00:00&lt;00:00, 54.4MB/s]
+ 77%|#######7  | 32.0M/41.5M [00:00&lt;00:00, 53.0MB/s]
+ 92%|#########2| 38.3M/41.5M [00:00&lt;00:00, 51.4MB/s]
+100%|##########| 41.5M/41.5M [00:00&lt;00:00, 51.0MB/s]
 </pre></div>
 </div>
 </div>
diff --git a/docs/how_to/compile_models/from_pytorch.html b/docs/how_to/compile_models/from_pytorch.html
index 84b33bf9a5..dd09806e79 100644
--- a/docs/how_to/compile_models/from_pytorch.html
+++ b/docs/how_to/compile_models/from_pytorch.html
@@ -432,12 +432,9 @@ be unstable.</p>
 Downloading: &quot;https://download.pytorch.org/models/resnet18-f37072fd.pth&quot; to /workspace/.cache/torch/hub/checkpoints/resnet18-f37072fd.pth
 
   0%|          | 0.00/44.7M [00:00&lt;?, ?B/s]
- 18%|#7        | 7.99M/44.7M [00:00&lt;00:00, 66.9MB/s]
- 32%|###2      | 14.4M/44.7M [00:00&lt;00:00, 56.9MB/s]
- 54%|#####3    | 24.0M/44.7M [00:00&lt;00:00, 60.0MB/s]
- 66%|######6   | 29.7M/44.7M [00:00&lt;00:00, 59.4MB/s]
- 90%|########9 | 40.0M/44.7M [00:00&lt;00:00, 63.0MB/s]
-100%|##########| 44.7M/44.7M [00:00&lt;00:00, 66.9MB/s]
+ 30%|##9       | 13.2M/44.7M [00:00&lt;00:00, 63.0MB/s]
+ 71%|#######1  | 31.9M/44.7M [00:00&lt;00:00, 116MB/s]
+100%|##########| 44.7M/44.7M [00:00&lt;00:00, 118MB/s]
 </pre></div>
 </div>
 </div>
diff --git a/docs/how_to/compile_models/from_tensorflow.html b/docs/how_to/compile_models/from_tensorflow.html
index b6b608bd62..9b25f78caf 100644
--- a/docs/how_to/compile_models/from_tensorflow.html
+++ b/docs/how_to/compile_models/from_tensorflow.html
@@ -649,7 +649,7 @@ banana (score = 0.00022)
 desk (score = 0.00019)
 </pre></div>
 </div>
-<p class="sphx-glr-timing"><strong>Total running time of the script:</strong> ( 1 minutes  12.122 seconds)</p>
+<p class="sphx-glr-timing"><strong>Total running time of the script:</strong> ( 1 minutes  10.824 seconds)</p>
 <div class="sphx-glr-footer sphx-glr-footer-example docutils container" id="sphx-glr-download-how-to-compile-models-from-tensorflow-py">
 <div class="sphx-glr-download sphx-glr-download-python docutils container">
 <p><a class="reference download internal" download="" href="../../_downloads/7f1d3d1b878694c201c614c807cdebc8/from_tensorflow.py"><code class="xref download docutils literal notranslate"><span class="pre">Download</span> <span class="pre">Python</span> <span class="pre">source</span> <span class="pre">code:</span> <span class="pre">from_tensorflow.py</span></code></a></p>
diff --git a/docs/how_to/compile_models/sg_execution_times.html b/docs/how_to/compile_models/sg_execution_times.html
index 80c39bc689..f655d04927 100644
--- a/docs/how_to/compile_models/sg_execution_times.html
+++ b/docs/how_to/compile_models/sg_execution_times.html
@@ -340,7 +340,7 @@
             
   <div class="section" id="computation-times">
 <span id="sphx-glr-how-to-compile-models-sg-execution-times"></span><h1>Computation times<a class="headerlink" href="#computation-times" title="Permalink to this headline">¶</a></h1>
-<p><strong>05:47.551</strong> total execution time for <strong>how_to_compile_models</strong> files:</p>
+<p><strong>05:38.207</strong> total execution time for <strong>how_to_compile_models</strong> files:</p>
 <table class="docutils align-default">
 <colgroup>
 <col style="width: 81%" />
@@ -349,43 +349,43 @@
 </colgroup>
 <tbody>
 <tr class="row-odd"><td><p><a class="reference internal" href="from_tensorflow.html#sphx-glr-how-to-compile-models-from-tensorflow-py"><span class="std std-ref">Compile Tensorflow Models</span></a> (<code class="docutils literal notranslate"><span class="pre">from_tensorflow.py</span></code>)</p></td>
-<td><p>01:12.122</p></td>
+<td><p>01:10.824</p></td>
 <td><p>0.0 MB</p></td>
 </tr>
 <tr class="row-even"><td><p><a class="reference internal" href="from_darknet.html#sphx-glr-how-to-compile-models-from-darknet-py"><span class="std std-ref">Compile YOLO-V2 and YOLO-V3 in DarkNet Models</span></a> (<code class="docutils literal notranslate"><span class="pre">from_darknet.py</span></code>)</p></td>
-<td><p>01:10.448</p></td>
+<td><p>01:09.859</p></td>
 <td><p>0.0 MB</p></td>
 </tr>
 <tr class="row-odd"><td><p><a class="reference internal" href="from_paddle.html#sphx-glr-how-to-compile-models-from-paddle-py"><span class="std std-ref">Compile PaddlePaddle Models</span></a> (<code class="docutils literal notranslate"><span class="pre">from_paddle.py</span></code>)</p></td>
-<td><p>00:47.930</p></td>
+<td><p>00:45.970</p></td>
 <td><p>0.0 MB</p></td>
 </tr>
 <tr class="row-even"><td><p><a class="reference internal" href="from_oneflow.html#sphx-glr-how-to-compile-models-from-oneflow-py"><span class="std std-ref">Compile OneFlow Models</span></a> (<code class="docutils literal notranslate"><span class="pre">from_oneflow.py</span></code>)</p></td>
-<td><p>00:32.848</p></td>
+<td><p>00:31.237</p></td>
 <td><p>0.0 MB</p></td>
 </tr>
 <tr class="row-odd"><td><p><a class="reference internal" href="from_mxnet.html#sphx-glr-how-to-compile-models-from-mxnet-py"><span class="std std-ref">Compile MXNet Models</span></a> (<code class="docutils literal notranslate"><span class="pre">from_mxnet.py</span></code>)</p></td>
-<td><p>00:28.909</p></td>
+<td><p>00:27.711</p></td>
 <td><p>0.0 MB</p></td>
 </tr>
 <tr class="row-even"><td><p><a class="reference internal" href="from_coreml.html#sphx-glr-how-to-compile-models-from-coreml-py"><span class="std std-ref">Compile CoreML Models</span></a> (<code class="docutils literal notranslate"><span class="pre">from_coreml.py</span></code>)</p></td>
-<td><p>00:27.055</p></td>
+<td><p>00:26.331</p></td>
 <td><p>0.0 MB</p></td>
 </tr>
 <tr class="row-odd"><td><p><a class="reference internal" href="from_tflite.html#sphx-glr-how-to-compile-models-from-tflite-py"><span class="std std-ref">Compile TFLite Models</span></a> (<code class="docutils literal notranslate"><span class="pre">from_tflite.py</span></code>)</p></td>
-<td><p>00:25.686</p></td>
+<td><p>00:25.096</p></td>
 <td><p>0.0 MB</p></td>
 </tr>
 <tr class="row-even"><td><p><a class="reference internal" href="from_pytorch.html#sphx-glr-how-to-compile-models-from-pytorch-py"><span class="std std-ref">Compile PyTorch Models</span></a> (<code class="docutils literal notranslate"><span class="pre">from_pytorch.py</span></code>)</p></td>
-<td><p>00:23.471</p></td>
+<td><p>00:22.291</p></td>
 <td><p>0.0 MB</p></td>
 </tr>
 <tr class="row-odd"><td><p><a class="reference internal" href="from_keras.html#sphx-glr-how-to-compile-models-from-keras-py"><span class="std std-ref">Compile Keras Models</span></a> (<code class="docutils literal notranslate"><span class="pre">from_keras.py</span></code>)</p></td>
-<td><p>00:16.617</p></td>
+<td><p>00:16.429</p></td>
 <td><p>0.0 MB</p></td>
 </tr>
 <tr class="row-even"><td><p><a class="reference internal" href="from_onnx.html#sphx-glr-how-to-compile-models-from-onnx-py"><span class="std std-ref">Compile ONNX Models</span></a> (<code class="docutils literal notranslate"><span class="pre">from_onnx.py</span></code>)</p></td>
-<td><p>00:02.464</p></td>
+<td><p>00:02.459</p></td>
 <td><p>0.0 MB</p></td>
 </tr>
 </tbody>
diff --git a/docs/how_to/deploy_models/deploy_model_on_adreno.html b/docs/how_to/deploy_models/deploy_model_on_adreno.html
index bc7d78891e..62de30dbf3 100644
--- a/docs/how_to/deploy_models/deploy_model_on_adreno.html
+++ b/docs/how_to/deploy_models/deploy_model_on_adreno.html
@@ -920,7 +920,7 @@ Top5 predictions:
 Evaluate inference time cost...
 Execution time summary:
  mean (ms)   median (ms)    max (ms)     min (ms)     std (ms)
- 2755.8343    2754.5199    2766.7212    2752.4388      3.8831
+ 2754.4249    2754.0660    2757.6122    2753.0408      1.3760
 </pre></div>
 </div>
 <div class="sphx-glr-footer sphx-glr-footer-example docutils container" id="sphx-glr-download-how-to-deploy-models-deploy-model-on-adreno-py">
diff --git a/docs/how_to/deploy_models/deploy_model_on_android.html b/docs/how_to/deploy_models/deploy_model_on_android.html
index deb8672136..609ed0e3ed 100644
--- a/docs/how_to/deploy_models/deploy_model_on_android.html
+++ b/docs/how_to/deploy_models/deploy_model_on_android.html
@@ -662,7 +662,7 @@ to the remote android device.</p>
 Evaluate inference time cost...
 Execution time summary:
  mean (ms)   median (ms)    max (ms)     min (ms)     std (ms)
-  16.5264      16.3593      17.2411      15.9385       0.5255
+  15.9780      15.8666      16.8045      15.7341       0.3105
 </pre></div>
 </div>
 </div>
diff --git a/docs/how_to/deploy_models/deploy_object_detection_pytorch.html b/docs/how_to/deploy_models/deploy_object_detection_pytorch.html
index e9556d7702..3cc878bb85 100644
--- a/docs/how_to/deploy_models/deploy_object_detection_pytorch.html
+++ b/docs/how_to/deploy_models/deploy_object_detection_pytorch.html
@@ -454,27 +454,23 @@ be unstable.</p>
 Downloading: &quot;https://download.pytorch.org/models/maskrcnn_resnet50_fpn_coco-bf2d0c1e.pth&quot; to /workspace/.cache/torch/hub/checkpoints/maskrcnn_resnet50_fpn_coco-bf2d0c1e.pth
 
   0%|          | 0.00/170M [00:00&lt;?, ?B/s]
-  5%|4         | 7.99M/170M [00:00&lt;00:02, 78.8MB/s]
-  9%|9         | 16.0M/170M [00:00&lt;00:02, 64.8MB/s]
- 14%|#4        | 24.0M/170M [00:00&lt;00:02, 55.3MB/s]
- 19%|#8        | 32.0M/170M [00:00&lt;00:02, 55.4MB/s]
- 22%|##2       | 37.4M/170M [00:00&lt;00:02, 48.5MB/s]
- 25%|##4       | 42.1M/170M [00:00&lt;00:02, 46.9MB/s]
- 28%|##8       | 48.0M/170M [00:01&lt;00:03, 40.1MB/s]
- 37%|###6      | 62.3M/170M [00:01&lt;00:02, 52.8MB/s]
- 40%|####      | 68.8M/170M [00:01&lt;00:01, 55.9MB/s]
- 46%|####6     | 78.8M/170M [00:01&lt;00:01, 67.1MB/s]
- 52%|#####1    | 88.0M/170M [00:01&lt;00:01, 65.6MB/s]
- 57%|#####6    | 96.0M/170M [00:01&lt;00:01, 67.1MB/s]
- 61%|######1   | 104M/170M [00:01&lt;00:01, 68.7MB/s]
- 66%|######6   | 112M/170M [00:01&lt;00:00, 72.0MB/s]
- 71%|#######   | 120M/170M [00:02&lt;00:00, 73.4MB/s]
- 75%|#######5  | 128M/170M [00:02&lt;00:00, 69.8MB/s]
- 80%|########  | 136M/170M [00:02&lt;00:00, 69.8MB/s]
- 87%|########6 | 147M/170M [00:02&lt;00:00, 82.8MB/s]
- 92%|#########1| 156M/170M [00:02&lt;00:00, 74.1MB/s]
- 96%|#########5| 163M/170M [00:02&lt;00:00, 71.9MB/s]
-100%|##########| 170M/170M [00:02&lt;00:00, 64.3MB/s]
+  5%|4         | 7.99M/170M [00:00&lt;00:02, 79.9MB/s]
+  9%|9         | 16.0M/170M [00:00&lt;00:01, 81.3MB/s]
+ 17%|#7        | 28.9M/170M [00:00&lt;00:01, 105MB/s]
+ 24%|##3       | 40.0M/170M [00:00&lt;00:01, 89.8MB/s]
+ 29%|##8       | 48.9M/170M [00:00&lt;00:01, 87.5MB/s]
+ 34%|###3      | 57.5M/170M [00:00&lt;00:01, 76.2MB/s]
+ 38%|###8      | 65.0M/170M [00:00&lt;00:01, 68.9MB/s]
+ 47%|####7     | 80.0M/170M [00:01&lt;00:01, 77.9MB/s]
+ 56%|#####5    | 94.9M/170M [00:01&lt;00:00, 96.4MB/s]
+ 62%|######1   | 105M/170M [00:01&lt;00:00, 86.3MB/s]
+ 71%|#######   | 120M/170M [00:01&lt;00:00, 95.8MB/s]
+ 76%|#######6  | 129M/170M [00:01&lt;00:00, 92.9MB/s]
+ 82%|########1 | 139M/170M [00:01&lt;00:00, 91.0MB/s]
+ 87%|########7 | 148M/170M [00:01&lt;00:00, 94.1MB/s]
+ 93%|#########2| 157M/170M [00:01&lt;00:00, 89.8MB/s]
+ 98%|#########7| 166M/170M [00:01&lt;00:00, 89.3MB/s]
+100%|##########| 170M/170M [00:02&lt;00:00, 88.0MB/s]
 /venv/apache-tvm-py3.7/lib/python3.7/site-packages/torch/nn/functional.py:3897: UserWarning: To copy construct from a tensor, it is recommended to use sourceTensor.clone().detach() or sourceTensor.clone().detach().requires_grad_(True), rather than torch.tensor(sourceTensor).
   for i in range(dim)
 /venv/apache-tvm-py3.7/lib/python3.7/site-packages/torchvision/models/detection/anchor_utils.py:124: UserWarning: __floordiv__ is deprecated, and its behavior will change in a future version of pytorch. It currently rounds toward 0 (like the &#39;trunc&#39; function NOT &#39;floor&#39;). This results in incorrect rounding for negative values. To keep the current behavior, use torch.div(a, b, rounding_mode=&#39;trunc&#39;), or for actual floor division, use torch.div(a, b, rounding_mode=& [...]
@@ -572,7 +568,7 @@ torchvision rcnn models.</p>
 <div class="sphx-glr-script-out highlight-none notranslate"><div class="highlight"><pre><span></span>Get 9 valid boxes
 </pre></div>
 </div>
-<p class="sphx-glr-timing"><strong>Total running time of the script:</strong> ( 3 minutes  19.797 seconds)</p>
+<p class="sphx-glr-timing"><strong>Total running time of the script:</strong> ( 3 minutes  14.579 seconds)</p>
 <div class="sphx-glr-footer sphx-glr-footer-example docutils container" id="sphx-glr-download-how-to-deploy-models-deploy-object-detection-pytorch-py">
 <div class="sphx-glr-download sphx-glr-download-python docutils container">
 <p><a class="reference download internal" download="" href="../../_downloads/7795da4b258c8feff986668b95ef57ad/deploy_object_detection_pytorch.py"><code class="xref download docutils literal notranslate"><span class="pre">Download</span> <span class="pre">Python</span> <span class="pre">source</span> <span class="pre">code:</span> <span class="pre">deploy_object_detection_pytorch.py</span></code></a></p>
diff --git a/docs/how_to/deploy_models/deploy_prequantized.html b/docs/how_to/deploy_models/deploy_prequantized.html
index c7fa570e37..c63ca2b094 100644
--- a/docs/how_to/deploy_models/deploy_prequantized.html
+++ b/docs/how_to/deploy_models/deploy_prequantized.html
@@ -498,8 +498,8 @@ training. Other models require a full post training calibration.</p>
 Downloading: &quot;https://download.pytorch.org/models/mobilenet_v2-b0353104.pth&quot; to /workspace/.cache/torch/hub/checkpoints/mobilenet_v2-b0353104.pth
 
   0%|          | 0.00/13.6M [00:00&lt;?, ?B/s]
- 59%|#####8    | 7.99M/13.6M [00:00&lt;00:00, 57.1MB/s]
-100%|##########| 13.6M/13.6M [00:00&lt;00:00, 79.7MB/s]
+ 60%|######    | 8.14M/13.6M [00:00&lt;00:00, 85.3MB/s]
+100%|##########| 13.6M/13.6M [00:00&lt;00:00, 112MB/s]
 </pre></div>
 </div>
 </div>
@@ -590,7 +590,7 @@ output values are identical out of 1000 outputs from mobilenet v2.</p>
 </div>
 <div class="sphx-glr-script-out highlight-none notranslate"><div class="highlight"><pre><span></span>Execution time summary:
  mean (ms)   median (ms)    max (ms)     min (ms)     std (ms)
-  90.5098      90.4528      92.3407      90.1770       0.2807
+  90.3708      90.2761      91.4479      90.0140       0.2925
 </pre></div>
 </div>
 <div class="admonition note">
@@ -629,7 +629,7 @@ This includes support for the VNNI 8 bit dot product instruction (CascadeLake or
 <div class="section" id="deploy-a-quantized-tflite-model">
 <h2>Deploy a quantized TFLite Model<a class="headerlink" href="#deploy-a-quantized-tflite-model" title="Permalink to this headline">¶</a></h2>
 <p>TODO</p>
-<p class="sphx-glr-timing"><strong>Total running time of the script:</strong> ( 1 minutes  7.152 seconds)</p>
+<p class="sphx-glr-timing"><strong>Total running time of the script:</strong> ( 1 minutes  5.873 seconds)</p>
 <div class="sphx-glr-footer sphx-glr-footer-example docutils container" id="sphx-glr-download-how-to-deploy-models-deploy-prequantized-py">
 <div class="sphx-glr-download sphx-glr-download-python docutils container">
 <p><a class="reference download internal" download="" href="../../_downloads/fb8217c13f4351224c6cf3aacf1a87fc/deploy_prequantized.py"><code class="xref download docutils literal notranslate"><span class="pre">Download</span> <span class="pre">Python</span> <span class="pre">source</span> <span class="pre">code:</span> <span class="pre">deploy_prequantized.py</span></code></a></p>
diff --git a/docs/how_to/deploy_models/deploy_prequantized_tflite.html b/docs/how_to/deploy_models/deploy_prequantized_tflite.html
index 95ec00f3e8..3fb9695504 100644
--- a/docs/how_to/deploy_models/deploy_prequantized_tflite.html
+++ b/docs/how_to/deploy_models/deploy_prequantized_tflite.html
@@ -583,7 +583,7 @@ TFLite Top-5 labels: [387 102 386 341 349]
 </div>
 <div class="sphx-glr-script-out highlight-none notranslate"><div class="highlight"><pre><span></span>Execution time summary:
  mean (ms)   median (ms)    max (ms)     min (ms)     std (ms)
-  121.3945     121.2771     125.1450     120.3455      0.6445
+  120.5424     120.5388     122.1327     119.7057      0.3824
 </pre></div>
 </div>
 <div class="admonition note">
@@ -611,7 +611,7 @@ network for ARM CPU</span></a>.</p></li>
 </ul>
 </div></blockquote>
 </div>
-<p class="sphx-glr-timing"><strong>Total running time of the script:</strong> ( 2 minutes  33.878 seconds)</p>
+<p class="sphx-glr-timing"><strong>Total running time of the script:</strong> ( 2 minutes  28.183 seconds)</p>
 <div class="sphx-glr-footer sphx-glr-footer-example docutils container" id="sphx-glr-download-how-to-deploy-models-deploy-prequantized-tflite-py">
 <div class="sphx-glr-download sphx-glr-download-python docutils container">
 <p><a class="reference download internal" download="" href="../../_downloads/56691c7a27d45da61d112276334640d3/deploy_prequantized_tflite.py"><code class="xref download docutils literal notranslate"><span class="pre">Download</span> <span class="pre">Python</span> <span class="pre">source</span> <span class="pre">code:</span> <span class="pre">deploy_prequantized_tflite.py</span></code></a></p>
diff --git a/docs/how_to/deploy_models/deploy_quantized.html b/docs/how_to/deploy_models/deploy_quantized.html
index 475a2c2d4f..9fe43ad75d 100644
--- a/docs/how_to/deploy_models/deploy_quantized.html
+++ b/docs/how_to/deploy_models/deploy_quantized.html
@@ -521,7 +521,7 @@ for calibration. But the accuracy might be impacted.</p>
   DeprecationWarning,
 </pre></div>
 </div>
-<p class="sphx-glr-timing"><strong>Total running time of the script:</strong> ( 1 minutes  27.833 seconds)</p>
+<p class="sphx-glr-timing"><strong>Total running time of the script:</strong> ( 1 minutes  32.385 seconds)</p>
 <div class="sphx-glr-footer sphx-glr-footer-example docutils container" id="sphx-glr-download-how-to-deploy-models-deploy-quantized-py">
 <div class="sphx-glr-download sphx-glr-download-python docutils container">
 <p><a class="reference download internal" download="" href="../../_downloads/7810ecf51bfc05f7d5e8a400ac3e815d/deploy_quantized.py"><code class="xref download docutils literal notranslate"><span class="pre">Download</span> <span class="pre">Python</span> <span class="pre">source</span> <span class="pre">code:</span> <span class="pre">deploy_quantized.py</span></code></a></p>
diff --git a/docs/how_to/deploy_models/deploy_ssd_gluoncv.html b/docs/how_to/deploy_models/deploy_ssd_gluoncv.html
index f02b8ce3ac..568dd4c56d 100644
--- a/docs/how_to/deploy_models/deploy_ssd_gluoncv.html
+++ b/docs/how_to/deploy_models/deploy_ssd_gluoncv.html
@@ -463,22 +463,22 @@ to your device.</p>
 Downloading /workspace/.mxnet/models/ssd_512_resnet50_v1_voc-9c8b225a.zip from https://apache-mxnet.s3-accelerate.dualstack.amazonaws.com/gluon/models/ssd_512_resnet50_v1_voc-9c8b225a.zip...
 
   0%|          | 0/132723 [00:00&lt;?, ?KB/s]
-  5%|5         | 6816/132723 [00:00&lt;00:01, 68151.94KB/s]
- 12%|#1        | 15413/132723 [00:00&lt;00:01, 78503.22KB/s]
- 18%|#8        | 24101/132723 [00:00&lt;00:01, 82319.91KB/s]
- 25%|##4       | 32802/132723 [00:00&lt;00:01, 84165.21KB/s]
- 31%|###1      | 41555/132723 [00:00&lt;00:01, 85374.95KB/s]
- 38%|###7      | 50208/132723 [00:00&lt;00:00, 85764.79KB/s]
- 44%|####4     | 58999/132723 [00:00&lt;00:00, 86464.57KB/s]
- 51%|#####1    | 67902/132723 [00:00&lt;00:00, 87278.52KB/s]
- 58%|#####7    | 76936/132723 [00:00&lt;00:00, 88231.93KB/s]
- 65%|######4   | 85905/132723 [00:01&lt;00:00, 88680.67KB/s]
- 72%|#######1  | 94935/132723 [00:01&lt;00:00, 89173.39KB/s]
- 78%|#######8  | 103923/132723 [00:01&lt;00:00, 89386.54KB/s]
- 85%|########5 | 112884/132723 [00:01&lt;00:00, 89452.96KB/s]
- 92%|#########1| 121856/132723 [00:01&lt;00:00, 89528.03KB/s]
- 99%|#########8| 130809/132723 [00:01&lt;00:00, 89515.98KB/s]
-100%|##########| 132723/132723 [00:01&lt;00:00, 87096.42KB/s]
+  5%|4         | 6005/132723 [00:00&lt;00:02, 60036.75KB/s]
+ 11%|#         | 13948/132723 [00:00&lt;00:01, 71441.04KB/s]
+ 17%|#6        | 22552/132723 [00:00&lt;00:01, 78105.37KB/s]
+ 24%|##3       | 31193/132723 [00:00&lt;00:01, 81382.58KB/s]
+ 30%|##9       | 39332/132723 [00:00&lt;00:01, 80918.42KB/s]
+ 36%|###6      | 47973/132723 [00:00&lt;00:01, 82776.58KB/s]
+ 43%|####2     | 56665/132723 [00:00&lt;00:00, 84124.80KB/s]
+ 49%|####9     | 65314/132723 [00:00&lt;00:00, 84874.39KB/s]
+ 56%|#####5    | 73903/132723 [00:00&lt;00:00, 85188.53KB/s]
+ 62%|######2   | 82532/132723 [00:01&lt;00:00, 85526.10KB/s]
+ 69%|######8   | 91169/132723 [00:01&lt;00:00, 85781.17KB/s]
+ 75%|#######5  | 99909/132723 [00:01&lt;00:00, 86272.37KB/s]
+ 82%|########1 | 108553/132723 [00:01&lt;00:00, 86321.98KB/s]
+ 88%|########8 | 117186/132723 [00:01&lt;00:00, 86139.24KB/s]
+ 95%|#########4| 125801/132723 [00:01&lt;00:00, 86084.69KB/s]
+100%|##########| 132723/132723 [00:01&lt;00:00, 83737.23KB/s]
 </pre></div>
 </div>
 <p>Create TVM runtime and do inference
@@ -517,7 +517,7 @@ Downloading /workspace/.mxnet/models/ssd_512_resnet50_v1_voc-9c8b225a.zip from h
 <span class="n">plt</span><span class="o">.</span><span class="n">show</span><span class="p">()</span>
 </pre></div>
 </div>
-<img src="../../_images/sphx_glr_deploy_ssd_gluoncv_001.png" srcset="../../_images/sphx_glr_deploy_ssd_gluoncv_001.png" alt="deploy ssd gluoncv" class = "sphx-glr-single-img"/><p class="sphx-glr-timing"><strong>Total running time of the script:</strong> ( 3 minutes  12.197 seconds)</p>
+<img src="../../_images/sphx_glr_deploy_ssd_gluoncv_001.png" srcset="../../_images/sphx_glr_deploy_ssd_gluoncv_001.png" alt="deploy ssd gluoncv" class = "sphx-glr-single-img"/><p class="sphx-glr-timing"><strong>Total running time of the script:</strong> ( 3 minutes  5.615 seconds)</p>
 <div class="sphx-glr-footer sphx-glr-footer-example docutils container" id="sphx-glr-download-how-to-deploy-models-deploy-ssd-gluoncv-py">
 <div class="sphx-glr-download sphx-glr-download-python docutils container">
 <p><a class="reference download internal" download="" href="../../_downloads/cccb17d28e5e8b2e94ea8cd5ec59f6ed/deploy_ssd_gluoncv.py"><code class="xref download docutils literal notranslate"><span class="pre">Download</span> <span class="pre">Python</span> <span class="pre">source</span> <span class="pre">code:</span> <span class="pre">deploy_ssd_gluoncv.py</span></code></a></p>
diff --git a/docs/how_to/deploy_models/sg_execution_times.html b/docs/how_to/deploy_models/sg_execution_times.html
index 6b8d4bb72a..76eab1cdd6 100644
--- a/docs/how_to/deploy_models/sg_execution_times.html
+++ b/docs/how_to/deploy_models/sg_execution_times.html
@@ -340,7 +340,7 @@
             
   <div class="section" id="computation-times">
 <span id="sphx-glr-how-to-deploy-models-sg-execution-times"></span><h1>Computation times<a class="headerlink" href="#computation-times" title="Permalink to this headline">¶</a></h1>
-<p><strong>14:03.078</strong> total execution time for <strong>how_to_deploy_models</strong> files:</p>
+<p><strong>13:46.850</strong> total execution time for <strong>how_to_deploy_models</strong> files:</p>
 <table class="docutils align-default">
 <colgroup>
 <col style="width: 86%" />
@@ -349,39 +349,39 @@
 </colgroup>
 <tbody>
 <tr class="row-odd"><td><p><a class="reference internal" href="deploy_object_detection_pytorch.html#sphx-glr-how-to-deploy-models-deploy-object-detection-pytorch-py"><span class="std std-ref">Compile PyTorch Object Detection Models</span></a> (<code class="docutils literal notranslate"><span class="pre">deploy_object_detection_pytorch.py</span></code>)</p></td>
-<td><p>03:19.797</p></td>
+<td><p>03:14.579</p></td>
 <td><p>0.0 MB</p></td>
 </tr>
 <tr class="row-even"><td><p><a class="reference internal" href="deploy_ssd_gluoncv.html#sphx-glr-how-to-deploy-models-deploy-ssd-gluoncv-py"><span class="std std-ref">Deploy Single Shot Multibox Detector(SSD) model</span></a> (<code class="docutils literal notranslate"><span class="pre">deploy_ssd_gluoncv.py</span></code>)</p></td>
-<td><p>03:12.197</p></td>
+<td><p>03:05.615</p></td>
 <td><p>0.0 MB</p></td>
 </tr>
 <tr class="row-odd"><td><p><a class="reference internal" href="deploy_prequantized_tflite.html#sphx-glr-how-to-deploy-models-deploy-prequantized-tflite-py"><span class="std std-ref">Deploy a Framework-prequantized Model with TVM - Part 3 (TFLite)</span></a> (<code class="docutils literal notranslate"><span class="pre">deploy_prequantized_tflite.py</span></code>)</p></td>
-<td><p>02:33.878</p></td>
+<td><p>02:28.183</p></td>
 <td><p>0.0 MB</p></td>
 </tr>
 <tr class="row-even"><td><p><a class="reference internal" href="deploy_quantized.html#sphx-glr-how-to-deploy-models-deploy-quantized-py"><span class="std std-ref">Deploy a Quantized Model on Cuda</span></a> (<code class="docutils literal notranslate"><span class="pre">deploy_quantized.py</span></code>)</p></td>
-<td><p>01:27.833</p></td>
+<td><p>01:32.385</p></td>
 <td><p>0.0 MB</p></td>
 </tr>
 <tr class="row-odd"><td><p><a class="reference internal" href="deploy_prequantized.html#sphx-glr-how-to-deploy-models-deploy-prequantized-py"><span class="std std-ref">Deploy a Framework-prequantized Model with TVM</span></a> (<code class="docutils literal notranslate"><span class="pre">deploy_prequantized.py</span></code>)</p></td>
-<td><p>01:07.152</p></td>
+<td><p>01:05.873</p></td>
 <td><p>0.0 MB</p></td>
 </tr>
 <tr class="row-even"><td><p><a class="reference internal" href="deploy_model_on_adreno.html#sphx-glr-how-to-deploy-models-deploy-model-on-adreno-py"><span class="std std-ref">Deploy the Pretrained Model on Adreno</span></a> (<code class="docutils literal notranslate"><span class="pre">deploy_model_on_adreno.py</span></code>)</p></td>
-<td><p>00:54.484</p></td>
+<td><p>00:53.857</p></td>
 <td><p>0.0 MB</p></td>
 </tr>
 <tr class="row-odd"><td><p><a class="reference internal" href="deploy_model_on_android.html#sphx-glr-how-to-deploy-models-deploy-model-on-android-py"><span class="std std-ref">Deploy the Pretrained Model on Android</span></a> (<code class="docutils literal notranslate"><span class="pre">deploy_model_on_android.py</span></code>)</p></td>
-<td><p>00:36.060</p></td>
+<td><p>00:35.316</p></td>
 <td><p>0.0 MB</p></td>
 </tr>
 <tr class="row-even"><td><p><a class="reference internal" href="deploy_model_on_nano.html#sphx-glr-how-to-deploy-models-deploy-model-on-nano-py"><span class="std std-ref">Deploy the Pretrained Model on Jetson Nano</span></a> (<code class="docutils literal notranslate"><span class="pre">deploy_model_on_nano.py</span></code>)</p></td>
-<td><p>00:26.002</p></td>
+<td><p>00:25.701</p></td>
 <td><p>0.0 MB</p></td>
 </tr>
 <tr class="row-odd"><td><p><a class="reference internal" href="deploy_model_on_rasp.html#sphx-glr-how-to-deploy-models-deploy-model-on-rasp-py"><span class="std std-ref">Deploy the Pretrained Model on Raspberry Pi</span></a> (<code class="docutils literal notranslate"><span class="pre">deploy_model_on_rasp.py</span></code>)</p></td>
-<td><p>00:25.669</p></td>
+<td><p>00:25.334</p></td>
 <td><p>0.0 MB</p></td>
 </tr>
 <tr class="row-even"><td><p><a class="reference internal" href="deploy_sparse.html#sphx-glr-how-to-deploy-models-deploy-sparse-py"><span class="std std-ref">Deploy a Hugging Face Pruned Model on CPU</span></a> (<code class="docutils literal notranslate"><span class="pre">deploy_sparse.py</span></code>)</p></td>
diff --git a/docs/how_to/extend_tvm/bring_your_own_datatypes.html b/docs/how_to/extend_tvm/bring_your_own_datatypes.html
index 803a9ac8cf..9062a3c430 100644
--- a/docs/how_to/extend_tvm/bring_your_own_datatypes.html
+++ b/docs/how_to/extend_tvm/bring_your_own_datatypes.html
@@ -622,7 +622,7 @@ In this alpha state of the Bring Your Own Datatypes framework, we have not imple
 <span class="n">module</span><span class="p">,</span> <a href="https://docs.python.org/3/library/stdtypes.html#dict" title="builtins.dict" class="sphx-glr-backref-module-builtins sphx-glr-backref-type-py-class sphx-glr-backref-instance"><span class="n">params</span></a> <span class="o">=</span> <span class="n">get_mobilenet</span><span class="p">()</span>
 </pre></div>
 </div>
-<div class="sphx-glr-script-out highlight-none notranslate"><div class="highlight"><pre><span></span>Downloading /workspace/.mxnet/models/mobilenet0.25-9f83e440.zipcba088a8-c84b-423d-b81b-c9e460cb7f68 from https://apache-mxnet.s3-accelerate.dualstack.amazonaws.com/gluon/models/mobilenet0.25-9f83e440.zip...
+<div class="sphx-glr-script-out highlight-none notranslate"><div class="highlight"><pre><span></span>Downloading /workspace/.mxnet/models/mobilenet0.25-9f83e440.zip258d8e55-a0b4-49bf-8a93-02ba6b31ee53 from https://apache-mxnet.s3-accelerate.dualstack.amazonaws.com/gluon/models/mobilenet0.25-9f83e440.zip...
 </pre></div>
 </div>
 <p>It’s easy to execute MobileNet with native TVM:</p>
diff --git a/docs/how_to/extend_tvm/sg_execution_times.html b/docs/how_to/extend_tvm/sg_execution_times.html
index 0474b80eb2..074ce5ab0e 100644
--- a/docs/how_to/extend_tvm/sg_execution_times.html
+++ b/docs/how_to/extend_tvm/sg_execution_times.html
@@ -340,7 +340,7 @@
             
   <div class="section" id="computation-times">
 <span id="sphx-glr-how-to-extend-tvm-sg-execution-times"></span><h1>Computation times<a class="headerlink" href="#computation-times" title="Permalink to this headline">¶</a></h1>
-<p><strong>00:48.648</strong> total execution time for <strong>how_to_extend_tvm</strong> files:</p>
+<p><strong>00:47.361</strong> total execution time for <strong>how_to_extend_tvm</strong> files:</p>
 <table class="docutils align-default">
 <colgroup>
 <col style="width: 84%" />
@@ -349,19 +349,19 @@
 </colgroup>
 <tbody>
 <tr class="row-odd"><td><p><a class="reference internal" href="bring_your_own_datatypes.html#sphx-glr-how-to-extend-tvm-bring-your-own-datatypes-py"><span class="std std-ref">Bring Your Own Datatypes to TVM</span></a> (<code class="docutils literal notranslate"><span class="pre">bring_your_own_datatypes.py</span></code>)</p></td>
-<td><p>00:45.102</p></td>
+<td><p>00:43.920</p></td>
 <td><p>0.0 MB</p></td>
 </tr>
 <tr class="row-even"><td><p><a class="reference internal" href="use_pass_instrument.html#sphx-glr-how-to-extend-tvm-use-pass-instrument-py"><span class="std std-ref">How to Use TVM Pass Instrument</span></a> (<code class="docutils literal notranslate"><span class="pre">use_pass_instrument.py</span></code>)</p></td>
-<td><p>00:02.478</p></td>
+<td><p>00:02.409</p></td>
 <td><p>0.0 MB</p></td>
 </tr>
 <tr class="row-odd"><td><p><a class="reference internal" href="use_pass_infra.html#sphx-glr-how-to-extend-tvm-use-pass-infra-py"><span class="std std-ref">How to Use TVM Pass Infra</span></a> (<code class="docutils literal notranslate"><span class="pre">use_pass_infra.py</span></code>)</p></td>
-<td><p>00:01.059</p></td>
+<td><p>00:01.024</p></td>
 <td><p>0.0 MB</p></td>
 </tr>
 <tr class="row-even"><td><p><a class="reference internal" href="low_level_custom_pass.html#sphx-glr-how-to-extend-tvm-low-level-custom-pass-py"><span class="std std-ref">Writing a Customized Pass</span></a> (<code class="docutils literal notranslate"><span class="pre">low_level_custom_pass.py</span></code>)</p></td>
-<td><p>00:00.009</p></td>
+<td><p>00:00.008</p></td>
 <td><p>0.0 MB</p></td>
 </tr>
 </tbody>
diff --git a/docs/how_to/extend_tvm/use_pass_instrument.html b/docs/how_to/extend_tvm/use_pass_instrument.html
index f8245cacc7..5fae859a76 100644
--- a/docs/how_to/extend_tvm/use_pass_instrument.html
+++ b/docs/how_to/extend_tvm/use_pass_instrument.html
@@ -526,10 +526,10 @@ profile the execution time of each passes.</p>
 </pre></div>
 </div>
 <div class="sphx-glr-script-out highlight-none notranslate"><div class="highlight"><pre><span></span>Printing results of timing profile...
-InferType: 7467us [7467us] (47.17%; 47.17%)
-FoldScaleAxis: 8363us [7us] (52.83%; 52.83%)
-        FoldConstant: 8356us [1717us] (52.79%; 99.92%)
-                InferType: 6639us [6639us] (41.94%; 79.45%)
+InferType: 7196us [7196us] (46.44%; 46.44%)
+FoldScaleAxis: 8299us [7us] (53.56%; 53.56%)
+        FoldConstant: 8292us [1698us] (53.52%; 99.92%)
+                InferType: 6594us [6594us] (42.56%; 79.53%)
 </pre></div>
 </div>
 </div>
@@ -551,10 +551,10 @@ Refer to following sections and <a class="reference internal" href="../../refere
 </pre></div>
 </div>
 <div class="sphx-glr-script-out highlight-none notranslate"><div class="highlight"><pre><span></span>Printing results of timing profile...
-InferType: 6707us [6707us] (44.67%; 44.67%)
-FoldScaleAxis: 8307us [6us] (55.33%; 55.33%)
-        FoldConstant: 8301us [1715us] (55.29%; 99.93%)
-                InferType: 6586us [6586us] (43.87%; 79.34%)
+InferType: 6608us [6608us] (44.86%; 44.86%)
+FoldScaleAxis: 8123us [6us] (55.14%; 55.14%)
+        FoldConstant: 8118us [1676us] (55.10%; 99.93%)
+                InferType: 6441us [6441us] (43.72%; 79.35%)
 </pre></div>
 </div>
 <p>Register empty list to clear existing instruments.</p>
diff --git a/docs/how_to/optimize_operators/opt_conv_cuda.html b/docs/how_to/optimize_operators/opt_conv_cuda.html
index 4709b7a576..ed53b5cafe 100644
--- a/docs/how_to/optimize_operators/opt_conv_cuda.html
+++ b/docs/how_to/optimize_operators/opt_conv_cuda.html
@@ -578,7 +578,7 @@ latency of convolution.</p>
 <span class="nb">print</span><span class="p">(</span><span class="s2">&quot;Convolution: </span><span class="si">%f</span><span class="s2"> ms&quot;</span> <span class="o">%</span> <span class="p">(</span><span class="n">evaluator</span><span class="p">(</span><span class="n">a</span><span class="p">,</span> <span class="n">w</span><span class="p">,</span> <span class="n">b</span><span class="p">)</span><span class="o">.</span><span class="n">mean</span> <span class="o">*</span> <span cl [...]
 </pre></div>
 </div>
-<div class="sphx-glr-script-out highlight-none notranslate"><div class="highlight"><pre><span></span>Convolution: 54.149791 ms
+<div class="sphx-glr-script-out highlight-none notranslate"><div class="highlight"><pre><span></span>Convolution: 53.049343 ms
 </pre></div>
 </div>
 <div class="sphx-glr-footer sphx-glr-footer-example docutils container" id="sphx-glr-download-how-to-optimize-operators-opt-conv-cuda-py">
diff --git a/docs/how_to/optimize_operators/opt_conv_tensorcore.html b/docs/how_to/optimize_operators/opt_conv_tensorcore.html
index faa64d86b0..e95c0d682c 100644
--- a/docs/how_to/optimize_operators/opt_conv_tensorcore.html
+++ b/docs/how_to/optimize_operators/opt_conv_tensorcore.html
@@ -915,7 +915,7 @@ be able to run on our build server</p>
     <span class="nb">print</span><span class="p">(</span><span class="s2">&quot;conv2d with tensor core: </span><span class="si">%f</span><span class="s2"> ms&quot;</span> <span class="o">%</span> <span class="p">(</span><span class="n">evaluator</span><span class="p">(</span><span class="n">a</span><span class="p">,</span> <span class="n">w</span><span class="p">,</span> <span class="n">c</span><span class="p">)</span><span class="o">.</span><span class="n">mean</span> <span class="o">* [...]
 </pre></div>
 </div>
-<div class="sphx-glr-script-out highlight-none notranslate"><div class="highlight"><pre><span></span>conv2d with tensor core: 13.359639 ms
+<div class="sphx-glr-script-out highlight-none notranslate"><div class="highlight"><pre><span></span>conv2d with tensor core: 13.356647 ms
 </pre></div>
 </div>
 </div>
diff --git a/docs/how_to/optimize_operators/opt_gemm.html b/docs/how_to/optimize_operators/opt_gemm.html
index 8c6a80d38e..f81018c9d6 100644
--- a/docs/how_to/optimize_operators/opt_gemm.html
+++ b/docs/how_to/optimize_operators/opt_gemm.html
@@ -475,8 +475,8 @@ Then we write a baseline implementation, the simplest way to write a matrix mult
 <span class="nb">print</span><span class="p">(</span><span class="s2">&quot;Baseline: </span><span class="si">%f</span><span class="s2">&quot;</span> <span class="o">%</span> <span class="n">evaluator</span><span class="p">(</span><span class="n">a</span><span class="p">,</span> <span class="n">b</span><span class="p">,</span> <span class="n">c</span><span class="p">)</span><span class="o">.</span><span class="n">mean</span><span class="p">)</span>
 </pre></div>
 </div>
-<div class="sphx-glr-script-out highlight-none notranslate"><div class="highlight"><pre><span></span>Numpy running time: 0.019011
-Baseline: 3.401171
+<div class="sphx-glr-script-out highlight-none notranslate"><div class="highlight"><pre><span></span>Numpy running time: 0.018666
+Baseline: 3.409708
 </pre></div>
 </div>
 <p>In TVM, we can always inspect lower level IR to debug or optimize our schedule.
@@ -535,7 +535,7 @@ fill 32 * 32 * sizeof(float) which is 4KB in the cache whose total size is 32KB
 <span class="nb">print</span><span class="p">(</span><span class="s2">&quot;Opt1: </span><span class="si">%f</span><span class="s2">&quot;</span> <span class="o">%</span> <span class="n">evaluator</span><span class="p">(</span><span class="n">a</span><span class="p">,</span> <span class="n">b</span><span class="p">,</span> <span class="n">c</span><span class="p">)</span><span class="o">.</span><span class="n">mean</span><span class="p">)</span>
 </pre></div>
 </div>
-<div class="sphx-glr-script-out highlight-none notranslate"><div class="highlight"><pre><span></span>Opt1: 0.325391
+<div class="sphx-glr-script-out highlight-none notranslate"><div class="highlight"><pre><span></span>Opt1: 0.303196
 </pre></div>
 </div>
 <p>Here is the generated IR after blocking.</p>
@@ -601,7 +601,7 @@ vastly.</p>
 <span class="nb">print</span><span class="p">(</span><span class="s2">&quot;Opt2: </span><span class="si">%f</span><span class="s2">&quot;</span> <span class="o">%</span> <span class="n">evaluator</span><span class="p">(</span><span class="n">a</span><span class="p">,</span> <span class="n">b</span><span class="p">,</span> <span class="n">c</span><span class="p">)</span><span class="o">.</span><span class="n">mean</span><span class="p">)</span>
 </pre></div>
 </div>
-<div class="sphx-glr-script-out highlight-none notranslate"><div class="highlight"><pre><span></span>Opt2: 0.350191
+<div class="sphx-glr-script-out highlight-none notranslate"><div class="highlight"><pre><span></span>Opt2: 0.339247
 </pre></div>
 </div>
 <p>Here is the generated IR after vectorization.</p>
@@ -661,7 +661,7 @@ the access pattern for A matrix is more cache friendly.</p>
 <span class="nb">print</span><span class="p">(</span><span class="s2">&quot;Opt3: </span><span class="si">%f</span><span class="s2">&quot;</span> <span class="o">%</span> <span class="n">evaluator</span><span class="p">(</span><span class="n">a</span><span class="p">,</span> <span class="n">b</span><span class="p">,</span> <span class="n">c</span><span class="p">)</span><span class="o">.</span><span class="n">mean</span><span class="p">)</span>
 </pre></div>
 </div>
-<div class="sphx-glr-script-out highlight-none notranslate"><div class="highlight"><pre><span></span>Opt3: 0.119524
+<div class="sphx-glr-script-out highlight-none notranslate"><div class="highlight"><pre><span></span>Opt3: 0.117116
 </pre></div>
 </div>
 <p>Here is the generated IR after loop permutation.</p>
@@ -743,7 +743,7 @@ flattening.</p>
 <span class="nb">print</span><span class="p">(</span><span class="s2">&quot;Opt4: </span><span class="si">%f</span><span class="s2">&quot;</span> <span class="o">%</span> <span class="n">evaluator</span><span class="p">(</span><span class="n">a</span><span class="p">,</span> <span class="n">b</span><span class="p">,</span> <span class="n">c</span><span class="p">)</span><span class="o">.</span><span class="n">mean</span><span class="p">)</span>
 </pre></div>
 </div>
-<div class="sphx-glr-script-out highlight-none notranslate"><div class="highlight"><pre><span></span>Opt4: 0.109798
+<div class="sphx-glr-script-out highlight-none notranslate"><div class="highlight"><pre><span></span>Opt4: 0.109304
 </pre></div>
 </div>
 <p>Here is the generated IR after array packing.</p>
@@ -828,7 +828,7 @@ write to C when all the block results are ready.</p>
 <span class="nb">print</span><span class="p">(</span><span class="s2">&quot;Opt5: </span><span class="si">%f</span><span class="s2">&quot;</span> <span class="o">%</span> <span class="n">evaluator</span><span class="p">(</span><span class="n">a</span><span class="p">,</span> <span class="n">b</span><span class="p">,</span> <span class="n">c</span><span class="p">)</span><span class="o">.</span><span class="n">mean</span><span class="p">)</span>
 </pre></div>
 </div>
-<div class="sphx-glr-script-out highlight-none notranslate"><div class="highlight"><pre><span></span>Opt5: 0.111408
+<div class="sphx-glr-script-out highlight-none notranslate"><div class="highlight"><pre><span></span>Opt5: 0.110906
 </pre></div>
 </div>
 <p>Here is the generated IR after blocking.</p>
@@ -917,7 +917,7 @@ write to C when all the block results are ready.</p>
 <span class="nb">print</span><span class="p">(</span><span class="s2">&quot;Opt6: </span><span class="si">%f</span><span class="s2">&quot;</span> <span class="o">%</span> <span class="n">opt6_time</span><span class="p">)</span>
 </pre></div>
 </div>
-<div class="sphx-glr-script-out highlight-none notranslate"><div class="highlight"><pre><span></span>Opt6: 0.147337
+<div class="sphx-glr-script-out highlight-none notranslate"><div class="highlight"><pre><span></span>Opt6: 0.147578
 </pre></div>
 </div>
 <p>Here is the generated IR after parallelization.</p>
diff --git a/docs/how_to/optimize_operators/sg_execution_times.html b/docs/how_to/optimize_operators/sg_execution_times.html
index aaaf46fb17..87428a752d 100644
--- a/docs/how_to/optimize_operators/sg_execution_times.html
+++ b/docs/how_to/optimize_operators/sg_execution_times.html
@@ -340,7 +340,7 @@
             
   <div class="section" id="computation-times">
 <span id="sphx-glr-how-to-optimize-operators-sg-execution-times"></span><h1>Computation times<a class="headerlink" href="#computation-times" title="Permalink to this headline">¶</a></h1>
-<p><strong>00:35.675</strong> total execution time for <strong>how_to_optimize_operators</strong> files:</p>
+<p><strong>00:35.161</strong> total execution time for <strong>how_to_optimize_operators</strong> files:</p>
 <table class="docutils align-default">
 <colgroup>
 <col style="width: 83%" />
@@ -349,15 +349,15 @@
 </colgroup>
 <tbody>
 <tr class="row-odd"><td><p><a class="reference internal" href="opt_gemm.html#sphx-glr-how-to-optimize-operators-opt-gemm-py"><span class="std std-ref">How to optimize GEMM on CPU</span></a> (<code class="docutils literal notranslate"><span class="pre">opt_gemm.py</span></code>)</p></td>
-<td><p>00:32.872</p></td>
+<td><p>00:32.403</p></td>
 <td><p>0.0 MB</p></td>
 </tr>
 <tr class="row-even"><td><p><a class="reference internal" href="opt_conv_tensorcore.html#sphx-glr-how-to-optimize-operators-opt-conv-tensorcore-py"><span class="std std-ref">How to optimize convolution using TensorCores</span></a> (<code class="docutils literal notranslate"><span class="pre">opt_conv_tensorcore.py</span></code>)</p></td>
-<td><p>00:01.627</p></td>
+<td><p>00:01.553</p></td>
 <td><p>0.0 MB</p></td>
 </tr>
 <tr class="row-odd"><td><p><a class="reference internal" href="opt_conv_cuda.html#sphx-glr-how-to-optimize-operators-opt-conv-cuda-py"><span class="std std-ref">How to optimize convolution on GPU</span></a> (<code class="docutils literal notranslate"><span class="pre">opt_conv_cuda.py</span></code>)</p></td>
-<td><p>00:01.175</p></td>
+<td><p>00:01.206</p></td>
 <td><p>0.0 MB</p></td>
 </tr>
 </tbody>
diff --git a/docs/how_to/tune_with_autoscheduler/sg_execution_times.html b/docs/how_to/tune_with_autoscheduler/sg_execution_times.html
index a7a2faa453..3bcc7bbbd3 100644
--- a/docs/how_to/tune_with_autoscheduler/sg_execution_times.html
+++ b/docs/how_to/tune_with_autoscheduler/sg_execution_times.html
@@ -340,7 +340,7 @@
             
   <div class="section" id="computation-times">
 <span id="sphx-glr-how-to-tune-with-autoscheduler-sg-execution-times"></span><h1>Computation times<a class="headerlink" href="#computation-times" title="Permalink to this headline">¶</a></h1>
-<p><strong>09:13.652</strong> total execution time for <strong>how_to_tune_with_autoscheduler</strong> files:</p>
+<p><strong>09:03.580</strong> total execution time for <strong>how_to_tune_with_autoscheduler</strong> files:</p>
 <table class="docutils align-default">
 <colgroup>
 <col style="width: 85%" />
@@ -349,27 +349,27 @@
 </colgroup>
 <tbody>
 <tr class="row-odd"><td><p><a class="reference internal" href="tune_conv2d_layer_cuda.html#sphx-glr-how-to-tune-with-autoscheduler-tune-conv2d-layer-cuda-py"><span class="std std-ref">Auto-scheduling a Convolution Layer for GPU</span></a> (<code class="docutils literal notranslate"><span class="pre">tune_conv2d_layer_cuda.py</span></code>)</p></td>
-<td><p>05:43.471</p></td>
+<td><p>05:37.856</p></td>
 <td><p>0.0 MB</p></td>
 </tr>
 <tr class="row-even"><td><p><a class="reference internal" href="tune_network_x86.html#sphx-glr-how-to-tune-with-autoscheduler-tune-network-x86-py"><span class="std std-ref">Auto-scheduling a Neural Network for x86 CPU</span></a> (<code class="docutils literal notranslate"><span class="pre">tune_network_x86.py</span></code>)</p></td>
-<td><p>01:34.285</p></td>
+<td><p>01:31.696</p></td>
 <td><p>0.0 MB</p></td>
 </tr>
 <tr class="row-odd"><td><p><a class="reference internal" href="tune_network_cuda.html#sphx-glr-how-to-tune-with-autoscheduler-tune-network-cuda-py"><span class="std std-ref">Auto-scheduling a Neural Network for NVIDIA GPU</span></a> (<code class="docutils literal notranslate"><span class="pre">tune_network_cuda.py</span></code>)</p></td>
-<td><p>01:02.866</p></td>
+<td><p>01:01.645</p></td>
 <td><p>0.0 MB</p></td>
 </tr>
 <tr class="row-even"><td><p><a class="reference internal" href="tune_sparse_x86.html#sphx-glr-how-to-tune-with-autoscheduler-tune-sparse-x86-py"><span class="std std-ref">Auto-scheduling Sparse Matrix Multiplication on CPU with Custom Sketch Rule</span></a> (<code class="docutils literal notranslate"><span class="pre">tune_sparse_x86.py</span></code>)</p></td>
-<td><p>00:29.070</p></td>
+<td><p>00:29.326</p></td>
 <td><p>0.0 MB</p></td>
 </tr>
 <tr class="row-odd"><td><p><a class="reference internal" href="tune_network_arm.html#sphx-glr-how-to-tune-with-autoscheduler-tune-network-arm-py"><span class="std std-ref">Auto-scheduling a Neural Network for ARM CPU</span></a> (<code class="docutils literal notranslate"><span class="pre">tune_network_arm.py</span></code>)</p></td>
-<td><p>00:12.384</p></td>
+<td><p>00:12.006</p></td>
 <td><p>0.0 MB</p></td>
 </tr>
 <tr class="row-even"><td><p><a class="reference internal" href="tune_network_mali.html#sphx-glr-how-to-tune-with-autoscheduler-tune-network-mali-py"><span class="std std-ref">Auto-scheduling a Neural Network for mali GPU</span></a> (<code class="docutils literal notranslate"><span class="pre">tune_network_mali.py</span></code>)</p></td>
-<td><p>00:11.575</p></td>
+<td><p>00:11.050</p></td>
 <td><p>0.0 MB</p></td>
 </tr>
 </tbody>
diff --git a/docs/how_to/tune_with_autoscheduler/tune_conv2d_layer_cuda.html b/docs/how_to/tune_with_autoscheduler/tune_conv2d_layer_cuda.html
index b62791fa5a..1fca0fcb9e 100644
--- a/docs/how_to/tune_with_autoscheduler/tune_conv2d_layer_cuda.html
+++ b/docs/how_to/tune_with_autoscheduler/tune_conv2d_layer_cuda.html
@@ -489,9 +489,6 @@ file and apply it.</p>
 <span class="k">del</span> <span class="n">measure_ctx</span>
 </pre></div>
 </div>
-<div class="sphx-glr-script-out highlight-none notranslate"><div class="highlight"><pre><span></span>.T
-</pre></div>
-</div>
 <p>We can lower the schedule to see the IR after auto-scheduling.
 The auto-scheduler correctly performs optimizations including multi-level tiling,
 cooperative fetching, unrolling and operator fusion.</p>
@@ -509,10 +506,10 @@ cooperative fetching, unrolling and operator fusion.</p>
   buffer_map = {data_1: data, kernel_1: kernel, bias_1: bias, compute_1: compute} {
   attr [IterVar(blockIdx.x: int32, (nullptr), &quot;ThreadIndex&quot;, &quot;blockIdx.x&quot;)] &quot;thread_extent&quot; = 28;
   allocate(conv2d_nchw: Pointer(local float32), float32, [14]), storage_scope = local;
-  allocate(pad_temp.shared: Pointer(shared float32), float32, [72]), storage_scope = shared;
-  allocate(kernel.shared: Pointer(shared float32), float32, [3072]), storage_scope = shared;
+  allocate(pad_temp.shared: Pointer(shared float32), float32, [216]), storage_scope = shared;
+  allocate(kernel.shared: Pointer(shared float32), float32, [9216]), storage_scope = shared;
   attr [IterVar(threadIdx.x: int32, (nullptr), &quot;ThreadIndex&quot;, &quot;threadIdx.x&quot;)] &quot;thread_extent&quot; = 64 {
-    conv2d_nchw_1: Buffer(conv2d_nchw, float32, [14], [], scope=&quot;local&quot;, align=32)[0] = 0f32
+    conv2d_nchw_1: Buffer(conv2d_nchw, float32, [1], [], scope=&quot;local&quot;, align=4)[0] = 0f32
     conv2d_nchw_1[1] = 0f32
     conv2d_nchw_1[2] = 0f32
     conv2d_nchw_1[3] = 0f32
@@ -527,464 +524,1331 @@ cooperative fetching, unrolling and operator fusion.</p>
     conv2d_nchw_1[12] = 0f32
     conv2d_nchw_1[13] = 0f32
     for (rc.outer.outer: int32, 0, 64) {
-      for (ry.outer.outer: int32, 0, 3) {
-        let cse_var_2: int32 = (rc.outer.outer*72)
-        let cse_var_1: int32 = (ry.outer.outer*3)
-         {
-          attr [IterVar(threadIdx.x_1: int32, (nullptr), &quot;ThreadIndex&quot;, &quot;threadIdx.x&quot;)] &quot;thread_extent&quot; = 64 {
-            if @tir.likely((threadIdx.x_1 &lt; 18), dtype=bool) {
-              pad_temp.shared_1: Buffer(pad_temp.shared, float32, [72], [], scope=&quot;shared&quot;)[(threadIdx.x_1*4)] = @tir.if_then_else(((((1 &lt;= (ry.outer.outer + floormod(blockIdx.x, 7))) &amp;&amp; ((ry.outer.outer + floormod(blockIdx.x, 7)) &lt; 8)) &amp;&amp; (1 &lt;= floormod((threadIdx.x_1*4), 9))) &amp;&amp; (floormod((threadIdx.x_1*4), 9) &lt; 8)), data_3: Buffer(data_2, float32, [25088], [])[((((((rc.outer.outer*392) + (floordiv((threadIdx.x_1*4), 9)*49)) + (ry.outer.out [...]
-            }
-            if @tir.likely((threadIdx.x_1 &lt; 18), dtype=bool) {
-              pad_temp.shared_1[((threadIdx.x_1*4) + 1)] = @tir.if_then_else(((((1 &lt;= (ry.outer.outer + floormod(blockIdx.x, 7))) &amp;&amp; ((ry.outer.outer + floormod(blockIdx.x, 7)) &lt; 8)) &amp;&amp; (1 &lt;= floormod(((threadIdx.x_1*4) + 1), 9))) &amp;&amp; (floormod(((threadIdx.x_1*4) + 1), 9) &lt; 8)), data_3[((((((rc.outer.outer*392) + (floordiv(((threadIdx.x_1*4) + 1), 9)*49)) + (ry.outer.outer*7)) + (floormod(blockIdx.x, 7)*7)) + floormod(((threadIdx.x_1*4) + 1), 9)) - 8)], [...]
-            }
-            if @tir.likely((threadIdx.x_1 &lt; 18), dtype=bool) {
-              pad_temp.shared_1[((threadIdx.x_1*4) + 2)] = @tir.if_then_else(((((1 &lt;= (ry.outer.outer + floormod(blockIdx.x, 7))) &amp;&amp; ((ry.outer.outer + floormod(blockIdx.x, 7)) &lt; 8)) &amp;&amp; (1 &lt;= floormod(((threadIdx.x_1*4) + 2), 9))) &amp;&amp; (floormod(((threadIdx.x_1*4) + 2), 9) &lt; 8)), data_3[((((((rc.outer.outer*392) + (floordiv(((threadIdx.x_1*4) + 2), 9)*49)) + (ry.outer.outer*7)) + (floormod(blockIdx.x, 7)*7)) + floormod(((threadIdx.x_1*4) + 2), 9)) - 8)], [...]
-            }
-            if @tir.likely((threadIdx.x_1 &lt; 18), dtype=bool) {
-              pad_temp.shared_1[((threadIdx.x_1*4) + 3)] = @tir.if_then_else(((((1 &lt;= (ry.outer.outer + floormod(blockIdx.x, 7))) &amp;&amp; ((ry.outer.outer + floormod(blockIdx.x, 7)) &lt; 8)) &amp;&amp; (1 &lt;= floormod(((threadIdx.x_1*4) + 3), 9))) &amp;&amp; (floormod(((threadIdx.x_1*4) + 3), 9) &lt; 8)), data_3[((((((rc.outer.outer*392) + (floordiv(((threadIdx.x_1*4) + 3), 9)*49)) + (ry.outer.outer*7)) + (floormod(blockIdx.x, 7)*7)) + floormod(((threadIdx.x_1*4) + 3), 9)) - 8)], [...]
-            }
-          }
-          attr [IterVar(threadIdx.x_2: int32, (nullptr), &quot;ThreadIndex&quot;, &quot;threadIdx.x&quot;)] &quot;thread_extent&quot; = 64;
-          kernel.shared_1: Buffer(kernel.shared, float32, [3072], [], scope=&quot;shared&quot;)[threadIdx.x_2] = kernel_3: Buffer(kernel_2, float32, [2359296], [])[((((((floordiv(blockIdx.x, 7)*589824) + (floordiv(threadIdx.x_2, 24)*4608)) + cse_var_2) + (floordiv(floormod(threadIdx.x_2, 24), 3)*9)) + cse_var_1) + floormod(threadIdx.x_2, 3))]
-          attr [IterVar(threadIdx.x_2, (nullptr), &quot;ThreadIndex&quot;, &quot;threadIdx.x&quot;)] &quot;thread_extent&quot; = 64;
-          kernel.shared_1[(threadIdx.x_2 + 64)] = kernel_3[((((((floordiv(blockIdx.x, 7)*589824) + (floordiv((threadIdx.x_2 + 64), 24)*4608)) + cse_var_2) + (floordiv(floormod((threadIdx.x_2 + 16), 24), 3)*9)) + cse_var_1) + floormod((threadIdx.x_2 + 1), 3))]
-          attr [IterVar(threadIdx.x_2, (nullptr), &quot;ThreadIndex&quot;, &quot;threadIdx.x&quot;)] &quot;thread_extent&quot; = 64;
-          kernel.shared_1[(threadIdx.x_2 + 128)] = kernel_3[((((((floordiv(blockIdx.x, 7)*589824) + (floordiv((threadIdx.x_2 + 128), 24)*4608)) + cse_var_2) + (floordiv(floormod((threadIdx.x_2 + 8), 24), 3)*9)) + cse_var_1) + floormod((threadIdx.x_2 + 2), 3))]
-          attr [IterVar(threadIdx.x_2, (nullptr), &quot;ThreadIndex&quot;, &quot;threadIdx.x&quot;)] &quot;thread_extent&quot; = 64;
-          kernel.shared_1[(threadIdx.x_2 + 192)] = kernel_3[(((((((floordiv(blockIdx.x, 7)*589824) + (floordiv(threadIdx.x_2, 24)*4608)) + cse_var_2) + (floordiv(floormod(threadIdx.x_2, 24), 3)*9)) + cse_var_1) + floormod(threadIdx.x_2, 3)) + 36864)]
-          attr [IterVar(threadIdx.x_2, (nullptr), &quot;ThreadIndex&quot;, &quot;threadIdx.x&quot;)] &quot;thread_extent&quot; = 64;
-          kernel.shared_1[(threadIdx.x_2 + 256)] = kernel_3[((((((floordiv(blockIdx.x, 7)*589824) + (floordiv((threadIdx.x_2 + 256), 24)*4608)) + cse_var_2) + (floordiv(floormod((threadIdx.x_2 + 16), 24), 3)*9)) + cse_var_1) + floormod((threadIdx.x_2 + 1), 3))]
-          attr [IterVar(threadIdx.x_2, (nullptr), &quot;ThreadIndex&quot;, &quot;threadIdx.x&quot;)] &quot;thread_extent&quot; = 64;
-          kernel.shared_1[(threadIdx.x_2 + 320)] = kernel_3[((((((floordiv(blockIdx.x, 7)*589824) + (floordiv((threadIdx.x_2 + 320), 24)*4608)) + cse_var_2) + (floordiv(floormod((threadIdx.x_2 + 8), 24), 3)*9)) + cse_var_1) + floormod((threadIdx.x_2 + 2), 3))]
-          attr [IterVar(threadIdx.x_2, (nullptr), &quot;ThreadIndex&quot;, &quot;threadIdx.x&quot;)] &quot;thread_extent&quot; = 64;
-          kernel.shared_1[(threadIdx.x_2 + 384)] = kernel_3[(((((((floordiv(blockIdx.x, 7)*589824) + (floordiv(threadIdx.x_2, 24)*4608)) + cse_var_2) + (floordiv(floormod(threadIdx.x_2, 24), 3)*9)) + cse_var_1) + floormod(threadIdx.x_2, 3)) + 73728)]
-          attr [IterVar(threadIdx.x_2, (nullptr), &quot;ThreadIndex&quot;, &quot;threadIdx.x&quot;)] &quot;thread_extent&quot; = 64;
-          kernel.shared_1[(threadIdx.x_2 + 448)] = kernel_3[((((((floordiv(blockIdx.x, 7)*589824) + (floordiv((threadIdx.x_2 + 448), 24)*4608)) + cse_var_2) + (floordiv(floormod((threadIdx.x_2 + 16), 24), 3)*9)) + cse_var_1) + floormod((threadIdx.x_2 + 1), 3))]
-          attr [IterVar(threadIdx.x_2, (nullptr), &quot;ThreadIndex&quot;, &quot;threadIdx.x&quot;)] &quot;thread_extent&quot; = 64;
-          kernel.shared_1[(threadIdx.x_2 + 512)] = kernel_3[((((((floordiv(blockIdx.x, 7)*589824) + (floordiv((threadIdx.x_2 + 512), 24)*4608)) + cse_var_2) + (floordiv(floormod((threadIdx.x_2 + 8), 24), 3)*9)) + cse_var_1) + floormod((threadIdx.x_2 + 2), 3))]
-          attr [IterVar(threadIdx.x_2, (nullptr), &quot;ThreadIndex&quot;, &quot;threadIdx.x&quot;)] &quot;thread_extent&quot; = 64;
-          kernel.shared_1[(threadIdx.x_2 + 576)] = kernel_3[(((((((floordiv(blockIdx.x, 7)*589824) + (floordiv(threadIdx.x_2, 24)*4608)) + cse_var_2) + (floordiv(floormod(threadIdx.x_2, 24), 3)*9)) + cse_var_1) + floormod(threadIdx.x_2, 3)) + 110592)]
-          attr [IterVar(threadIdx.x_2, (nullptr), &quot;ThreadIndex&quot;, &quot;threadIdx.x&quot;)] &quot;thread_extent&quot; = 64;
-          kernel.shared_1[(threadIdx.x_2 + 640)] = kernel_3[((((((floordiv(blockIdx.x, 7)*589824) + (floordiv((threadIdx.x_2 + 640), 24)*4608)) + cse_var_2) + (floordiv(floormod((threadIdx.x_2 + 16), 24), 3)*9)) + cse_var_1) + floormod((threadIdx.x_2 + 1), 3))]
-          attr [IterVar(threadIdx.x_2, (nullptr), &quot;ThreadIndex&quot;, &quot;threadIdx.x&quot;)] &quot;thread_extent&quot; = 64;
-          kernel.shared_1[(threadIdx.x_2 + 704)] = kernel_3[((((((floordiv(blockIdx.x, 7)*589824) + (floordiv((threadIdx.x_2 + 704), 24)*4608)) + cse_var_2) + (floordiv(floormod((threadIdx.x_2 + 8), 24), 3)*9)) + cse_var_1) + floormod((threadIdx.x_2 + 2), 3))]
-          attr [IterVar(threadIdx.x_2, (nullptr), &quot;ThreadIndex&quot;, &quot;threadIdx.x&quot;)] &quot;thread_extent&quot; = 64;
-          kernel.shared_1[(threadIdx.x_2 + 768)] = kernel_3[(((((((floordiv(blockIdx.x, 7)*589824) + (floordiv(threadIdx.x_2, 24)*4608)) + cse_var_2) + (floordiv(floormod(threadIdx.x_2, 24), 3)*9)) + cse_var_1) + floormod(threadIdx.x_2, 3)) + 147456)]
-          attr [IterVar(threadIdx.x_2, (nullptr), &quot;ThreadIndex&quot;, &quot;threadIdx.x&quot;)] &quot;thread_extent&quot; = 64;
-          kernel.shared_1[(threadIdx.x_2 + 832)] = kernel_3[((((((floordiv(blockIdx.x, 7)*589824) + (floordiv((threadIdx.x_2 + 832), 24)*4608)) + cse_var_2) + (floordiv(floormod((threadIdx.x_2 + 16), 24), 3)*9)) + cse_var_1) + floormod((threadIdx.x_2 + 1), 3))]
-          attr [IterVar(threadIdx.x_2, (nullptr), &quot;ThreadIndex&quot;, &quot;threadIdx.x&quot;)] &quot;thread_extent&quot; = 64;
-          kernel.shared_1[(threadIdx.x_2 + 896)] = kernel_3[((((((floordiv(blockIdx.x, 7)*589824) + (floordiv((threadIdx.x_2 + 896), 24)*4608)) + cse_var_2) + (floordiv(floormod((threadIdx.x_2 + 8), 24), 3)*9)) + cse_var_1) + floormod((threadIdx.x_2 + 2), 3))]
-          attr [IterVar(threadIdx.x_2, (nullptr), &quot;ThreadIndex&quot;, &quot;threadIdx.x&quot;)] &quot;thread_extent&quot; = 64;
-          kernel.shared_1[(threadIdx.x_2 + 960)] = kernel_3[(((((((floordiv(blockIdx.x, 7)*589824) + (floordiv(threadIdx.x_2, 24)*4608)) + cse_var_2) + (floordiv(floormod(threadIdx.x_2, 24), 3)*9)) + cse_var_1) + floormod(threadIdx.x_2, 3)) + 184320)]
-          attr [IterVar(threadIdx.x_2, (nullptr), &quot;ThreadIndex&quot;, &quot;threadIdx.x&quot;)] &quot;thread_extent&quot; = 64;
-          kernel.shared_1[(threadIdx.x_2 + 1024)] = kernel_3[((((((floordiv(blockIdx.x, 7)*589824) + (floordiv((threadIdx.x_2 + 1024), 24)*4608)) + cse_var_2) + (floordiv(floormod((threadIdx.x_2 + 16), 24), 3)*9)) + cse_var_1) + floormod((threadIdx.x_2 + 1), 3))]
-          attr [IterVar(threadIdx.x_2, (nullptr), &quot;ThreadIndex&quot;, &quot;threadIdx.x&quot;)] &quot;thread_extent&quot; = 64;
-          kernel.shared_1[(threadIdx.x_2 + 1088)] = kernel_3[((((((floordiv(blockIdx.x, 7)*589824) + (floordiv((threadIdx.x_2 + 1088), 24)*4608)) + cse_var_2) + (floordiv(floormod((threadIdx.x_2 + 8), 24), 3)*9)) + cse_var_1) + floormod((threadIdx.x_2 + 2), 3))]
-          attr [IterVar(threadIdx.x_2, (nullptr), &quot;ThreadIndex&quot;, &quot;threadIdx.x&quot;)] &quot;thread_extent&quot; = 64;
-          kernel.shared_1[(threadIdx.x_2 + 1152)] = kernel_3[(((((((floordiv(blockIdx.x, 7)*589824) + (floordiv(threadIdx.x_2, 24)*4608)) + cse_var_2) + (floordiv(floormod(threadIdx.x_2, 24), 3)*9)) + cse_var_1) + floormod(threadIdx.x_2, 3)) + 221184)]
-          attr [IterVar(threadIdx.x_2, (nullptr), &quot;ThreadIndex&quot;, &quot;threadIdx.x&quot;)] &quot;thread_extent&quot; = 64;
-          kernel.shared_1[(threadIdx.x_2 + 1216)] = kernel_3[((((((floordiv(blockIdx.x, 7)*589824) + (floordiv((threadIdx.x_2 + 1216), 24)*4608)) + cse_var_2) + (floordiv(floormod((threadIdx.x_2 + 16), 24), 3)*9)) + cse_var_1) + floormod((threadIdx.x_2 + 1), 3))]
-          attr [IterVar(threadIdx.x_2, (nullptr), &quot;ThreadIndex&quot;, &quot;threadIdx.x&quot;)] &quot;thread_extent&quot; = 64;
-          kernel.shared_1[(threadIdx.x_2 + 1280)] = kernel_3[((((((floordiv(blockIdx.x, 7)*589824) + (floordiv((threadIdx.x_2 + 1280), 24)*4608)) + cse_var_2) + (floordiv(floormod((threadIdx.x_2 + 8), 24), 3)*9)) + cse_var_1) + floormod((threadIdx.x_2 + 2), 3))]
-          attr [IterVar(threadIdx.x_2, (nullptr), &quot;ThreadIndex&quot;, &quot;threadIdx.x&quot;)] &quot;thread_extent&quot; = 64;
-          kernel.shared_1[(threadIdx.x_2 + 1344)] = kernel_3[(((((((floordiv(blockIdx.x, 7)*589824) + (floordiv(threadIdx.x_2, 24)*4608)) + cse_var_2) + (floordiv(floormod(threadIdx.x_2, 24), 3)*9)) + cse_var_1) + floormod(threadIdx.x_2, 3)) + 258048)]
-          attr [IterVar(threadIdx.x_2, (nullptr), &quot;ThreadIndex&quot;, &quot;threadIdx.x&quot;)] &quot;thread_extent&quot; = 64;
-          kernel.shared_1[(threadIdx.x_2 + 1408)] = kernel_3[((((((floordiv(blockIdx.x, 7)*589824) + (floordiv((threadIdx.x_2 + 1408), 24)*4608)) + cse_var_2) + (floordiv(floormod((threadIdx.x_2 + 16), 24), 3)*9)) + cse_var_1) + floormod((threadIdx.x_2 + 1), 3))]
-          attr [IterVar(threadIdx.x_2, (nullptr), &quot;ThreadIndex&quot;, &quot;threadIdx.x&quot;)] &quot;thread_extent&quot; = 64;
-          kernel.shared_1[(threadIdx.x_2 + 1472)] = kernel_3[((((((floordiv(blockIdx.x, 7)*589824) + (floordiv((threadIdx.x_2 + 1472), 24)*4608)) + cse_var_2) + (floordiv(floormod((threadIdx.x_2 + 8), 24), 3)*9)) + cse_var_1) + floormod((threadIdx.x_2 + 2), 3))]
-          attr [IterVar(threadIdx.x_2, (nullptr), &quot;ThreadIndex&quot;, &quot;threadIdx.x&quot;)] &quot;thread_extent&quot; = 64;
-          kernel.shared_1[(threadIdx.x_2 + 1536)] = kernel_3[(((((((floordiv(blockIdx.x, 7)*589824) + (floordiv(threadIdx.x_2, 24)*4608)) + cse_var_2) + (floordiv(floormod(threadIdx.x_2, 24), 3)*9)) + cse_var_1) + floormod(threadIdx.x_2, 3)) + 294912)]
-          attr [IterVar(threadIdx.x_2, (nullptr), &quot;ThreadIndex&quot;, &quot;threadIdx.x&quot;)] &quot;thread_extent&quot; = 64;
-          kernel.shared_1[(threadIdx.x_2 + 1600)] = kernel_3[((((((floordiv(blockIdx.x, 7)*589824) + (floordiv((threadIdx.x_2 + 1600), 24)*4608)) + cse_var_2) + (floordiv(floormod((threadIdx.x_2 + 16), 24), 3)*9)) + cse_var_1) + floormod((threadIdx.x_2 + 1), 3))]
-          attr [IterVar(threadIdx.x_2, (nullptr), &quot;ThreadIndex&quot;, &quot;threadIdx.x&quot;)] &quot;thread_extent&quot; = 64;
-          kernel.shared_1[(threadIdx.x_2 + 1664)] = kernel_3[((((((floordiv(blockIdx.x, 7)*589824) + (floordiv((threadIdx.x_2 + 1664), 24)*4608)) + cse_var_2) + (floordiv(floormod((threadIdx.x_2 + 8), 24), 3)*9)) + cse_var_1) + floormod((threadIdx.x_2 + 2), 3))]
-          attr [IterVar(threadIdx.x_2, (nullptr), &quot;ThreadIndex&quot;, &quot;threadIdx.x&quot;)] &quot;thread_extent&quot; = 64;
-          kernel.shared_1[(threadIdx.x_2 + 1728)] = kernel_3[(((((((floordiv(blockIdx.x, 7)*589824) + (floordiv(threadIdx.x_2, 24)*4608)) + cse_var_2) + (floordiv(floormod(threadIdx.x_2, 24), 3)*9)) + cse_var_1) + floormod(threadIdx.x_2, 3)) + 331776)]
-          attr [IterVar(threadIdx.x_2, (nullptr), &quot;ThreadIndex&quot;, &quot;threadIdx.x&quot;)] &quot;thread_extent&quot; = 64;
-          kernel.shared_1[(threadIdx.x_2 + 1792)] = kernel_3[((((((floordiv(blockIdx.x, 7)*589824) + (floordiv((threadIdx.x_2 + 1792), 24)*4608)) + cse_var_2) + (floordiv(floormod((threadIdx.x_2 + 16), 24), 3)*9)) + cse_var_1) + floormod((threadIdx.x_2 + 1), 3))]
-          attr [IterVar(threadIdx.x_2, (nullptr), &quot;ThreadIndex&quot;, &quot;threadIdx.x&quot;)] &quot;thread_extent&quot; = 64;
-          kernel.shared_1[(threadIdx.x_2 + 1856)] = kernel_3[((((((floordiv(blockIdx.x, 7)*589824) + (floordiv((threadIdx.x_2 + 1856), 24)*4608)) + cse_var_2) + (floordiv(floormod((threadIdx.x_2 + 8), 24), 3)*9)) + cse_var_1) + floormod((threadIdx.x_2 + 2), 3))]
-          attr [IterVar(threadIdx.x_2, (nullptr), &quot;ThreadIndex&quot;, &quot;threadIdx.x&quot;)] &quot;thread_extent&quot; = 64;
-          kernel.shared_1[(threadIdx.x_2 + 1920)] = kernel_3[(((((((floordiv(blockIdx.x, 7)*589824) + (floordiv(threadIdx.x_2, 24)*4608)) + cse_var_2) + (floordiv(floormod(threadIdx.x_2, 24), 3)*9)) + cse_var_1) + floormod(threadIdx.x_2, 3)) + 368640)]
-          attr [IterVar(threadIdx.x_2, (nullptr), &quot;ThreadIndex&quot;, &quot;threadIdx.x&quot;)] &quot;thread_extent&quot; = 64;
-          kernel.shared_1[(threadIdx.x_2 + 1984)] = kernel_3[((((((floordiv(blockIdx.x, 7)*589824) + (floordiv((threadIdx.x_2 + 1984), 24)*4608)) + cse_var_2) + (floordiv(floormod((threadIdx.x_2 + 16), 24), 3)*9)) + cse_var_1) + floormod((threadIdx.x_2 + 1), 3))]
-          attr [IterVar(threadIdx.x_2, (nullptr), &quot;ThreadIndex&quot;, &quot;threadIdx.x&quot;)] &quot;thread_extent&quot; = 64;
-          kernel.shared_1[(threadIdx.x_2 + 2048)] = kernel_3[((((((floordiv(blockIdx.x, 7)*589824) + (floordiv((threadIdx.x_2 + 2048), 24)*4608)) + cse_var_2) + (floordiv(floormod((threadIdx.x_2 + 8), 24), 3)*9)) + cse_var_1) + floormod((threadIdx.x_2 + 2), 3))]
-          attr [IterVar(threadIdx.x_2, (nullptr), &quot;ThreadIndex&quot;, &quot;threadIdx.x&quot;)] &quot;thread_extent&quot; = 64;
-          kernel.shared_1[(threadIdx.x_2 + 2112)] = kernel_3[(((((((floordiv(blockIdx.x, 7)*589824) + (floordiv(threadIdx.x_2, 24)*4608)) + cse_var_2) + (floordiv(floormod(threadIdx.x_2, 24), 3)*9)) + cse_var_1) + floormod(threadIdx.x_2, 3)) + 405504)]
-          attr [IterVar(threadIdx.x_2, (nullptr), &quot;ThreadIndex&quot;, &quot;threadIdx.x&quot;)] &quot;thread_extent&quot; = 64;
-          kernel.shared_1[(threadIdx.x_2 + 2176)] = kernel_3[((((((floordiv(blockIdx.x, 7)*589824) + (floordiv((threadIdx.x_2 + 2176), 24)*4608)) + cse_var_2) + (floordiv(floormod((threadIdx.x_2 + 16), 24), 3)*9)) + cse_var_1) + floormod((threadIdx.x_2 + 1), 3))]
-          attr [IterVar(threadIdx.x_2, (nullptr), &quot;ThreadIndex&quot;, &quot;threadIdx.x&quot;)] &quot;thread_extent&quot; = 64;
-          kernel.shared_1[(threadIdx.x_2 + 2240)] = kernel_3[((((((floordiv(blockIdx.x, 7)*589824) + (floordiv((threadIdx.x_2 + 2240), 24)*4608)) + cse_var_2) + (floordiv(floormod((threadIdx.x_2 + 8), 24), 3)*9)) + cse_var_1) + floormod((threadIdx.x_2 + 2), 3))]
-          attr [IterVar(threadIdx.x_2, (nullptr), &quot;ThreadIndex&quot;, &quot;threadIdx.x&quot;)] &quot;thread_extent&quot; = 64;
-          kernel.shared_1[(threadIdx.x_2 + 2304)] = kernel_3[(((((((floordiv(blockIdx.x, 7)*589824) + (floordiv(threadIdx.x_2, 24)*4608)) + cse_var_2) + (floordiv(floormod(threadIdx.x_2, 24), 3)*9)) + cse_var_1) + floormod(threadIdx.x_2, 3)) + 442368)]
-          attr [IterVar(threadIdx.x_2, (nullptr), &quot;ThreadIndex&quot;, &quot;threadIdx.x&quot;)] &quot;thread_extent&quot; = 64;
-          kernel.shared_1[(threadIdx.x_2 + 2368)] = kernel_3[((((((floordiv(blockIdx.x, 7)*589824) + (floordiv((threadIdx.x_2 + 2368), 24)*4608)) + cse_var_2) + (floordiv(floormod((threadIdx.x_2 + 16), 24), 3)*9)) + cse_var_1) + floormod((threadIdx.x_2 + 1), 3))]
-          attr [IterVar(threadIdx.x_2, (nullptr), &quot;ThreadIndex&quot;, &quot;threadIdx.x&quot;)] &quot;thread_extent&quot; = 64;
-          kernel.shared_1[(threadIdx.x_2 + 2432)] = kernel_3[((((((floordiv(blockIdx.x, 7)*589824) + (floordiv((threadIdx.x_2 + 2432), 24)*4608)) + cse_var_2) + (floordiv(floormod((threadIdx.x_2 + 8), 24), 3)*9)) + cse_var_1) + floormod((threadIdx.x_2 + 2), 3))]
-          attr [IterVar(threadIdx.x_2, (nullptr), &quot;ThreadIndex&quot;, &quot;threadIdx.x&quot;)] &quot;thread_extent&quot; = 64;
-          kernel.shared_1[(threadIdx.x_2 + 2496)] = kernel_3[(((((((floordiv(blockIdx.x, 7)*589824) + (floordiv(threadIdx.x_2, 24)*4608)) + cse_var_2) + (floordiv(floormod(threadIdx.x_2, 24), 3)*9)) + cse_var_1) + floormod(threadIdx.x_2, 3)) + 479232)]
-          attr [IterVar(threadIdx.x_2, (nullptr), &quot;ThreadIndex&quot;, &quot;threadIdx.x&quot;)] &quot;thread_extent&quot; = 64;
-          kernel.shared_1[(threadIdx.x_2 + 2560)] = kernel_3[((((((floordiv(blockIdx.x, 7)*589824) + (floordiv((threadIdx.x_2 + 2560), 24)*4608)) + cse_var_2) + (floordiv(floormod((threadIdx.x_2 + 16), 24), 3)*9)) + cse_var_1) + floormod((threadIdx.x_2 + 1), 3))]
-          attr [IterVar(threadIdx.x_2, (nullptr), &quot;ThreadIndex&quot;, &quot;threadIdx.x&quot;)] &quot;thread_extent&quot; = 64;
-          kernel.shared_1[(threadIdx.x_2 + 2624)] = kernel_3[((((((floordiv(blockIdx.x, 7)*589824) + (floordiv((threadIdx.x_2 + 2624), 24)*4608)) + cse_var_2) + (floordiv(floormod((threadIdx.x_2 + 8), 24), 3)*9)) + cse_var_1) + floormod((threadIdx.x_2 + 2), 3))]
-          attr [IterVar(threadIdx.x_2, (nullptr), &quot;ThreadIndex&quot;, &quot;threadIdx.x&quot;)] &quot;thread_extent&quot; = 64;
-          kernel.shared_1[(threadIdx.x_2 + 2688)] = kernel_3[(((((((floordiv(blockIdx.x, 7)*589824) + (floordiv(threadIdx.x_2, 24)*4608)) + cse_var_2) + (floordiv(floormod(threadIdx.x_2, 24), 3)*9)) + cse_var_1) + floormod(threadIdx.x_2, 3)) + 516096)]
-          attr [IterVar(threadIdx.x_2, (nullptr), &quot;ThreadIndex&quot;, &quot;threadIdx.x&quot;)] &quot;thread_extent&quot; = 64;
-          kernel.shared_1[(threadIdx.x_2 + 2752)] = kernel_3[((((((floordiv(blockIdx.x, 7)*589824) + (floordiv((threadIdx.x_2 + 2752), 24)*4608)) + cse_var_2) + (floordiv(floormod((threadIdx.x_2 + 16), 24), 3)*9)) + cse_var_1) + floormod((threadIdx.x_2 + 1), 3))]
-          attr [IterVar(threadIdx.x_2, (nullptr), &quot;ThreadIndex&quot;, &quot;threadIdx.x&quot;)] &quot;thread_extent&quot; = 64;
-          kernel.shared_1[(threadIdx.x_2 + 2816)] = kernel_3[((((((floordiv(blockIdx.x, 7)*589824) + (floordiv((threadIdx.x_2 + 2816), 24)*4608)) + cse_var_2) + (floordiv(floormod((threadIdx.x_2 + 8), 24), 3)*9)) + cse_var_1) + floormod((threadIdx.x_2 + 2), 3))]
-          attr [IterVar(threadIdx.x_2, (nullptr), &quot;ThreadIndex&quot;, &quot;threadIdx.x&quot;)] &quot;thread_extent&quot; = 64;
-          kernel.shared_1[(threadIdx.x_2 + 2880)] = kernel_3[(((((((floordiv(blockIdx.x, 7)*589824) + (floordiv(threadIdx.x_2, 24)*4608)) + cse_var_2) + (floordiv(floormod(threadIdx.x_2, 24), 3)*9)) + cse_var_1) + floormod(threadIdx.x_2, 3)) + 552960)]
-          attr [IterVar(threadIdx.x_2, (nullptr), &quot;ThreadIndex&quot;, &quot;threadIdx.x&quot;)] &quot;thread_extent&quot; = 64;
-          kernel.shared_1[(threadIdx.x_2 + 2944)] = kernel_3[((((((floordiv(blockIdx.x, 7)*589824) + (floordiv((threadIdx.x_2 + 2944), 24)*4608)) + cse_var_2) + (floordiv(floormod((threadIdx.x_2 + 16), 24), 3)*9)) + cse_var_1) + floormod((threadIdx.x_2 + 1), 3))]
-          attr [IterVar(threadIdx.x_2, (nullptr), &quot;ThreadIndex&quot;, &quot;threadIdx.x&quot;)] &quot;thread_extent&quot; = 64;
-          kernel.shared_1[(threadIdx.x_2 + 3008)] = kernel_3[((((((floordiv(blockIdx.x, 7)*589824) + (floordiv((threadIdx.x_2 + 3008), 24)*4608)) + cse_var_2) + (floordiv(floormod((threadIdx.x_2 + 8), 24), 3)*9)) + cse_var_1) + floormod((threadIdx.x_2 + 2), 3))]
-          conv2d_nchw_1[0] = (conv2d_nchw_1[0] + (pad_temp.shared_1[0]*kernel.shared_1[(threadIdx.x*48)]))
-          conv2d_nchw_1[0] = (conv2d_nchw_1[0] + (pad_temp.shared_1[9]*kernel.shared_1[((threadIdx.x*48) + 3)]))
-          conv2d_nchw_1[1] = (conv2d_nchw_1[1] + (pad_temp.shared_1[1]*kernel.shared_1[(threadIdx.x*48)]))
-          conv2d_nchw_1[1] = (conv2d_nchw_1[1] + (pad_temp.shared_1[10]*kernel.shared_1[((threadIdx.x*48) + 3)]))
-          conv2d_nchw_1[2] = (conv2d_nchw_1[2] + (pad_temp.shared_1[2]*kernel.shared_1[(threadIdx.x*48)]))
-          conv2d_nchw_1[2] = (conv2d_nchw_1[2] + (pad_temp.shared_1[11]*kernel.shared_1[((threadIdx.x*48) + 3)]))
-          conv2d_nchw_1[3] = (conv2d_nchw_1[3] + (pad_temp.shared_1[3]*kernel.shared_1[(threadIdx.x*48)]))
-          conv2d_nchw_1[3] = (conv2d_nchw_1[3] + (pad_temp.shared_1[12]*kernel.shared_1[((threadIdx.x*48) + 3)]))
-          conv2d_nchw_1[4] = (conv2d_nchw_1[4] + (pad_temp.shared_1[4]*kernel.shared_1[(threadIdx.x*48)]))
-          conv2d_nchw_1[4] = (conv2d_nchw_1[4] + (pad_temp.shared_1[13]*kernel.shared_1[((threadIdx.x*48) + 3)]))
-          conv2d_nchw_1[5] = (conv2d_nchw_1[5] + (pad_temp.shared_1[5]*kernel.shared_1[(threadIdx.x*48)]))
-          conv2d_nchw_1[5] = (conv2d_nchw_1[5] + (pad_temp.shared_1[14]*kernel.shared_1[((threadIdx.x*48) + 3)]))
-          conv2d_nchw_1[6] = (conv2d_nchw_1[6] + (pad_temp.shared_1[6]*kernel.shared_1[(threadIdx.x*48)]))
-          conv2d_nchw_1[6] = (conv2d_nchw_1[6] + (pad_temp.shared_1[15]*kernel.shared_1[((threadIdx.x*48) + 3)]))
-          conv2d_nchw_1[7] = (conv2d_nchw_1[7] + (pad_temp.shared_1[0]*kernel.shared_1[((threadIdx.x*48) + 24)]))
-          conv2d_nchw_1[7] = (conv2d_nchw_1[7] + (pad_temp.shared_1[9]*kernel.shared_1[((threadIdx.x*48) + 27)]))
-          conv2d_nchw_1[8] = (conv2d_nchw_1[8] + (pad_temp.shared_1[1]*kernel.shared_1[((threadIdx.x*48) + 24)]))
-          conv2d_nchw_1[8] = (conv2d_nchw_1[8] + (pad_temp.shared_1[10]*kernel.shared_1[((threadIdx.x*48) + 27)]))
-          conv2d_nchw_1[9] = (conv2d_nchw_1[9] + (pad_temp.shared_1[2]*kernel.shared_1[((threadIdx.x*48) + 24)]))
-          conv2d_nchw_1[9] = (conv2d_nchw_1[9] + (pad_temp.shared_1[11]*kernel.shared_1[((threadIdx.x*48) + 27)]))
-          conv2d_nchw_1[10] = (conv2d_nchw_1[10] + (pad_temp.shared_1[3]*kernel.shared_1[((threadIdx.x*48) + 24)]))
-          conv2d_nchw_1[10] = (conv2d_nchw_1[10] + (pad_temp.shared_1[12]*kernel.shared_1[((threadIdx.x*48) + 27)]))
-          conv2d_nchw_1[11] = (conv2d_nchw_1[11] + (pad_temp.shared_1[4]*kernel.shared_1[((threadIdx.x*48) + 24)]))
-          conv2d_nchw_1[11] = (conv2d_nchw_1[11] + (pad_temp.shared_1[13]*kernel.shared_1[((threadIdx.x*48) + 27)]))
-          conv2d_nchw_1[12] = (conv2d_nchw_1[12] + (pad_temp.shared_1[5]*kernel.shared_1[((threadIdx.x*48) + 24)]))
-          conv2d_nchw_1[12] = (conv2d_nchw_1[12] + (pad_temp.shared_1[14]*kernel.shared_1[((threadIdx.x*48) + 27)]))
-          conv2d_nchw_1[13] = (conv2d_nchw_1[13] + (pad_temp.shared_1[6]*kernel.shared_1[((threadIdx.x*48) + 24)]))
-          conv2d_nchw_1[13] = (conv2d_nchw_1[13] + (pad_temp.shared_1[15]*kernel.shared_1[((threadIdx.x*48) + 27)]))
-          conv2d_nchw_1[0] = (conv2d_nchw_1[0] + (pad_temp.shared_1[1]*kernel.shared_1[((threadIdx.x*48) + 1)]))
-          conv2d_nchw_1[0] = (conv2d_nchw_1[0] + (pad_temp.shared_1[10]*kernel.shared_1[((threadIdx.x*48) + 4)]))
-          conv2d_nchw_1[1] = (conv2d_nchw_1[1] + (pad_temp.shared_1[2]*kernel.shared_1[((threadIdx.x*48) + 1)]))
-          conv2d_nchw_1[1] = (conv2d_nchw_1[1] + (pad_temp.shared_1[11]*kernel.shared_1[((threadIdx.x*48) + 4)]))
-          conv2d_nchw_1[2] = (conv2d_nchw_1[2] + (pad_temp.shared_1[3]*kernel.shared_1[((threadIdx.x*48) + 1)]))
-          conv2d_nchw_1[2] = (conv2d_nchw_1[2] + (pad_temp.shared_1[12]*kernel.shared_1[((threadIdx.x*48) + 4)]))
-          conv2d_nchw_1[3] = (conv2d_nchw_1[3] + (pad_temp.shared_1[4]*kernel.shared_1[((threadIdx.x*48) + 1)]))
-          conv2d_nchw_1[3] = (conv2d_nchw_1[3] + (pad_temp.shared_1[13]*kernel.shared_1[((threadIdx.x*48) + 4)]))
-          conv2d_nchw_1[4] = (conv2d_nchw_1[4] + (pad_temp.shared_1[5]*kernel.shared_1[((threadIdx.x*48) + 1)]))
-          conv2d_nchw_1[4] = (conv2d_nchw_1[4] + (pad_temp.shared_1[14]*kernel.shared_1[((threadIdx.x*48) + 4)]))
-          conv2d_nchw_1[5] = (conv2d_nchw_1[5] + (pad_temp.shared_1[6]*kernel.shared_1[((threadIdx.x*48) + 1)]))
-          conv2d_nchw_1[5] = (conv2d_nchw_1[5] + (pad_temp.shared_1[15]*kernel.shared_1[((threadIdx.x*48) + 4)]))
-          conv2d_nchw_1[6] = (conv2d_nchw_1[6] + (pad_temp.shared_1[7]*kernel.shared_1[((threadIdx.x*48) + 1)]))
-          conv2d_nchw_1[6] = (conv2d_nchw_1[6] + (pad_temp.shared_1[16]*kernel.shared_1[((threadIdx.x*48) + 4)]))
-          conv2d_nchw_1[7] = (conv2d_nchw_1[7] + (pad_temp.shared_1[1]*kernel.shared_1[((threadIdx.x*48) + 25)]))
-          conv2d_nchw_1[7] = (conv2d_nchw_1[7] + (pad_temp.shared_1[10]*kernel.shared_1[((threadIdx.x*48) + 28)]))
-          conv2d_nchw_1[8] = (conv2d_nchw_1[8] + (pad_temp.shared_1[2]*kernel.shared_1[((threadIdx.x*48) + 25)]))
-          conv2d_nchw_1[8] = (conv2d_nchw_1[8] + (pad_temp.shared_1[11]*kernel.shared_1[((threadIdx.x*48) + 28)]))
-          conv2d_nchw_1[9] = (conv2d_nchw_1[9] + (pad_temp.shared_1[3]*kernel.shared_1[((threadIdx.x*48) + 25)]))
-          conv2d_nchw_1[9] = (conv2d_nchw_1[9] + (pad_temp.shared_1[12]*kernel.shared_1[((threadIdx.x*48) + 28)]))
-          conv2d_nchw_1[10] = (conv2d_nchw_1[10] + (pad_temp.shared_1[4]*kernel.shared_1[((threadIdx.x*48) + 25)]))
-          conv2d_nchw_1[10] = (conv2d_nchw_1[10] + (pad_temp.shared_1[13]*kernel.shared_1[((threadIdx.x*48) + 28)]))
-          conv2d_nchw_1[11] = (conv2d_nchw_1[11] + (pad_temp.shared_1[5]*kernel.shared_1[((threadIdx.x*48) + 25)]))
-          conv2d_nchw_1[11] = (conv2d_nchw_1[11] + (pad_temp.shared_1[14]*kernel.shared_1[((threadIdx.x*48) + 28)]))
-          conv2d_nchw_1[12] = (conv2d_nchw_1[12] + (pad_temp.shared_1[6]*kernel.shared_1[((threadIdx.x*48) + 25)]))
-          conv2d_nchw_1[12] = (conv2d_nchw_1[12] + (pad_temp.shared_1[15]*kernel.shared_1[((threadIdx.x*48) + 28)]))
-          conv2d_nchw_1[13] = (conv2d_nchw_1[13] + (pad_temp.shared_1[7]*kernel.shared_1[((threadIdx.x*48) + 25)]))
-          conv2d_nchw_1[13] = (conv2d_nchw_1[13] + (pad_temp.shared_1[16]*kernel.shared_1[((threadIdx.x*48) + 28)]))
-          conv2d_nchw_1[0] = (conv2d_nchw_1[0] + (pad_temp.shared_1[2]*kernel.shared_1[((threadIdx.x*48) + 2)]))
-          conv2d_nchw_1[0] = (conv2d_nchw_1[0] + (pad_temp.shared_1[11]*kernel.shared_1[((threadIdx.x*48) + 5)]))
-          conv2d_nchw_1[1] = (conv2d_nchw_1[1] + (pad_temp.shared_1[3]*kernel.shared_1[((threadIdx.x*48) + 2)]))
-          conv2d_nchw_1[1] = (conv2d_nchw_1[1] + (pad_temp.shared_1[12]*kernel.shared_1[((threadIdx.x*48) + 5)]))
-          conv2d_nchw_1[2] = (conv2d_nchw_1[2] + (pad_temp.shared_1[4]*kernel.shared_1[((threadIdx.x*48) + 2)]))
-          conv2d_nchw_1[2] = (conv2d_nchw_1[2] + (pad_temp.shared_1[13]*kernel.shared_1[((threadIdx.x*48) + 5)]))
-          conv2d_nchw_1[3] = (conv2d_nchw_1[3] + (pad_temp.shared_1[5]*kernel.shared_1[((threadIdx.x*48) + 2)]))
-          conv2d_nchw_1[3] = (conv2d_nchw_1[3] + (pad_temp.shared_1[14]*kernel.shared_1[((threadIdx.x*48) + 5)]))
-          conv2d_nchw_1[4] = (conv2d_nchw_1[4] + (pad_temp.shared_1[6]*kernel.shared_1[((threadIdx.x*48) + 2)]))
-          conv2d_nchw_1[4] = (conv2d_nchw_1[4] + (pad_temp.shared_1[15]*kernel.shared_1[((threadIdx.x*48) + 5)]))
-          conv2d_nchw_1[5] = (conv2d_nchw_1[5] + (pad_temp.shared_1[7]*kernel.shared_1[((threadIdx.x*48) + 2)]))
-          conv2d_nchw_1[5] = (conv2d_nchw_1[5] + (pad_temp.shared_1[16]*kernel.shared_1[((threadIdx.x*48) + 5)]))
-          conv2d_nchw_1[6] = (conv2d_nchw_1[6] + (pad_temp.shared_1[8]*kernel.shared_1[((threadIdx.x*48) + 2)]))
-          conv2d_nchw_1[6] = (conv2d_nchw_1[6] + (pad_temp.shared_1[17]*kernel.shared_1[((threadIdx.x*48) + 5)]))
-          conv2d_nchw_1[7] = (conv2d_nchw_1[7] + (pad_temp.shared_1[2]*kernel.shared_1[((threadIdx.x*48) + 26)]))
-          conv2d_nchw_1[7] = (conv2d_nchw_1[7] + (pad_temp.shared_1[11]*kernel.shared_1[((threadIdx.x*48) + 29)]))
-          conv2d_nchw_1[8] = (conv2d_nchw_1[8] + (pad_temp.shared_1[3]*kernel.shared_1[((threadIdx.x*48) + 26)]))
-          conv2d_nchw_1[8] = (conv2d_nchw_1[8] + (pad_temp.shared_1[12]*kernel.shared_1[((threadIdx.x*48) + 29)]))
-          conv2d_nchw_1[9] = (conv2d_nchw_1[9] + (pad_temp.shared_1[4]*kernel.shared_1[((threadIdx.x*48) + 26)]))
-          conv2d_nchw_1[9] = (conv2d_nchw_1[9] + (pad_temp.shared_1[13]*kernel.shared_1[((threadIdx.x*48) + 29)]))
-          conv2d_nchw_1[10] = (conv2d_nchw_1[10] + (pad_temp.shared_1[5]*kernel.shared_1[((threadIdx.x*48) + 26)]))
-          conv2d_nchw_1[10] = (conv2d_nchw_1[10] + (pad_temp.shared_1[14]*kernel.shared_1[((threadIdx.x*48) + 29)]))
-          conv2d_nchw_1[11] = (conv2d_nchw_1[11] + (pad_temp.shared_1[6]*kernel.shared_1[((threadIdx.x*48) + 26)]))
-          conv2d_nchw_1[11] = (conv2d_nchw_1[11] + (pad_temp.shared_1[15]*kernel.shared_1[((threadIdx.x*48) + 29)]))
-          conv2d_nchw_1[12] = (conv2d_nchw_1[12] + (pad_temp.shared_1[7]*kernel.shared_1[((threadIdx.x*48) + 26)]))
-          conv2d_nchw_1[12] = (conv2d_nchw_1[12] + (pad_temp.shared_1[16]*kernel.shared_1[((threadIdx.x*48) + 29)]))
-          conv2d_nchw_1[13] = (conv2d_nchw_1[13] + (pad_temp.shared_1[8]*kernel.shared_1[((threadIdx.x*48) + 26)]))
-          conv2d_nchw_1[13] = (conv2d_nchw_1[13] + (pad_temp.shared_1[17]*kernel.shared_1[((threadIdx.x*48) + 29)]))
-          conv2d_nchw_1[0] = (conv2d_nchw_1[0] + (pad_temp.shared_1[18]*kernel.shared_1[((threadIdx.x*48) + 6)]))
-          conv2d_nchw_1[0] = (conv2d_nchw_1[0] + (pad_temp.shared_1[27]*kernel.shared_1[((threadIdx.x*48) + 9)]))
-          conv2d_nchw_1[1] = (conv2d_nchw_1[1] + (pad_temp.shared_1[19]*kernel.shared_1[((threadIdx.x*48) + 6)]))
-          conv2d_nchw_1[1] = (conv2d_nchw_1[1] + (pad_temp.shared_1[28]*kernel.shared_1[((threadIdx.x*48) + 9)]))
-          conv2d_nchw_1[2] = (conv2d_nchw_1[2] + (pad_temp.shared_1[20]*kernel.shared_1[((threadIdx.x*48) + 6)]))
-          conv2d_nchw_1[2] = (conv2d_nchw_1[2] + (pad_temp.shared_1[29]*kernel.shared_1[((threadIdx.x*48) + 9)]))
-          conv2d_nchw_1[3] = (conv2d_nchw_1[3] + (pad_temp.shared_1[21]*kernel.shared_1[((threadIdx.x*48) + 6)]))
-          conv2d_nchw_1[3] = (conv2d_nchw_1[3] + (pad_temp.shared_1[30]*kernel.shared_1[((threadIdx.x*48) + 9)]))
-          conv2d_nchw_1[4] = (conv2d_nchw_1[4] + (pad_temp.shared_1[22]*kernel.shared_1[((threadIdx.x*48) + 6)]))
-          conv2d_nchw_1[4] = (conv2d_nchw_1[4] + (pad_temp.shared_1[31]*kernel.shared_1[((threadIdx.x*48) + 9)]))
-          conv2d_nchw_1[5] = (conv2d_nchw_1[5] + (pad_temp.shared_1[23]*kernel.shared_1[((threadIdx.x*48) + 6)]))
-          conv2d_nchw_1[5] = (conv2d_nchw_1[5] + (pad_temp.shared_1[32]*kernel.shared_1[((threadIdx.x*48) + 9)]))
-          conv2d_nchw_1[6] = (conv2d_nchw_1[6] + (pad_temp.shared_1[24]*kernel.shared_1[((threadIdx.x*48) + 6)]))
-          conv2d_nchw_1[6] = (conv2d_nchw_1[6] + (pad_temp.shared_1[33]*kernel.shared_1[((threadIdx.x*48) + 9)]))
-          conv2d_nchw_1[7] = (conv2d_nchw_1[7] + (pad_temp.shared_1[18]*kernel.shared_1[((threadIdx.x*48) + 30)]))
-          conv2d_nchw_1[7] = (conv2d_nchw_1[7] + (pad_temp.shared_1[27]*kernel.shared_1[((threadIdx.x*48) + 33)]))
-          conv2d_nchw_1[8] = (conv2d_nchw_1[8] + (pad_temp.shared_1[19]*kernel.shared_1[((threadIdx.x*48) + 30)]))
-          conv2d_nchw_1[8] = (conv2d_nchw_1[8] + (pad_temp.shared_1[28]*kernel.shared_1[((threadIdx.x*48) + 33)]))
-          conv2d_nchw_1[9] = (conv2d_nchw_1[9] + (pad_temp.shared_1[20]*kernel.shared_1[((threadIdx.x*48) + 30)]))
-          conv2d_nchw_1[9] = (conv2d_nchw_1[9] + (pad_temp.shared_1[29]*kernel.shared_1[((threadIdx.x*48) + 33)]))
-          conv2d_nchw_1[10] = (conv2d_nchw_1[10] + (pad_temp.shared_1[21]*kernel.shared_1[((threadIdx.x*48) + 30)]))
-          conv2d_nchw_1[10] = (conv2d_nchw_1[10] + (pad_temp.shared_1[30]*kernel.shared_1[((threadIdx.x*48) + 33)]))
-          conv2d_nchw_1[11] = (conv2d_nchw_1[11] + (pad_temp.shared_1[22]*kernel.shared_1[((threadIdx.x*48) + 30)]))
-          conv2d_nchw_1[11] = (conv2d_nchw_1[11] + (pad_temp.shared_1[31]*kernel.shared_1[((threadIdx.x*48) + 33)]))
-          conv2d_nchw_1[12] = (conv2d_nchw_1[12] + (pad_temp.shared_1[23]*kernel.shared_1[((threadIdx.x*48) + 30)]))
-          conv2d_nchw_1[12] = (conv2d_nchw_1[12] + (pad_temp.shared_1[32]*kernel.shared_1[((threadIdx.x*48) + 33)]))
-          conv2d_nchw_1[13] = (conv2d_nchw_1[13] + (pad_temp.shared_1[24]*kernel.shared_1[((threadIdx.x*48) + 30)]))
-          conv2d_nchw_1[13] = (conv2d_nchw_1[13] + (pad_temp.shared_1[33]*kernel.shared_1[((threadIdx.x*48) + 33)]))
-          conv2d_nchw_1[0] = (conv2d_nchw_1[0] + (pad_temp.shared_1[19]*kernel.shared_1[((threadIdx.x*48) + 7)]))
-          conv2d_nchw_1[0] = (conv2d_nchw_1[0] + (pad_temp.shared_1[28]*kernel.shared_1[((threadIdx.x*48) + 10)]))
-          conv2d_nchw_1[1] = (conv2d_nchw_1[1] + (pad_temp.shared_1[20]*kernel.shared_1[((threadIdx.x*48) + 7)]))
-          conv2d_nchw_1[1] = (conv2d_nchw_1[1] + (pad_temp.shared_1[29]*kernel.shared_1[((threadIdx.x*48) + 10)]))
-          conv2d_nchw_1[2] = (conv2d_nchw_1[2] + (pad_temp.shared_1[21]*kernel.shared_1[((threadIdx.x*48) + 7)]))
-          conv2d_nchw_1[2] = (conv2d_nchw_1[2] + (pad_temp.shared_1[30]*kernel.shared_1[((threadIdx.x*48) + 10)]))
-          conv2d_nchw_1[3] = (conv2d_nchw_1[3] + (pad_temp.shared_1[22]*kernel.shared_1[((threadIdx.x*48) + 7)]))
-          conv2d_nchw_1[3] = (conv2d_nchw_1[3] + (pad_temp.shared_1[31]*kernel.shared_1[((threadIdx.x*48) + 10)]))
-          conv2d_nchw_1[4] = (conv2d_nchw_1[4] + (pad_temp.shared_1[23]*kernel.shared_1[((threadIdx.x*48) + 7)]))
-          conv2d_nchw_1[4] = (conv2d_nchw_1[4] + (pad_temp.shared_1[32]*kernel.shared_1[((threadIdx.x*48) + 10)]))
-          conv2d_nchw_1[5] = (conv2d_nchw_1[5] + (pad_temp.shared_1[24]*kernel.shared_1[((threadIdx.x*48) + 7)]))
-          conv2d_nchw_1[5] = (conv2d_nchw_1[5] + (pad_temp.shared_1[33]*kernel.shared_1[((threadIdx.x*48) + 10)]))
-          conv2d_nchw_1[6] = (conv2d_nchw_1[6] + (pad_temp.shared_1[25]*kernel.shared_1[((threadIdx.x*48) + 7)]))
-          conv2d_nchw_1[6] = (conv2d_nchw_1[6] + (pad_temp.shared_1[34]*kernel.shared_1[((threadIdx.x*48) + 10)]))
-          conv2d_nchw_1[7] = (conv2d_nchw_1[7] + (pad_temp.shared_1[19]*kernel.shared_1[((threadIdx.x*48) + 31)]))
-          conv2d_nchw_1[7] = (conv2d_nchw_1[7] + (pad_temp.shared_1[28]*kernel.shared_1[((threadIdx.x*48) + 34)]))
-          conv2d_nchw_1[8] = (conv2d_nchw_1[8] + (pad_temp.shared_1[20]*kernel.shared_1[((threadIdx.x*48) + 31)]))
-          conv2d_nchw_1[8] = (conv2d_nchw_1[8] + (pad_temp.shared_1[29]*kernel.shared_1[((threadIdx.x*48) + 34)]))
-          conv2d_nchw_1[9] = (conv2d_nchw_1[9] + (pad_temp.shared_1[21]*kernel.shared_1[((threadIdx.x*48) + 31)]))
-          conv2d_nchw_1[9] = (conv2d_nchw_1[9] + (pad_temp.shared_1[30]*kernel.shared_1[((threadIdx.x*48) + 34)]))
-          conv2d_nchw_1[10] = (conv2d_nchw_1[10] + (pad_temp.shared_1[22]*kernel.shared_1[((threadIdx.x*48) + 31)]))
-          conv2d_nchw_1[10] = (conv2d_nchw_1[10] + (pad_temp.shared_1[31]*kernel.shared_1[((threadIdx.x*48) + 34)]))
-          conv2d_nchw_1[11] = (conv2d_nchw_1[11] + (pad_temp.shared_1[23]*kernel.shared_1[((threadIdx.x*48) + 31)]))
-          conv2d_nchw_1[11] = (conv2d_nchw_1[11] + (pad_temp.shared_1[32]*kernel.shared_1[((threadIdx.x*48) + 34)]))
-          conv2d_nchw_1[12] = (conv2d_nchw_1[12] + (pad_temp.shared_1[24]*kernel.shared_1[((threadIdx.x*48) + 31)]))
-          conv2d_nchw_1[12] = (conv2d_nchw_1[12] + (pad_temp.shared_1[33]*kernel.shared_1[((threadIdx.x*48) + 34)]))
-          conv2d_nchw_1[13] = (conv2d_nchw_1[13] + (pad_temp.shared_1[25]*kernel.shared_1[((threadIdx.x*48) + 31)]))
-          conv2d_nchw_1[13] = (conv2d_nchw_1[13] + (pad_temp.shared_1[34]*kernel.shared_1[((threadIdx.x*48) + 34)]))
-          conv2d_nchw_1[0] = (conv2d_nchw_1[0] + (pad_temp.shared_1[20]*kernel.shared_1[((threadIdx.x*48) + 8)]))
-          conv2d_nchw_1[0] = (conv2d_nchw_1[0] + (pad_temp.shared_1[29]*kernel.shared_1[((threadIdx.x*48) + 11)]))
-          conv2d_nchw_1[1] = (conv2d_nchw_1[1] + (pad_temp.shared_1[21]*kernel.shared_1[((threadIdx.x*48) + 8)]))
-          conv2d_nchw_1[1] = (conv2d_nchw_1[1] + (pad_temp.shared_1[30]*kernel.shared_1[((threadIdx.x*48) + 11)]))
-          conv2d_nchw_1[2] = (conv2d_nchw_1[2] + (pad_temp.shared_1[22]*kernel.shared_1[((threadIdx.x*48) + 8)]))
-          conv2d_nchw_1[2] = (conv2d_nchw_1[2] + (pad_temp.shared_1[31]*kernel.shared_1[((threadIdx.x*48) + 11)]))
-          conv2d_nchw_1[3] = (conv2d_nchw_1[3] + (pad_temp.shared_1[23]*kernel.shared_1[((threadIdx.x*48) + 8)]))
-          conv2d_nchw_1[3] = (conv2d_nchw_1[3] + (pad_temp.shared_1[32]*kernel.shared_1[((threadIdx.x*48) + 11)]))
-          conv2d_nchw_1[4] = (conv2d_nchw_1[4] + (pad_temp.shared_1[24]*kernel.shared_1[((threadIdx.x*48) + 8)]))
-          conv2d_nchw_1[4] = (conv2d_nchw_1[4] + (pad_temp.shared_1[33]*kernel.shared_1[((threadIdx.x*48) + 11)]))
-          conv2d_nchw_1[5] = (conv2d_nchw_1[5] + (pad_temp.shared_1[25]*kernel.shared_1[((threadIdx.x*48) + 8)]))
-          conv2d_nchw_1[5] = (conv2d_nchw_1[5] + (pad_temp.shared_1[34]*kernel.shared_1[((threadIdx.x*48) + 11)]))
-          conv2d_nchw_1[6] = (conv2d_nchw_1[6] + (pad_temp.shared_1[26]*kernel.shared_1[((threadIdx.x*48) + 8)]))
-          conv2d_nchw_1[6] = (conv2d_nchw_1[6] + (pad_temp.shared_1[35]*kernel.shared_1[((threadIdx.x*48) + 11)]))
-          conv2d_nchw_1[7] = (conv2d_nchw_1[7] + (pad_temp.shared_1[20]*kernel.shared_1[((threadIdx.x*48) + 32)]))
-          conv2d_nchw_1[7] = (conv2d_nchw_1[7] + (pad_temp.shared_1[29]*kernel.shared_1[((threadIdx.x*48) + 35)]))
-          conv2d_nchw_1[8] = (conv2d_nchw_1[8] + (pad_temp.shared_1[21]*kernel.shared_1[((threadIdx.x*48) + 32)]))
-          conv2d_nchw_1[8] = (conv2d_nchw_1[8] + (pad_temp.shared_1[30]*kernel.shared_1[((threadIdx.x*48) + 35)]))
-          conv2d_nchw_1[9] = (conv2d_nchw_1[9] + (pad_temp.shared_1[22]*kernel.shared_1[((threadIdx.x*48) + 32)]))
-          conv2d_nchw_1[9] = (conv2d_nchw_1[9] + (pad_temp.shared_1[31]*kernel.shared_1[((threadIdx.x*48) + 35)]))
-          conv2d_nchw_1[10] = (conv2d_nchw_1[10] + (pad_temp.shared_1[23]*kernel.shared_1[((threadIdx.x*48) + 32)]))
-          conv2d_nchw_1[10] = (conv2d_nchw_1[10] + (pad_temp.shared_1[32]*kernel.shared_1[((threadIdx.x*48) + 35)]))
-          conv2d_nchw_1[11] = (conv2d_nchw_1[11] + (pad_temp.shared_1[24]*kernel.shared_1[((threadIdx.x*48) + 32)]))
-          conv2d_nchw_1[11] = (conv2d_nchw_1[11] + (pad_temp.shared_1[33]*kernel.shared_1[((threadIdx.x*48) + 35)]))
-          conv2d_nchw_1[12] = (conv2d_nchw_1[12] + (pad_temp.shared_1[25]*kernel.shared_1[((threadIdx.x*48) + 32)]))
-          conv2d_nchw_1[12] = (conv2d_nchw_1[12] + (pad_temp.shared_1[34]*kernel.shared_1[((threadIdx.x*48) + 35)]))
-          conv2d_nchw_1[13] = (conv2d_nchw_1[13] + (pad_temp.shared_1[26]*kernel.shared_1[((threadIdx.x*48) + 32)]))
-          conv2d_nchw_1[13] = (conv2d_nchw_1[13] + (pad_temp.shared_1[35]*kernel.shared_1[((threadIdx.x*48) + 35)]))
-          conv2d_nchw_1[0] = (conv2d_nchw_1[0] + (pad_temp.shared_1[36]*kernel.shared_1[((threadIdx.x*48) + 12)]))
-          conv2d_nchw_1[0] = (conv2d_nchw_1[0] + (pad_temp.shared_1[45]*kernel.shared_1[((threadIdx.x*48) + 15)]))
-          conv2d_nchw_1[1] = (conv2d_nchw_1[1] + (pad_temp.shared_1[37]*kernel.shared_1[((threadIdx.x*48) + 12)]))
-          conv2d_nchw_1[1] = (conv2d_nchw_1[1] + (pad_temp.shared_1[46]*kernel.shared_1[((threadIdx.x*48) + 15)]))
-          conv2d_nchw_1[2] = (conv2d_nchw_1[2] + (pad_temp.shared_1[38]*kernel.shared_1[((threadIdx.x*48) + 12)]))
-          conv2d_nchw_1[2] = (conv2d_nchw_1[2] + (pad_temp.shared_1[47]*kernel.shared_1[((threadIdx.x*48) + 15)]))
-          conv2d_nchw_1[3] = (conv2d_nchw_1[3] + (pad_temp.shared_1[39]*kernel.shared_1[((threadIdx.x*48) + 12)]))
-          conv2d_nchw_1[3] = (conv2d_nchw_1[3] + (pad_temp.shared_1[48]*kernel.shared_1[((threadIdx.x*48) + 15)]))
-          conv2d_nchw_1[4] = (conv2d_nchw_1[4] + (pad_temp.shared_1[40]*kernel.shared_1[((threadIdx.x*48) + 12)]))
-          conv2d_nchw_1[4] = (conv2d_nchw_1[4] + (pad_temp.shared_1[49]*kernel.shared_1[((threadIdx.x*48) + 15)]))
-          conv2d_nchw_1[5] = (conv2d_nchw_1[5] + (pad_temp.shared_1[41]*kernel.shared_1[((threadIdx.x*48) + 12)]))
-          conv2d_nchw_1[5] = (conv2d_nchw_1[5] + (pad_temp.shared_1[50]*kernel.shared_1[((threadIdx.x*48) + 15)]))
-          conv2d_nchw_1[6] = (conv2d_nchw_1[6] + (pad_temp.shared_1[42]*kernel.shared_1[((threadIdx.x*48) + 12)]))
-          conv2d_nchw_1[6] = (conv2d_nchw_1[6] + (pad_temp.shared_1[51]*kernel.shared_1[((threadIdx.x*48) + 15)]))
-          conv2d_nchw_1[7] = (conv2d_nchw_1[7] + (pad_temp.shared_1[36]*kernel.shared_1[((threadIdx.x*48) + 36)]))
-          conv2d_nchw_1[7] = (conv2d_nchw_1[7] + (pad_temp.shared_1[45]*kernel.shared_1[((threadIdx.x*48) + 39)]))
-          conv2d_nchw_1[8] = (conv2d_nchw_1[8] + (pad_temp.shared_1[37]*kernel.shared_1[((threadIdx.x*48) + 36)]))
-          conv2d_nchw_1[8] = (conv2d_nchw_1[8] + (pad_temp.shared_1[46]*kernel.shared_1[((threadIdx.x*48) + 39)]))
-          conv2d_nchw_1[9] = (conv2d_nchw_1[9] + (pad_temp.shared_1[38]*kernel.shared_1[((threadIdx.x*48) + 36)]))
-          conv2d_nchw_1[9] = (conv2d_nchw_1[9] + (pad_temp.shared_1[47]*kernel.shared_1[((threadIdx.x*48) + 39)]))
-          conv2d_nchw_1[10] = (conv2d_nchw_1[10] + (pad_temp.shared_1[39]*kernel.shared_1[((threadIdx.x*48) + 36)]))
-          conv2d_nchw_1[10] = (conv2d_nchw_1[10] + (pad_temp.shared_1[48]*kernel.shared_1[((threadIdx.x*48) + 39)]))
-          conv2d_nchw_1[11] = (conv2d_nchw_1[11] + (pad_temp.shared_1[40]*kernel.shared_1[((threadIdx.x*48) + 36)]))
-          conv2d_nchw_1[11] = (conv2d_nchw_1[11] + (pad_temp.shared_1[49]*kernel.shared_1[((threadIdx.x*48) + 39)]))
-          conv2d_nchw_1[12] = (conv2d_nchw_1[12] + (pad_temp.shared_1[41]*kernel.shared_1[((threadIdx.x*48) + 36)]))
-          conv2d_nchw_1[12] = (conv2d_nchw_1[12] + (pad_temp.shared_1[50]*kernel.shared_1[((threadIdx.x*48) + 39)]))
-          conv2d_nchw_1[13] = (conv2d_nchw_1[13] + (pad_temp.shared_1[42]*kernel.shared_1[((threadIdx.x*48) + 36)]))
-          conv2d_nchw_1[13] = (conv2d_nchw_1[13] + (pad_temp.shared_1[51]*kernel.shared_1[((threadIdx.x*48) + 39)]))
-          conv2d_nchw_1[0] = (conv2d_nchw_1[0] + (pad_temp.shared_1[37]*kernel.shared_1[((threadIdx.x*48) + 13)]))
-          conv2d_nchw_1[0] = (conv2d_nchw_1[0] + (pad_temp.shared_1[46]*kernel.shared_1[((threadIdx.x*48) + 16)]))
-          conv2d_nchw_1[1] = (conv2d_nchw_1[1] + (pad_temp.shared_1[38]*kernel.shared_1[((threadIdx.x*48) + 13)]))
-          conv2d_nchw_1[1] = (conv2d_nchw_1[1] + (pad_temp.shared_1[47]*kernel.shared_1[((threadIdx.x*48) + 16)]))
-          conv2d_nchw_1[2] = (conv2d_nchw_1[2] + (pad_temp.shared_1[39]*kernel.shared_1[((threadIdx.x*48) + 13)]))
-          conv2d_nchw_1[2] = (conv2d_nchw_1[2] + (pad_temp.shared_1[48]*kernel.shared_1[((threadIdx.x*48) + 16)]))
-          conv2d_nchw_1[3] = (conv2d_nchw_1[3] + (pad_temp.shared_1[40]*kernel.shared_1[((threadIdx.x*48) + 13)]))
-          conv2d_nchw_1[3] = (conv2d_nchw_1[3] + (pad_temp.shared_1[49]*kernel.shared_1[((threadIdx.x*48) + 16)]))
-          conv2d_nchw_1[4] = (conv2d_nchw_1[4] + (pad_temp.shared_1[41]*kernel.shared_1[((threadIdx.x*48) + 13)]))
-          conv2d_nchw_1[4] = (conv2d_nchw_1[4] + (pad_temp.shared_1[50]*kernel.shared_1[((threadIdx.x*48) + 16)]))
-          conv2d_nchw_1[5] = (conv2d_nchw_1[5] + (pad_temp.shared_1[42]*kernel.shared_1[((threadIdx.x*48) + 13)]))
-          conv2d_nchw_1[5] = (conv2d_nchw_1[5] + (pad_temp.shared_1[51]*kernel.shared_1[((threadIdx.x*48) + 16)]))
-          conv2d_nchw_1[6] = (conv2d_nchw_1[6] + (pad_temp.shared_1[43]*kernel.shared_1[((threadIdx.x*48) + 13)]))
-          conv2d_nchw_1[6] = (conv2d_nchw_1[6] + (pad_temp.shared_1[52]*kernel.shared_1[((threadIdx.x*48) + 16)]))
-          conv2d_nchw_1[7] = (conv2d_nchw_1[7] + (pad_temp.shared_1[37]*kernel.shared_1[((threadIdx.x*48) + 37)]))
-          conv2d_nchw_1[7] = (conv2d_nchw_1[7] + (pad_temp.shared_1[46]*kernel.shared_1[((threadIdx.x*48) + 40)]))
-          conv2d_nchw_1[8] = (conv2d_nchw_1[8] + (pad_temp.shared_1[38]*kernel.shared_1[((threadIdx.x*48) + 37)]))
-          conv2d_nchw_1[8] = (conv2d_nchw_1[8] + (pad_temp.shared_1[47]*kernel.shared_1[((threadIdx.x*48) + 40)]))
-          conv2d_nchw_1[9] = (conv2d_nchw_1[9] + (pad_temp.shared_1[39]*kernel.shared_1[((threadIdx.x*48) + 37)]))
-          conv2d_nchw_1[9] = (conv2d_nchw_1[9] + (pad_temp.shared_1[48]*kernel.shared_1[((threadIdx.x*48) + 40)]))
-          conv2d_nchw_1[10] = (conv2d_nchw_1[10] + (pad_temp.shared_1[40]*kernel.shared_1[((threadIdx.x*48) + 37)]))
-          conv2d_nchw_1[10] = (conv2d_nchw_1[10] + (pad_temp.shared_1[49]*kernel.shared_1[((threadIdx.x*48) + 40)]))
-          conv2d_nchw_1[11] = (conv2d_nchw_1[11] + (pad_temp.shared_1[41]*kernel.shared_1[((threadIdx.x*48) + 37)]))
-          conv2d_nchw_1[11] = (conv2d_nchw_1[11] + (pad_temp.shared_1[50]*kernel.shared_1[((threadIdx.x*48) + 40)]))
-          conv2d_nchw_1[12] = (conv2d_nchw_1[12] + (pad_temp.shared_1[42]*kernel.shared_1[((threadIdx.x*48) + 37)]))
-          conv2d_nchw_1[12] = (conv2d_nchw_1[12] + (pad_temp.shared_1[51]*kernel.shared_1[((threadIdx.x*48) + 40)]))
-          conv2d_nchw_1[13] = (conv2d_nchw_1[13] + (pad_temp.shared_1[43]*kernel.shared_1[((threadIdx.x*48) + 37)]))
-          conv2d_nchw_1[13] = (conv2d_nchw_1[13] + (pad_temp.shared_1[52]*kernel.shared_1[((threadIdx.x*48) + 40)]))
-          conv2d_nchw_1[0] = (conv2d_nchw_1[0] + (pad_temp.shared_1[38]*kernel.shared_1[((threadIdx.x*48) + 14)]))
-          conv2d_nchw_1[0] = (conv2d_nchw_1[0] + (pad_temp.shared_1[47]*kernel.shared_1[((threadIdx.x*48) + 17)]))
-          conv2d_nchw_1[1] = (conv2d_nchw_1[1] + (pad_temp.shared_1[39]*kernel.shared_1[((threadIdx.x*48) + 14)]))
-          conv2d_nchw_1[1] = (conv2d_nchw_1[1] + (pad_temp.shared_1[48]*kernel.shared_1[((threadIdx.x*48) + 17)]))
-          conv2d_nchw_1[2] = (conv2d_nchw_1[2] + (pad_temp.shared_1[40]*kernel.shared_1[((threadIdx.x*48) + 14)]))
-          conv2d_nchw_1[2] = (conv2d_nchw_1[2] + (pad_temp.shared_1[49]*kernel.shared_1[((threadIdx.x*48) + 17)]))
-          conv2d_nchw_1[3] = (conv2d_nchw_1[3] + (pad_temp.shared_1[41]*kernel.shared_1[((threadIdx.x*48) + 14)]))
-          conv2d_nchw_1[3] = (conv2d_nchw_1[3] + (pad_temp.shared_1[50]*kernel.shared_1[((threadIdx.x*48) + 17)]))
-          conv2d_nchw_1[4] = (conv2d_nchw_1[4] + (pad_temp.shared_1[42]*kernel.shared_1[((threadIdx.x*48) + 14)]))
-          conv2d_nchw_1[4] = (conv2d_nchw_1[4] + (pad_temp.shared_1[51]*kernel.shared_1[((threadIdx.x*48) + 17)]))
-          conv2d_nchw_1[5] = (conv2d_nchw_1[5] + (pad_temp.shared_1[43]*kernel.shared_1[((threadIdx.x*48) + 14)]))
-          conv2d_nchw_1[5] = (conv2d_nchw_1[5] + (pad_temp.shared_1[52]*kernel.shared_1[((threadIdx.x*48) + 17)]))
-          conv2d_nchw_1[6] = (conv2d_nchw_1[6] + (pad_temp.shared_1[44]*kernel.shared_1[((threadIdx.x*48) + 14)]))
-          conv2d_nchw_1[6] = (conv2d_nchw_1[6] + (pad_temp.shared_1[53]*kernel.shared_1[((threadIdx.x*48) + 17)]))
-          conv2d_nchw_1[7] = (conv2d_nchw_1[7] + (pad_temp.shared_1[38]*kernel.shared_1[((threadIdx.x*48) + 38)]))
-          conv2d_nchw_1[7] = (conv2d_nchw_1[7] + (pad_temp.shared_1[47]*kernel.shared_1[((threadIdx.x*48) + 41)]))
-          conv2d_nchw_1[8] = (conv2d_nchw_1[8] + (pad_temp.shared_1[39]*kernel.shared_1[((threadIdx.x*48) + 38)]))
-          conv2d_nchw_1[8] = (conv2d_nchw_1[8] + (pad_temp.shared_1[48]*kernel.shared_1[((threadIdx.x*48) + 41)]))
-          conv2d_nchw_1[9] = (conv2d_nchw_1[9] + (pad_temp.shared_1[40]*kernel.shared_1[((threadIdx.x*48) + 38)]))
-          conv2d_nchw_1[9] = (conv2d_nchw_1[9] + (pad_temp.shared_1[49]*kernel.shared_1[((threadIdx.x*48) + 41)]))
-          conv2d_nchw_1[10] = (conv2d_nchw_1[10] + (pad_temp.shared_1[41]*kernel.shared_1[((threadIdx.x*48) + 38)]))
-          conv2d_nchw_1[10] = (conv2d_nchw_1[10] + (pad_temp.shared_1[50]*kernel.shared_1[((threadIdx.x*48) + 41)]))
-          conv2d_nchw_1[11] = (conv2d_nchw_1[11] + (pad_temp.shared_1[42]*kernel.shared_1[((threadIdx.x*48) + 38)]))
-          conv2d_nchw_1[11] = (conv2d_nchw_1[11] + (pad_temp.shared_1[51]*kernel.shared_1[((threadIdx.x*48) + 41)]))
-          conv2d_nchw_1[12] = (conv2d_nchw_1[12] + (pad_temp.shared_1[43]*kernel.shared_1[((threadIdx.x*48) + 38)]))
-          conv2d_nchw_1[12] = (conv2d_nchw_1[12] + (pad_temp.shared_1[52]*kernel.shared_1[((threadIdx.x*48) + 41)]))
-          conv2d_nchw_1[13] = (conv2d_nchw_1[13] + (pad_temp.shared_1[44]*kernel.shared_1[((threadIdx.x*48) + 38)]))
-          conv2d_nchw_1[13] = (conv2d_nchw_1[13] + (pad_temp.shared_1[53]*kernel.shared_1[((threadIdx.x*48) + 41)]))
-          conv2d_nchw_1[0] = (conv2d_nchw_1[0] + (pad_temp.shared_1[54]*kernel.shared_1[((threadIdx.x*48) + 18)]))
-          conv2d_nchw_1[0] = (conv2d_nchw_1[0] + (pad_temp.shared_1[63]*kernel.shared_1[((threadIdx.x*48) + 21)]))
-          conv2d_nchw_1[1] = (conv2d_nchw_1[1] + (pad_temp.shared_1[55]*kernel.shared_1[((threadIdx.x*48) + 18)]))
-          conv2d_nchw_1[1] = (conv2d_nchw_1[1] + (pad_temp.shared_1[64]*kernel.shared_1[((threadIdx.x*48) + 21)]))
-          conv2d_nchw_1[2] = (conv2d_nchw_1[2] + (pad_temp.shared_1[56]*kernel.shared_1[((threadIdx.x*48) + 18)]))
-          conv2d_nchw_1[2] = (conv2d_nchw_1[2] + (pad_temp.shared_1[65]*kernel.shared_1[((threadIdx.x*48) + 21)]))
-          conv2d_nchw_1[3] = (conv2d_nchw_1[3] + (pad_temp.shared_1[57]*kernel.shared_1[((threadIdx.x*48) + 18)]))
-          conv2d_nchw_1[3] = (conv2d_nchw_1[3] + (pad_temp.shared_1[66]*kernel.shared_1[((threadIdx.x*48) + 21)]))
-          conv2d_nchw_1[4] = (conv2d_nchw_1[4] + (pad_temp.shared_1[58]*kernel.shared_1[((threadIdx.x*48) + 18)]))
-          conv2d_nchw_1[4] = (conv2d_nchw_1[4] + (pad_temp.shared_1[67]*kernel.shared_1[((threadIdx.x*48) + 21)]))
-          conv2d_nchw_1[5] = (conv2d_nchw_1[5] + (pad_temp.shared_1[59]*kernel.shared_1[((threadIdx.x*48) + 18)]))
-          conv2d_nchw_1[5] = (conv2d_nchw_1[5] + (pad_temp.shared_1[68]*kernel.shared_1[((threadIdx.x*48) + 21)]))
-          conv2d_nchw_1[6] = (conv2d_nchw_1[6] + (pad_temp.shared_1[60]*kernel.shared_1[((threadIdx.x*48) + 18)]))
-          conv2d_nchw_1[6] = (conv2d_nchw_1[6] + (pad_temp.shared_1[69]*kernel.shared_1[((threadIdx.x*48) + 21)]))
-          conv2d_nchw_1[7] = (conv2d_nchw_1[7] + (pad_temp.shared_1[54]*kernel.shared_1[((threadIdx.x*48) + 42)]))
-          conv2d_nchw_1[7] = (conv2d_nchw_1[7] + (pad_temp.shared_1[63]*kernel.shared_1[((threadIdx.x*48) + 45)]))
-          conv2d_nchw_1[8] = (conv2d_nchw_1[8] + (pad_temp.shared_1[55]*kernel.shared_1[((threadIdx.x*48) + 42)]))
-          conv2d_nchw_1[8] = (conv2d_nchw_1[8] + (pad_temp.shared_1[64]*kernel.shared_1[((threadIdx.x*48) + 45)]))
-          conv2d_nchw_1[9] = (conv2d_nchw_1[9] + (pad_temp.shared_1[56]*kernel.shared_1[((threadIdx.x*48) + 42)]))
-          conv2d_nchw_1[9] = (conv2d_nchw_1[9] + (pad_temp.shared_1[65]*kernel.shared_1[((threadIdx.x*48) + 45)]))
-          conv2d_nchw_1[10] = (conv2d_nchw_1[10] + (pad_temp.shared_1[57]*kernel.shared_1[((threadIdx.x*48) + 42)]))
-          conv2d_nchw_1[10] = (conv2d_nchw_1[10] + (pad_temp.shared_1[66]*kernel.shared_1[((threadIdx.x*48) + 45)]))
-          conv2d_nchw_1[11] = (conv2d_nchw_1[11] + (pad_temp.shared_1[58]*kernel.shared_1[((threadIdx.x*48) + 42)]))
-          conv2d_nchw_1[11] = (conv2d_nchw_1[11] + (pad_temp.shared_1[67]*kernel.shared_1[((threadIdx.x*48) + 45)]))
-          conv2d_nchw_1[12] = (conv2d_nchw_1[12] + (pad_temp.shared_1[59]*kernel.shared_1[((threadIdx.x*48) + 42)]))
-          conv2d_nchw_1[12] = (conv2d_nchw_1[12] + (pad_temp.shared_1[68]*kernel.shared_1[((threadIdx.x*48) + 45)]))
-          conv2d_nchw_1[13] = (conv2d_nchw_1[13] + (pad_temp.shared_1[60]*kernel.shared_1[((threadIdx.x*48) + 42)]))
-          conv2d_nchw_1[13] = (conv2d_nchw_1[13] + (pad_temp.shared_1[69]*kernel.shared_1[((threadIdx.x*48) + 45)]))
-          conv2d_nchw_1[0] = (conv2d_nchw_1[0] + (pad_temp.shared_1[55]*kernel.shared_1[((threadIdx.x*48) + 19)]))
-          conv2d_nchw_1[0] = (conv2d_nchw_1[0] + (pad_temp.shared_1[64]*kernel.shared_1[((threadIdx.x*48) + 22)]))
-          conv2d_nchw_1[1] = (conv2d_nchw_1[1] + (pad_temp.shared_1[56]*kernel.shared_1[((threadIdx.x*48) + 19)]))
-          conv2d_nchw_1[1] = (conv2d_nchw_1[1] + (pad_temp.shared_1[65]*kernel.shared_1[((threadIdx.x*48) + 22)]))
-          conv2d_nchw_1[2] = (conv2d_nchw_1[2] + (pad_temp.shared_1[57]*kernel.shared_1[((threadIdx.x*48) + 19)]))
-          conv2d_nchw_1[2] = (conv2d_nchw_1[2] + (pad_temp.shared_1[66]*kernel.shared_1[((threadIdx.x*48) + 22)]))
-          conv2d_nchw_1[3] = (conv2d_nchw_1[3] + (pad_temp.shared_1[58]*kernel.shared_1[((threadIdx.x*48) + 19)]))
-          conv2d_nchw_1[3] = (conv2d_nchw_1[3] + (pad_temp.shared_1[67]*kernel.shared_1[((threadIdx.x*48) + 22)]))
-          conv2d_nchw_1[4] = (conv2d_nchw_1[4] + (pad_temp.shared_1[59]*kernel.shared_1[((threadIdx.x*48) + 19)]))
-          conv2d_nchw_1[4] = (conv2d_nchw_1[4] + (pad_temp.shared_1[68]*kernel.shared_1[((threadIdx.x*48) + 22)]))
-          conv2d_nchw_1[5] = (conv2d_nchw_1[5] + (pad_temp.shared_1[60]*kernel.shared_1[((threadIdx.x*48) + 19)]))
-          conv2d_nchw_1[5] = (conv2d_nchw_1[5] + (pad_temp.shared_1[69]*kernel.shared_1[((threadIdx.x*48) + 22)]))
-          conv2d_nchw_1[6] = (conv2d_nchw_1[6] + (pad_temp.shared_1[61]*kernel.shared_1[((threadIdx.x*48) + 19)]))
-          conv2d_nchw_1[6] = (conv2d_nchw_1[6] + (pad_temp.shared_1[70]*kernel.shared_1[((threadIdx.x*48) + 22)]))
-          conv2d_nchw_1[7] = (conv2d_nchw_1[7] + (pad_temp.shared_1[55]*kernel.shared_1[((threadIdx.x*48) + 43)]))
-          conv2d_nchw_1[7] = (conv2d_nchw_1[7] + (pad_temp.shared_1[64]*kernel.shared_1[((threadIdx.x*48) + 46)]))
-          conv2d_nchw_1[8] = (conv2d_nchw_1[8] + (pad_temp.shared_1[56]*kernel.shared_1[((threadIdx.x*48) + 43)]))
-          conv2d_nchw_1[8] = (conv2d_nchw_1[8] + (pad_temp.shared_1[65]*kernel.shared_1[((threadIdx.x*48) + 46)]))
-          conv2d_nchw_1[9] = (conv2d_nchw_1[9] + (pad_temp.shared_1[57]*kernel.shared_1[((threadIdx.x*48) + 43)]))
-          conv2d_nchw_1[9] = (conv2d_nchw_1[9] + (pad_temp.shared_1[66]*kernel.shared_1[((threadIdx.x*48) + 46)]))
-          conv2d_nchw_1[10] = (conv2d_nchw_1[10] + (pad_temp.shared_1[58]*kernel.shared_1[((threadIdx.x*48) + 43)]))
-          conv2d_nchw_1[10] = (conv2d_nchw_1[10] + (pad_temp.shared_1[67]*kernel.shared_1[((threadIdx.x*48) + 46)]))
-          conv2d_nchw_1[11] = (conv2d_nchw_1[11] + (pad_temp.shared_1[59]*kernel.shared_1[((threadIdx.x*48) + 43)]))
-          conv2d_nchw_1[11] = (conv2d_nchw_1[11] + (pad_temp.shared_1[68]*kernel.shared_1[((threadIdx.x*48) + 46)]))
-          conv2d_nchw_1[12] = (conv2d_nchw_1[12] + (pad_temp.shared_1[60]*kernel.shared_1[((threadIdx.x*48) + 43)]))
-          conv2d_nchw_1[12] = (conv2d_nchw_1[12] + (pad_temp.shared_1[69]*kernel.shared_1[((threadIdx.x*48) + 46)]))
-          conv2d_nchw_1[13] = (conv2d_nchw_1[13] + (pad_temp.shared_1[61]*kernel.shared_1[((threadIdx.x*48) + 43)]))
-          conv2d_nchw_1[13] = (conv2d_nchw_1[13] + (pad_temp.shared_1[70]*kernel.shared_1[((threadIdx.x*48) + 46)]))
-          conv2d_nchw_1[0] = (conv2d_nchw_1[0] + (pad_temp.shared_1[56]*kernel.shared_1[((threadIdx.x*48) + 20)]))
-          conv2d_nchw_1[0] = (conv2d_nchw_1[0] + (pad_temp.shared_1[65]*kernel.shared_1[((threadIdx.x*48) + 23)]))
-          conv2d_nchw_1[1] = (conv2d_nchw_1[1] + (pad_temp.shared_1[57]*kernel.shared_1[((threadIdx.x*48) + 20)]))
-          conv2d_nchw_1[1] = (conv2d_nchw_1[1] + (pad_temp.shared_1[66]*kernel.shared_1[((threadIdx.x*48) + 23)]))
-          conv2d_nchw_1[2] = (conv2d_nchw_1[2] + (pad_temp.shared_1[58]*kernel.shared_1[((threadIdx.x*48) + 20)]))
-          conv2d_nchw_1[2] = (conv2d_nchw_1[2] + (pad_temp.shared_1[67]*kernel.shared_1[((threadIdx.x*48) + 23)]))
-          conv2d_nchw_1[3] = (conv2d_nchw_1[3] + (pad_temp.shared_1[59]*kernel.shared_1[((threadIdx.x*48) + 20)]))
-          conv2d_nchw_1[3] = (conv2d_nchw_1[3] + (pad_temp.shared_1[68]*kernel.shared_1[((threadIdx.x*48) + 23)]))
-          conv2d_nchw_1[4] = (conv2d_nchw_1[4] + (pad_temp.shared_1[60]*kernel.shared_1[((threadIdx.x*48) + 20)]))
-          conv2d_nchw_1[4] = (conv2d_nchw_1[4] + (pad_temp.shared_1[69]*kernel.shared_1[((threadIdx.x*48) + 23)]))
-          conv2d_nchw_1[5] = (conv2d_nchw_1[5] + (pad_temp.shared_1[61]*kernel.shared_1[((threadIdx.x*48) + 20)]))
-          conv2d_nchw_1[5] = (conv2d_nchw_1[5] + (pad_temp.shared_1[70]*kernel.shared_1[((threadIdx.x*48) + 23)]))
-          conv2d_nchw_1[6] = (conv2d_nchw_1[6] + (pad_temp.shared_1[62]*kernel.shared_1[((threadIdx.x*48) + 20)]))
-          conv2d_nchw_1[6] = (conv2d_nchw_1[6] + (pad_temp.shared_1[71]*kernel.shared_1[((threadIdx.x*48) + 23)]))
-          conv2d_nchw_1[7] = (conv2d_nchw_1[7] + (pad_temp.shared_1[56]*kernel.shared_1[((threadIdx.x*48) + 44)]))
-          conv2d_nchw_1[7] = (conv2d_nchw_1[7] + (pad_temp.shared_1[65]*kernel.shared_1[((threadIdx.x*48) + 47)]))
-          conv2d_nchw_1[8] = (conv2d_nchw_1[8] + (pad_temp.shared_1[57]*kernel.shared_1[((threadIdx.x*48) + 44)]))
-          conv2d_nchw_1[8] = (conv2d_nchw_1[8] + (pad_temp.shared_1[66]*kernel.shared_1[((threadIdx.x*48) + 47)]))
-          conv2d_nchw_1[9] = (conv2d_nchw_1[9] + (pad_temp.shared_1[58]*kernel.shared_1[((threadIdx.x*48) + 44)]))
-          conv2d_nchw_1[9] = (conv2d_nchw_1[9] + (pad_temp.shared_1[67]*kernel.shared_1[((threadIdx.x*48) + 47)]))
-          conv2d_nchw_1[10] = (conv2d_nchw_1[10] + (pad_temp.shared_1[59]*kernel.shared_1[((threadIdx.x*48) + 44)]))
-          conv2d_nchw_1[10] = (conv2d_nchw_1[10] + (pad_temp.shared_1[68]*kernel.shared_1[((threadIdx.x*48) + 47)]))
-          conv2d_nchw_1[11] = (conv2d_nchw_1[11] + (pad_temp.shared_1[60]*kernel.shared_1[((threadIdx.x*48) + 44)]))
-          conv2d_nchw_1[11] = (conv2d_nchw_1[11] + (pad_temp.shared_1[69]*kernel.shared_1[((threadIdx.x*48) + 47)]))
-          conv2d_nchw_1[12] = (conv2d_nchw_1[12] + (pad_temp.shared_1[61]*kernel.shared_1[((threadIdx.x*48) + 44)]))
-          conv2d_nchw_1[12] = (conv2d_nchw_1[12] + (pad_temp.shared_1[70]*kernel.shared_1[((threadIdx.x*48) + 47)]))
-          conv2d_nchw_1[13] = (conv2d_nchw_1[13] + (pad_temp.shared_1[62]*kernel.shared_1[((threadIdx.x*48) + 44)]))
-          conv2d_nchw_1[13] = (conv2d_nchw_1[13] + (pad_temp.shared_1[71]*kernel.shared_1[((threadIdx.x*48) + 47)]))
+      let cse_var_2: int32 = (rc.outer.outer*392)
+      let cse_var_1: int32 = (rc.outer.outer*72)
+       {
+        attr [IterVar(threadIdx.x_1: int32, (nullptr), &quot;ThreadIndex&quot;, &quot;threadIdx.x&quot;)] &quot;thread_extent&quot; = 64;
+        pad_temp.shared_1: Buffer(pad_temp.shared, float32, [216], [], scope=&quot;shared&quot;)[threadIdx.x_1] = @tir.if_then_else(((((1 &lt;= (floordiv(floormod(threadIdx.x_1, 27), 9) + floormod(blockIdx.x, 7))) &amp;&amp; ((floordiv(floormod(threadIdx.x_1, 27), 9) + floormod(blockIdx.x, 7)) &lt; 8)) &amp;&amp; (1 &lt;= floormod(threadIdx.x_1, 9))) &amp;&amp; (floormod(threadIdx.x_1, 9) &lt; 8)), data_3: Buffer(data_2, float32, [25088], [])[(((((cse_var_2 + (floordiv(threadIdx.x_1, 27) [...]
+        attr [IterVar(threadIdx.x_1, (nullptr), &quot;ThreadIndex&quot;, &quot;threadIdx.x&quot;)] &quot;thread_extent&quot; = 64;
+        pad_temp.shared_1[(threadIdx.x_1 + 64)] = @tir.if_then_else(((((1 &lt;= (floordiv(floormod((threadIdx.x_1 + 10), 27), 9) + floormod(blockIdx.x, 7))) &amp;&amp; ((floordiv(floormod((threadIdx.x_1 + 10), 27), 9) + floormod(blockIdx.x, 7)) &lt; 8)) &amp;&amp; (1 &lt;= floormod((threadIdx.x_1 + 1), 9))) &amp;&amp; (floormod((threadIdx.x_1 + 1), 9) &lt; 8)), data_3[(((((cse_var_2 + (floordiv((threadIdx.x_1 + 64), 27)*49)) + (floordiv(floormod((threadIdx.x_1 + 10), 27), 9)*7)) + (floor [...]
+        attr [IterVar(threadIdx.x_1, (nullptr), &quot;ThreadIndex&quot;, &quot;threadIdx.x&quot;)] &quot;thread_extent&quot; = 64;
+        pad_temp.shared_1[(threadIdx.x_1 + 128)] = @tir.if_then_else(((((1 &lt;= (floordiv(floormod((threadIdx.x_1 + 20), 27), 9) + floormod(blockIdx.x, 7))) &amp;&amp; ((floordiv(floormod((threadIdx.x_1 + 20), 27), 9) + floormod(blockIdx.x, 7)) &lt; 8)) &amp;&amp; (1 &lt;= floormod((threadIdx.x_1 + 2), 9))) &amp;&amp; (floormod((threadIdx.x_1 + 2), 9) &lt; 8)), data_3[(((((cse_var_2 + (floordiv((threadIdx.x_1 + 128), 27)*49)) + (floordiv(floormod((threadIdx.x_1 + 20), 27), 9)*7)) + (flo [...]
+        attr [IterVar(threadIdx.x_1, (nullptr), &quot;ThreadIndex&quot;, &quot;threadIdx.x&quot;)] &quot;thread_extent&quot; = 64;
+        if @tir.likely((threadIdx.x_1 &lt; 24), dtype=bool) {
+          pad_temp.shared_1[(threadIdx.x_1 + 192)] = @tir.if_then_else(((((1 &lt;= (floordiv(floormod((threadIdx.x_1 + 3), 27), 9) + floormod(blockIdx.x, 7))) &amp;&amp; ((floordiv(floormod((threadIdx.x_1 + 3), 27), 9) + floormod(blockIdx.x, 7)) &lt; 8)) &amp;&amp; (1 &lt;= floormod((threadIdx.x_1 + 3), 9))) &amp;&amp; (floormod((threadIdx.x_1 + 3), 9) &lt; 8)), data_3[(((((cse_var_2 + (floordiv((threadIdx.x_1 + 192), 27)*49)) + (floordiv(floormod((threadIdx.x_1 + 3), 27), 9)*7)) + (floo [...]
         }
+        attr [IterVar(threadIdx.x_2: int32, (nullptr), &quot;ThreadIndex&quot;, &quot;threadIdx.x&quot;)] &quot;thread_extent&quot; = 64;
+        kernel.shared_1: Buffer(kernel.shared, float32, [9216], [], scope=&quot;shared&quot;)[threadIdx.x_2] = kernel_3: Buffer(kernel_2, float32, [2359296], [])[(((floordiv(blockIdx.x, 7)*589824) + cse_var_1) + threadIdx.x_2)]
+        attr [IterVar(threadIdx.x_2, (nullptr), &quot;ThreadIndex&quot;, &quot;threadIdx.x&quot;)] &quot;thread_extent&quot; = 64;
+        kernel.shared_1[(threadIdx.x_2 + 64)] = kernel_3[(((((floordiv(blockIdx.x, 7)*589824) + (floordiv((threadIdx.x_2 + 64), 72)*4608)) + cse_var_1) + (floordiv(floormod((threadIdx.x_2 + 64), 72), 3)*3)) + floormod((threadIdx.x_2 + 1), 3))]
+        attr [IterVar(threadIdx.x_2, (nullptr), &quot;ThreadIndex&quot;, &quot;threadIdx.x&quot;)] &quot;thread_extent&quot; = 64;
+        kernel.shared_1[(threadIdx.x_2 + 128)] = kernel_3[(((((floordiv(blockIdx.x, 7)*589824) + (floordiv((threadIdx.x_2 + 128), 72)*4608)) + cse_var_1) + (floordiv(floormod((threadIdx.x_2 + 56), 72), 3)*3)) + floormod((threadIdx.x_2 + 2), 3))]
+        attr [IterVar(threadIdx.x_2, (nullptr), &quot;ThreadIndex&quot;, &quot;threadIdx.x&quot;)] &quot;thread_extent&quot; = 64;
+        kernel.shared_1[(threadIdx.x_2 + 192)] = kernel_3[(((((floordiv(blockIdx.x, 7)*589824) + (floordiv((threadIdx.x_2 + 192), 72)*4608)) + cse_var_1) + (floormod((floordiv(threadIdx.x_2, 3) + 16), 24)*3)) + floormod(threadIdx.x_2, 3))]
+        attr [IterVar(threadIdx.x_2, (nullptr), &quot;ThreadIndex&quot;, &quot;threadIdx.x&quot;)] &quot;thread_extent&quot; = 64;
+        kernel.shared_1[(threadIdx.x_2 + 256)] = kernel_3[(((((floordiv(blockIdx.x, 7)*589824) + (floordiv((threadIdx.x_2 + 256), 72)*4608)) + cse_var_1) + (floordiv(floormod((threadIdx.x_2 + 40), 72), 3)*3)) + floormod((threadIdx.x_2 + 1), 3))]
+        attr [IterVar(threadIdx.x_2, (nullptr), &quot;ThreadIndex&quot;, &quot;threadIdx.x&quot;)] &quot;thread_extent&quot; = 64;
+        kernel.shared_1[(threadIdx.x_2 + 320)] = kernel_3[(((((floordiv(blockIdx.x, 7)*589824) + (floordiv((threadIdx.x_2 + 320), 72)*4608)) + cse_var_1) + (floordiv(floormod((threadIdx.x_2 + 32), 72), 3)*3)) + floormod((threadIdx.x_2 + 2), 3))]
+        attr [IterVar(threadIdx.x_2, (nullptr), &quot;ThreadIndex&quot;, &quot;threadIdx.x&quot;)] &quot;thread_extent&quot; = 64;
+        kernel.shared_1[(threadIdx.x_2 + 384)] = kernel_3[(((((floordiv(blockIdx.x, 7)*589824) + (floordiv((threadIdx.x_2 + 384), 72)*4608)) + cse_var_1) + (floormod((floordiv(threadIdx.x_2, 3) + 8), 24)*3)) + floormod(threadIdx.x_2, 3))]
+        attr [IterVar(threadIdx.x_2, (nullptr), &quot;ThreadIndex&quot;, &quot;threadIdx.x&quot;)] &quot;thread_extent&quot; = 64;
+        kernel.shared_1[(threadIdx.x_2 + 448)] = kernel_3[(((((floordiv(blockIdx.x, 7)*589824) + (floordiv((threadIdx.x_2 + 448), 72)*4608)) + cse_var_1) + (floordiv(floormod((threadIdx.x_2 + 16), 72), 3)*3)) + floormod((threadIdx.x_2 + 1), 3))]
+        attr [IterVar(threadIdx.x_2, (nullptr), &quot;ThreadIndex&quot;, &quot;threadIdx.x&quot;)] &quot;thread_extent&quot; = 64;
+        kernel.shared_1[(threadIdx.x_2 + 512)] = kernel_3[(((((floordiv(blockIdx.x, 7)*589824) + (floordiv((threadIdx.x_2 + 512), 72)*4608)) + cse_var_1) + (floordiv(floormod((threadIdx.x_2 + 8), 72), 3)*3)) + floormod((threadIdx.x_2 + 2), 3))]
+        attr [IterVar(threadIdx.x_2, (nullptr), &quot;ThreadIndex&quot;, &quot;threadIdx.x&quot;)] &quot;thread_extent&quot; = 64;
+        kernel.shared_1[(threadIdx.x_2 + 576)] = kernel_3[((((floordiv(blockIdx.x, 7)*589824) + cse_var_1) + threadIdx.x_2) + 36864)]
+        attr [IterVar(threadIdx.x_2, (nullptr), &quot;ThreadIndex&quot;, &quot;threadIdx.x&quot;)] &quot;thread_extent&quot; = 64;
+        kernel.shared_1[(threadIdx.x_2 + 640)] = kernel_3[(((((floordiv(blockIdx.x, 7)*589824) + (floordiv((threadIdx.x_2 + 640), 72)*4608)) + cse_var_1) + (floordiv(floormod((threadIdx.x_2 + 64), 72), 3)*3)) + floormod((threadIdx.x_2 + 1), 3))]
+        attr [IterVar(threadIdx.x_2, (nullptr), &quot;ThreadIndex&quot;, &quot;threadIdx.x&quot;)] &quot;thread_extent&quot; = 64;
+        kernel.shared_1[(threadIdx.x_2 + 704)] = kernel_3[(((((floordiv(blockIdx.x, 7)*589824) + (floordiv((threadIdx.x_2 + 704), 72)*4608)) + cse_var_1) + (floordiv(floormod((threadIdx.x_2 + 56), 72), 3)*3)) + floormod((threadIdx.x_2 + 2), 3))]
+        attr [IterVar(threadIdx.x_2, (nullptr), &quot;ThreadIndex&quot;, &quot;threadIdx.x&quot;)] &quot;thread_extent&quot; = 64;
+        kernel.shared_1[(threadIdx.x_2 + 768)] = kernel_3[(((((floordiv(blockIdx.x, 7)*589824) + (floordiv((threadIdx.x_2 + 768), 72)*4608)) + cse_var_1) + (floormod((floordiv(threadIdx.x_2, 3) + 16), 24)*3)) + floormod(threadIdx.x_2, 3))]
+        attr [IterVar(threadIdx.x_2, (nullptr), &quot;ThreadIndex&quot;, &quot;threadIdx.x&quot;)] &quot;thread_extent&quot; = 64;
+        kernel.shared_1[(threadIdx.x_2 + 832)] = kernel_3[(((((floordiv(blockIdx.x, 7)*589824) + (floordiv((threadIdx.x_2 + 832), 72)*4608)) + cse_var_1) + (floordiv(floormod((threadIdx.x_2 + 40), 72), 3)*3)) + floormod((threadIdx.x_2 + 1), 3))]
+        attr [IterVar(threadIdx.x_2, (nullptr), &quot;ThreadIndex&quot;, &quot;threadIdx.x&quot;)] &quot;thread_extent&quot; = 64;
+        kernel.shared_1[(threadIdx.x_2 + 896)] = kernel_3[(((((floordiv(blockIdx.x, 7)*589824) + (floordiv((threadIdx.x_2 + 896), 72)*4608)) + cse_var_1) + (floordiv(floormod((threadIdx.x_2 + 32), 72), 3)*3)) + floormod((threadIdx.x_2 + 2), 3))]
+        attr [IterVar(threadIdx.x_2, (nullptr), &quot;ThreadIndex&quot;, &quot;threadIdx.x&quot;)] &quot;thread_extent&quot; = 64;
+        kernel.shared_1[(threadIdx.x_2 + 960)] = kernel_3[(((((floordiv(blockIdx.x, 7)*589824) + (floordiv((threadIdx.x_2 + 960), 72)*4608)) + cse_var_1) + (floormod((floordiv(threadIdx.x_2, 3) + 8), 24)*3)) + floormod(threadIdx.x_2, 3))]
+        attr [IterVar(threadIdx.x_2, (nullptr), &quot;ThreadIndex&quot;, &quot;threadIdx.x&quot;)] &quot;thread_extent&quot; = 64;
+        kernel.shared_1[(threadIdx.x_2 + 1024)] = kernel_3[(((((floordiv(blockIdx.x, 7)*589824) + (floordiv((threadIdx.x_2 + 1024), 72)*4608)) + cse_var_1) + (floordiv(floormod((threadIdx.x_2 + 16), 72), 3)*3)) + floormod((threadIdx.x_2 + 1), 3))]
+        attr [IterVar(threadIdx.x_2, (nullptr), &quot;ThreadIndex&quot;, &quot;threadIdx.x&quot;)] &quot;thread_extent&quot; = 64;
+        kernel.shared_1[(threadIdx.x_2 + 1088)] = kernel_3[(((((floordiv(blockIdx.x, 7)*589824) + (floordiv((threadIdx.x_2 + 1088), 72)*4608)) + cse_var_1) + (floordiv(floormod((threadIdx.x_2 + 8), 72), 3)*3)) + floormod((threadIdx.x_2 + 2), 3))]
+        attr [IterVar(threadIdx.x_2, (nullptr), &quot;ThreadIndex&quot;, &quot;threadIdx.x&quot;)] &quot;thread_extent&quot; = 64;
+        kernel.shared_1[(threadIdx.x_2 + 1152)] = kernel_3[((((floordiv(blockIdx.x, 7)*589824) + cse_var_1) + threadIdx.x_2) + 73728)]
+        attr [IterVar(threadIdx.x_2, (nullptr), &quot;ThreadIndex&quot;, &quot;threadIdx.x&quot;)] &quot;thread_extent&quot; = 64;
+        kernel.shared_1[(threadIdx.x_2 + 1216)] = kernel_3[(((((floordiv(blockIdx.x, 7)*589824) + (floordiv((threadIdx.x_2 + 1216), 72)*4608)) + cse_var_1) + (floordiv(floormod((threadIdx.x_2 + 64), 72), 3)*3)) + floormod((threadIdx.x_2 + 1), 3))]
+        attr [IterVar(threadIdx.x_2, (nullptr), &quot;ThreadIndex&quot;, &quot;threadIdx.x&quot;)] &quot;thread_extent&quot; = 64;
+        kernel.shared_1[(threadIdx.x_2 + 1280)] = kernel_3[(((((floordiv(blockIdx.x, 7)*589824) + (floordiv((threadIdx.x_2 + 1280), 72)*4608)) + cse_var_1) + (floordiv(floormod((threadIdx.x_2 + 56), 72), 3)*3)) + floormod((threadIdx.x_2 + 2), 3))]
+        attr [IterVar(threadIdx.x_2, (nullptr), &quot;ThreadIndex&quot;, &quot;threadIdx.x&quot;)] &quot;thread_extent&quot; = 64;
+        kernel.shared_1[(threadIdx.x_2 + 1344)] = kernel_3[(((((floordiv(blockIdx.x, 7)*589824) + (floordiv((threadIdx.x_2 + 1344), 72)*4608)) + cse_var_1) + (floormod((floordiv(threadIdx.x_2, 3) + 16), 24)*3)) + floormod(threadIdx.x_2, 3))]
+        attr [IterVar(threadIdx.x_2, (nullptr), &quot;ThreadIndex&quot;, &quot;threadIdx.x&quot;)] &quot;thread_extent&quot; = 64;
+        kernel.shared_1[(threadIdx.x_2 + 1408)] = kernel_3[(((((floordiv(blockIdx.x, 7)*589824) + (floordiv((threadIdx.x_2 + 1408), 72)*4608)) + cse_var_1) + (floordiv(floormod((threadIdx.x_2 + 40), 72), 3)*3)) + floormod((threadIdx.x_2 + 1), 3))]
+        attr [IterVar(threadIdx.x_2, (nullptr), &quot;ThreadIndex&quot;, &quot;threadIdx.x&quot;)] &quot;thread_extent&quot; = 64;
+        kernel.shared_1[(threadIdx.x_2 + 1472)] = kernel_3[(((((floordiv(blockIdx.x, 7)*589824) + (floordiv((threadIdx.x_2 + 1472), 72)*4608)) + cse_var_1) + (floordiv(floormod((threadIdx.x_2 + 32), 72), 3)*3)) + floormod((threadIdx.x_2 + 2), 3))]
+        attr [IterVar(threadIdx.x_2, (nullptr), &quot;ThreadIndex&quot;, &quot;threadIdx.x&quot;)] &quot;thread_extent&quot; = 64;
+        kernel.shared_1[(threadIdx.x_2 + 1536)] = kernel_3[(((((floordiv(blockIdx.x, 7)*589824) + (floordiv((threadIdx.x_2 + 1536), 72)*4608)) + cse_var_1) + (floormod((floordiv(threadIdx.x_2, 3) + 8), 24)*3)) + floormod(threadIdx.x_2, 3))]
+        attr [IterVar(threadIdx.x_2, (nullptr), &quot;ThreadIndex&quot;, &quot;threadIdx.x&quot;)] &quot;thread_extent&quot; = 64;
+        kernel.shared_1[(threadIdx.x_2 + 1600)] = kernel_3[(((((floordiv(blockIdx.x, 7)*589824) + (floordiv((threadIdx.x_2 + 1600), 72)*4608)) + cse_var_1) + (floordiv(floormod((threadIdx.x_2 + 16), 72), 3)*3)) + floormod((threadIdx.x_2 + 1), 3))]
+        attr [IterVar(threadIdx.x_2, (nullptr), &quot;ThreadIndex&quot;, &quot;threadIdx.x&quot;)] &quot;thread_extent&quot; = 64;
+        kernel.shared_1[(threadIdx.x_2 + 1664)] = kernel_3[(((((floordiv(blockIdx.x, 7)*589824) + (floordiv((threadIdx.x_2 + 1664), 72)*4608)) + cse_var_1) + (floordiv(floormod((threadIdx.x_2 + 8), 72), 3)*3)) + floormod((threadIdx.x_2 + 2), 3))]
+        attr [IterVar(threadIdx.x_2, (nullptr), &quot;ThreadIndex&quot;, &quot;threadIdx.x&quot;)] &quot;thread_extent&quot; = 64;
+        kernel.shared_1[(threadIdx.x_2 + 1728)] = kernel_3[((((floordiv(blockIdx.x, 7)*589824) + cse_var_1) + threadIdx.x_2) + 110592)]
+        attr [IterVar(threadIdx.x_2, (nullptr), &quot;ThreadIndex&quot;, &quot;threadIdx.x&quot;)] &quot;thread_extent&quot; = 64;
+        kernel.shared_1[(threadIdx.x_2 + 1792)] = kernel_3[(((((floordiv(blockIdx.x, 7)*589824) + (floordiv((threadIdx.x_2 + 1792), 72)*4608)) + cse_var_1) + (floordiv(floormod((threadIdx.x_2 + 64), 72), 3)*3)) + floormod((threadIdx.x_2 + 1), 3))]
+        attr [IterVar(threadIdx.x_2, (nullptr), &quot;ThreadIndex&quot;, &quot;threadIdx.x&quot;)] &quot;thread_extent&quot; = 64;
+        kernel.shared_1[(threadIdx.x_2 + 1856)] = kernel_3[(((((floordiv(blockIdx.x, 7)*589824) + (floordiv((threadIdx.x_2 + 1856), 72)*4608)) + cse_var_1) + (floordiv(floormod((threadIdx.x_2 + 56), 72), 3)*3)) + floormod((threadIdx.x_2 + 2), 3))]
+        attr [IterVar(threadIdx.x_2, (nullptr), &quot;ThreadIndex&quot;, &quot;threadIdx.x&quot;)] &quot;thread_extent&quot; = 64;
+        kernel.shared_1[(threadIdx.x_2 + 1920)] = kernel_3[(((((floordiv(blockIdx.x, 7)*589824) + (floordiv((threadIdx.x_2 + 1920), 72)*4608)) + cse_var_1) + (floormod((floordiv(threadIdx.x_2, 3) + 16), 24)*3)) + floormod(threadIdx.x_2, 3))]
+        attr [IterVar(threadIdx.x_2, (nullptr), &quot;ThreadIndex&quot;, &quot;threadIdx.x&quot;)] &quot;thread_extent&quot; = 64;
+        kernel.shared_1[(threadIdx.x_2 + 1984)] = kernel_3[(((((floordiv(blockIdx.x, 7)*589824) + (floordiv((threadIdx.x_2 + 1984), 72)*4608)) + cse_var_1) + (floordiv(floormod((threadIdx.x_2 + 40), 72), 3)*3)) + floormod((threadIdx.x_2 + 1), 3))]
+        attr [IterVar(threadIdx.x_2, (nullptr), &quot;ThreadIndex&quot;, &quot;threadIdx.x&quot;)] &quot;thread_extent&quot; = 64;
+        kernel.shared_1[(threadIdx.x_2 + 2048)] = kernel_3[(((((floordiv(blockIdx.x, 7)*589824) + (floordiv((threadIdx.x_2 + 2048), 72)*4608)) + cse_var_1) + (floordiv(floormod((threadIdx.x_2 + 32), 72), 3)*3)) + floormod((threadIdx.x_2 + 2), 3))]
+        attr [IterVar(threadIdx.x_2, (nullptr), &quot;ThreadIndex&quot;, &quot;threadIdx.x&quot;)] &quot;thread_extent&quot; = 64;
+        kernel.shared_1[(threadIdx.x_2 + 2112)] = kernel_3[(((((floordiv(blockIdx.x, 7)*589824) + (floordiv((threadIdx.x_2 + 2112), 72)*4608)) + cse_var_1) + (floormod((floordiv(threadIdx.x_2, 3) + 8), 24)*3)) + floormod(threadIdx.x_2, 3))]
+        attr [IterVar(threadIdx.x_2, (nullptr), &quot;ThreadIndex&quot;, &quot;threadIdx.x&quot;)] &quot;thread_extent&quot; = 64;
+        kernel.shared_1[(threadIdx.x_2 + 2176)] = kernel_3[(((((floordiv(blockIdx.x, 7)*589824) + (floordiv((threadIdx.x_2 + 2176), 72)*4608)) + cse_var_1) + (floordiv(floormod((threadIdx.x_2 + 16), 72), 3)*3)) + floormod((threadIdx.x_2 + 1), 3))]
+        attr [IterVar(threadIdx.x_2, (nullptr), &quot;ThreadIndex&quot;, &quot;threadIdx.x&quot;)] &quot;thread_extent&quot; = 64;
+        kernel.shared_1[(threadIdx.x_2 + 2240)] = kernel_3[(((((floordiv(blockIdx.x, 7)*589824) + (floordiv((threadIdx.x_2 + 2240), 72)*4608)) + cse_var_1) + (floordiv(floormod((threadIdx.x_2 + 8), 72), 3)*3)) + floormod((threadIdx.x_2 + 2), 3))]
+        attr [IterVar(threadIdx.x_2, (nullptr), &quot;ThreadIndex&quot;, &quot;threadIdx.x&quot;)] &quot;thread_extent&quot; = 64;
+        kernel.shared_1[(threadIdx.x_2 + 2304)] = kernel_3[((((floordiv(blockIdx.x, 7)*589824) + cse_var_1) + threadIdx.x_2) + 147456)]
+        attr [IterVar(threadIdx.x_2, (nullptr), &quot;ThreadIndex&quot;, &quot;threadIdx.x&quot;)] &quot;thread_extent&quot; = 64;
+        kernel.shared_1[(threadIdx.x_2 + 2368)] = kernel_3[(((((floordiv(blockIdx.x, 7)*589824) + (floordiv((threadIdx.x_2 + 2368), 72)*4608)) + cse_var_1) + (floordiv(floormod((threadIdx.x_2 + 64), 72), 3)*3)) + floormod((threadIdx.x_2 + 1), 3))]
+        attr [IterVar(threadIdx.x_2, (nullptr), &quot;ThreadIndex&quot;, &quot;threadIdx.x&quot;)] &quot;thread_extent&quot; = 64;
+        kernel.shared_1[(threadIdx.x_2 + 2432)] = kernel_3[(((((floordiv(blockIdx.x, 7)*589824) + (floordiv((threadIdx.x_2 + 2432), 72)*4608)) + cse_var_1) + (floordiv(floormod((threadIdx.x_2 + 56), 72), 3)*3)) + floormod((threadIdx.x_2 + 2), 3))]
+        attr [IterVar(threadIdx.x_2, (nullptr), &quot;ThreadIndex&quot;, &quot;threadIdx.x&quot;)] &quot;thread_extent&quot; = 64;
+        kernel.shared_1[(threadIdx.x_2 + 2496)] = kernel_3[(((((floordiv(blockIdx.x, 7)*589824) + (floordiv((threadIdx.x_2 + 2496), 72)*4608)) + cse_var_1) + (floormod((floordiv(threadIdx.x_2, 3) + 16), 24)*3)) + floormod(threadIdx.x_2, 3))]
+        attr [IterVar(threadIdx.x_2, (nullptr), &quot;ThreadIndex&quot;, &quot;threadIdx.x&quot;)] &quot;thread_extent&quot; = 64;
+        kernel.shared_1[(threadIdx.x_2 + 2560)] = kernel_3[(((((floordiv(blockIdx.x, 7)*589824) + (floordiv((threadIdx.x_2 + 2560), 72)*4608)) + cse_var_1) + (floordiv(floormod((threadIdx.x_2 + 40), 72), 3)*3)) + floormod((threadIdx.x_2 + 1), 3))]
+        attr [IterVar(threadIdx.x_2, (nullptr), &quot;ThreadIndex&quot;, &quot;threadIdx.x&quot;)] &quot;thread_extent&quot; = 64;
+        kernel.shared_1[(threadIdx.x_2 + 2624)] = kernel_3[(((((floordiv(blockIdx.x, 7)*589824) + (floordiv((threadIdx.x_2 + 2624), 72)*4608)) + cse_var_1) + (floordiv(floormod((threadIdx.x_2 + 32), 72), 3)*3)) + floormod((threadIdx.x_2 + 2), 3))]
+        attr [IterVar(threadIdx.x_2, (nullptr), &quot;ThreadIndex&quot;, &quot;threadIdx.x&quot;)] &quot;thread_extent&quot; = 64;
+        kernel.shared_1[(threadIdx.x_2 + 2688)] = kernel_3[(((((floordiv(blockIdx.x, 7)*589824) + (floordiv((threadIdx.x_2 + 2688), 72)*4608)) + cse_var_1) + (floormod((floordiv(threadIdx.x_2, 3) + 8), 24)*3)) + floormod(threadIdx.x_2, 3))]
+        attr [IterVar(threadIdx.x_2, (nullptr), &quot;ThreadIndex&quot;, &quot;threadIdx.x&quot;)] &quot;thread_extent&quot; = 64;
+        kernel.shared_1[(threadIdx.x_2 + 2752)] = kernel_3[(((((floordiv(blockIdx.x, 7)*589824) + (floordiv((threadIdx.x_2 + 2752), 72)*4608)) + cse_var_1) + (floordiv(floormod((threadIdx.x_2 + 16), 72), 3)*3)) + floormod((threadIdx.x_2 + 1), 3))]
+        attr [IterVar(threadIdx.x_2, (nullptr), &quot;ThreadIndex&quot;, &quot;threadIdx.x&quot;)] &quot;thread_extent&quot; = 64;
+        kernel.shared_1[(threadIdx.x_2 + 2816)] = kernel_3[(((((floordiv(blockIdx.x, 7)*589824) + (floordiv((threadIdx.x_2 + 2816), 72)*4608)) + cse_var_1) + (floordiv(floormod((threadIdx.x_2 + 8), 72), 3)*3)) + floormod((threadIdx.x_2 + 2), 3))]
+        attr [IterVar(threadIdx.x_2, (nullptr), &quot;ThreadIndex&quot;, &quot;threadIdx.x&quot;)] &quot;thread_extent&quot; = 64;
+        kernel.shared_1[(threadIdx.x_2 + 2880)] = kernel_3[((((floordiv(blockIdx.x, 7)*589824) + cse_var_1) + threadIdx.x_2) + 184320)]
+        attr [IterVar(threadIdx.x_2, (nullptr), &quot;ThreadIndex&quot;, &quot;threadIdx.x&quot;)] &quot;thread_extent&quot; = 64;
+        kernel.shared_1[(threadIdx.x_2 + 2944)] = kernel_3[(((((floordiv(blockIdx.x, 7)*589824) + (floordiv((threadIdx.x_2 + 2944), 72)*4608)) + cse_var_1) + (floordiv(floormod((threadIdx.x_2 + 64), 72), 3)*3)) + floormod((threadIdx.x_2 + 1), 3))]
+        attr [IterVar(threadIdx.x_2, (nullptr), &quot;ThreadIndex&quot;, &quot;threadIdx.x&quot;)] &quot;thread_extent&quot; = 64;
+        kernel.shared_1[(threadIdx.x_2 + 3008)] = kernel_3[(((((floordiv(blockIdx.x, 7)*589824) + (floordiv((threadIdx.x_2 + 3008), 72)*4608)) + cse_var_1) + (floordiv(floormod((threadIdx.x_2 + 56), 72), 3)*3)) + floormod((threadIdx.x_2 + 2), 3))]
+        attr [IterVar(threadIdx.x_2, (nullptr), &quot;ThreadIndex&quot;, &quot;threadIdx.x&quot;)] &quot;thread_extent&quot; = 64;
+        kernel.shared_1[(threadIdx.x_2 + 3072)] = kernel_3[(((((floordiv(blockIdx.x, 7)*589824) + (floordiv((threadIdx.x_2 + 3072), 72)*4608)) + cse_var_1) + (floormod((floordiv(threadIdx.x_2, 3) + 16), 24)*3)) + floormod(threadIdx.x_2, 3))]
+        attr [IterVar(threadIdx.x_2, (nullptr), &quot;ThreadIndex&quot;, &quot;threadIdx.x&quot;)] &quot;thread_extent&quot; = 64;
+        kernel.shared_1[(threadIdx.x_2 + 3136)] = kernel_3[(((((floordiv(blockIdx.x, 7)*589824) + (floordiv((threadIdx.x_2 + 3136), 72)*4608)) + cse_var_1) + (floordiv(floormod((threadIdx.x_2 + 40), 72), 3)*3)) + floormod((threadIdx.x_2 + 1), 3))]
+        attr [IterVar(threadIdx.x_2, (nullptr), &quot;ThreadIndex&quot;, &quot;threadIdx.x&quot;)] &quot;thread_extent&quot; = 64;
+        kernel.shared_1[(threadIdx.x_2 + 3200)] = kernel_3[(((((floordiv(blockIdx.x, 7)*589824) + (floordiv((threadIdx.x_2 + 3200), 72)*4608)) + cse_var_1) + (floordiv(floormod((threadIdx.x_2 + 32), 72), 3)*3)) + floormod((threadIdx.x_2 + 2), 3))]
+        attr [IterVar(threadIdx.x_2, (nullptr), &quot;ThreadIndex&quot;, &quot;threadIdx.x&quot;)] &quot;thread_extent&quot; = 64;
+        kernel.shared_1[(threadIdx.x_2 + 3264)] = kernel_3[(((((floordiv(blockIdx.x, 7)*589824) + (floordiv((threadIdx.x_2 + 3264), 72)*4608)) + cse_var_1) + (floormod((floordiv(threadIdx.x_2, 3) + 8), 24)*3)) + floormod(threadIdx.x_2, 3))]
+        attr [IterVar(threadIdx.x_2, (nullptr), &quot;ThreadIndex&quot;, &quot;threadIdx.x&quot;)] &quot;thread_extent&quot; = 64;
+        kernel.shared_1[(threadIdx.x_2 + 3328)] = kernel_3[(((((floordiv(blockIdx.x, 7)*589824) + (floordiv((threadIdx.x_2 + 3328), 72)*4608)) + cse_var_1) + (floordiv(floormod((threadIdx.x_2 + 16), 72), 3)*3)) + floormod((threadIdx.x_2 + 1), 3))]
+        attr [IterVar(threadIdx.x_2, (nullptr), &quot;ThreadIndex&quot;, &quot;threadIdx.x&quot;)] &quot;thread_extent&quot; = 64;
+        kernel.shared_1[(threadIdx.x_2 + 3392)] = kernel_3[(((((floordiv(blockIdx.x, 7)*589824) + (floordiv((threadIdx.x_2 + 3392), 72)*4608)) + cse_var_1) + (floordiv(floormod((threadIdx.x_2 + 8), 72), 3)*3)) + floormod((threadIdx.x_2 + 2), 3))]
+        attr [IterVar(threadIdx.x_2, (nullptr), &quot;ThreadIndex&quot;, &quot;threadIdx.x&quot;)] &quot;thread_extent&quot; = 64;
+        kernel.shared_1[(threadIdx.x_2 + 3456)] = kernel_3[((((floordiv(blockIdx.x, 7)*589824) + cse_var_1) + threadIdx.x_2) + 221184)]
+        attr [IterVar(threadIdx.x_2, (nullptr), &quot;ThreadIndex&quot;, &quot;threadIdx.x&quot;)] &quot;thread_extent&quot; = 64;
+        kernel.shared_1[(threadIdx.x_2 + 3520)] = kernel_3[(((((floordiv(blockIdx.x, 7)*589824) + (floordiv((threadIdx.x_2 + 3520), 72)*4608)) + cse_var_1) + (floordiv(floormod((threadIdx.x_2 + 64), 72), 3)*3)) + floormod((threadIdx.x_2 + 1), 3))]
+        attr [IterVar(threadIdx.x_2, (nullptr), &quot;ThreadIndex&quot;, &quot;threadIdx.x&quot;)] &quot;thread_extent&quot; = 64;
+        kernel.shared_1[(threadIdx.x_2 + 3584)] = kernel_3[(((((floordiv(blockIdx.x, 7)*589824) + (floordiv((threadIdx.x_2 + 3584), 72)*4608)) + cse_var_1) + (floordiv(floormod((threadIdx.x_2 + 56), 72), 3)*3)) + floormod((threadIdx.x_2 + 2), 3))]
+        attr [IterVar(threadIdx.x_2, (nullptr), &quot;ThreadIndex&quot;, &quot;threadIdx.x&quot;)] &quot;thread_extent&quot; = 64;
+        kernel.shared_1[(threadIdx.x_2 + 3648)] = kernel_3[(((((floordiv(blockIdx.x, 7)*589824) + (floordiv((threadIdx.x_2 + 3648), 72)*4608)) + cse_var_1) + (floormod((floordiv(threadIdx.x_2, 3) + 16), 24)*3)) + floormod(threadIdx.x_2, 3))]
+        attr [IterVar(threadIdx.x_2, (nullptr), &quot;ThreadIndex&quot;, &quot;threadIdx.x&quot;)] &quot;thread_extent&quot; = 64;
+        kernel.shared_1[(threadIdx.x_2 + 3712)] = kernel_3[(((((floordiv(blockIdx.x, 7)*589824) + (floordiv((threadIdx.x_2 + 3712), 72)*4608)) + cse_var_1) + (floordiv(floormod((threadIdx.x_2 + 40), 72), 3)*3)) + floormod((threadIdx.x_2 + 1), 3))]
+        attr [IterVar(threadIdx.x_2, (nullptr), &quot;ThreadIndex&quot;, &quot;threadIdx.x&quot;)] &quot;thread_extent&quot; = 64;
+        kernel.shared_1[(threadIdx.x_2 + 3776)] = kernel_3[(((((floordiv(blockIdx.x, 7)*589824) + (floordiv((threadIdx.x_2 + 3776), 72)*4608)) + cse_var_1) + (floordiv(floormod((threadIdx.x_2 + 32), 72), 3)*3)) + floormod((threadIdx.x_2 + 2), 3))]
+        attr [IterVar(threadIdx.x_2, (nullptr), &quot;ThreadIndex&quot;, &quot;threadIdx.x&quot;)] &quot;thread_extent&quot; = 64;
+        kernel.shared_1[(threadIdx.x_2 + 3840)] = kernel_3[(((((floordiv(blockIdx.x, 7)*589824) + (floordiv((threadIdx.x_2 + 3840), 72)*4608)) + cse_var_1) + (floormod((floordiv(threadIdx.x_2, 3) + 8), 24)*3)) + floormod(threadIdx.x_2, 3))]
+        attr [IterVar(threadIdx.x_2, (nullptr), &quot;ThreadIndex&quot;, &quot;threadIdx.x&quot;)] &quot;thread_extent&quot; = 64;
+        kernel.shared_1[(threadIdx.x_2 + 3904)] = kernel_3[(((((floordiv(blockIdx.x, 7)*589824) + (floordiv((threadIdx.x_2 + 3904), 72)*4608)) + cse_var_1) + (floordiv(floormod((threadIdx.x_2 + 16), 72), 3)*3)) + floormod((threadIdx.x_2 + 1), 3))]
+        attr [IterVar(threadIdx.x_2, (nullptr), &quot;ThreadIndex&quot;, &quot;threadIdx.x&quot;)] &quot;thread_extent&quot; = 64;
+        kernel.shared_1[(threadIdx.x_2 + 3968)] = kernel_3[(((((floordiv(blockIdx.x, 7)*589824) + (floordiv((threadIdx.x_2 + 3968), 72)*4608)) + cse_var_1) + (floordiv(floormod((threadIdx.x_2 + 8), 72), 3)*3)) + floormod((threadIdx.x_2 + 2), 3))]
+        attr [IterVar(threadIdx.x_2, (nullptr), &quot;ThreadIndex&quot;, &quot;threadIdx.x&quot;)] &quot;thread_extent&quot; = 64;
+        kernel.shared_1[(threadIdx.x_2 + 4032)] = kernel_3[((((floordiv(blockIdx.x, 7)*589824) + cse_var_1) + threadIdx.x_2) + 258048)]
+        attr [IterVar(threadIdx.x_2, (nullptr), &quot;ThreadIndex&quot;, &quot;threadIdx.x&quot;)] &quot;thread_extent&quot; = 64;
+        kernel.shared_1[(threadIdx.x_2 + 4096)] = kernel_3[(((((floordiv(blockIdx.x, 7)*589824) + (floordiv((threadIdx.x_2 + 4096), 72)*4608)) + cse_var_1) + (floordiv(floormod((threadIdx.x_2 + 64), 72), 3)*3)) + floormod((threadIdx.x_2 + 1), 3))]
+        attr [IterVar(threadIdx.x_2, (nullptr), &quot;ThreadIndex&quot;, &quot;threadIdx.x&quot;)] &quot;thread_extent&quot; = 64;
+        kernel.shared_1[(threadIdx.x_2 + 4160)] = kernel_3[(((((floordiv(blockIdx.x, 7)*589824) + (floordiv((threadIdx.x_2 + 4160), 72)*4608)) + cse_var_1) + (floordiv(floormod((threadIdx.x_2 + 56), 72), 3)*3)) + floormod((threadIdx.x_2 + 2), 3))]
+        attr [IterVar(threadIdx.x_2, (nullptr), &quot;ThreadIndex&quot;, &quot;threadIdx.x&quot;)] &quot;thread_extent&quot; = 64;
+        kernel.shared_1[(threadIdx.x_2 + 4224)] = kernel_3[(((((floordiv(blockIdx.x, 7)*589824) + (floordiv((threadIdx.x_2 + 4224), 72)*4608)) + cse_var_1) + (floormod((floordiv(threadIdx.x_2, 3) + 16), 24)*3)) + floormod(threadIdx.x_2, 3))]
+        attr [IterVar(threadIdx.x_2, (nullptr), &quot;ThreadIndex&quot;, &quot;threadIdx.x&quot;)] &quot;thread_extent&quot; = 64;
+        kernel.shared_1[(threadIdx.x_2 + 4288)] = kernel_3[(((((floordiv(blockIdx.x, 7)*589824) + (floordiv((threadIdx.x_2 + 4288), 72)*4608)) + cse_var_1) + (floordiv(floormod((threadIdx.x_2 + 40), 72), 3)*3)) + floormod((threadIdx.x_2 + 1), 3))]
+        attr [IterVar(threadIdx.x_2, (nullptr), &quot;ThreadIndex&quot;, &quot;threadIdx.x&quot;)] &quot;thread_extent&quot; = 64;
+        kernel.shared_1[(threadIdx.x_2 + 4352)] = kernel_3[(((((floordiv(blockIdx.x, 7)*589824) + (floordiv((threadIdx.x_2 + 4352), 72)*4608)) + cse_var_1) + (floordiv(floormod((threadIdx.x_2 + 32), 72), 3)*3)) + floormod((threadIdx.x_2 + 2), 3))]
+        attr [IterVar(threadIdx.x_2, (nullptr), &quot;ThreadIndex&quot;, &quot;threadIdx.x&quot;)] &quot;thread_extent&quot; = 64;
+        kernel.shared_1[(threadIdx.x_2 + 4416)] = kernel_3[(((((floordiv(blockIdx.x, 7)*589824) + (floordiv((threadIdx.x_2 + 4416), 72)*4608)) + cse_var_1) + (floormod((floordiv(threadIdx.x_2, 3) + 8), 24)*3)) + floormod(threadIdx.x_2, 3))]
+        attr [IterVar(threadIdx.x_2, (nullptr), &quot;ThreadIndex&quot;, &quot;threadIdx.x&quot;)] &quot;thread_extent&quot; = 64;
+        kernel.shared_1[(threadIdx.x_2 + 4480)] = kernel_3[(((((floordiv(blockIdx.x, 7)*589824) + (floordiv((threadIdx.x_2 + 4480), 72)*4608)) + cse_var_1) + (floordiv(floormod((threadIdx.x_2 + 16), 72), 3)*3)) + floormod((threadIdx.x_2 + 1), 3))]
+        attr [IterVar(threadIdx.x_2, (nullptr), &quot;ThreadIndex&quot;, &quot;threadIdx.x&quot;)] &quot;thread_extent&quot; = 64;
+        kernel.shared_1[(threadIdx.x_2 + 4544)] = kernel_3[(((((floordiv(blockIdx.x, 7)*589824) + (floordiv((threadIdx.x_2 + 4544), 72)*4608)) + cse_var_1) + (floordiv(floormod((threadIdx.x_2 + 8), 72), 3)*3)) + floormod((threadIdx.x_2 + 2), 3))]
+        attr [IterVar(threadIdx.x_2, (nullptr), &quot;ThreadIndex&quot;, &quot;threadIdx.x&quot;)] &quot;thread_extent&quot; = 64;
+        kernel.shared_1[(threadIdx.x_2 + 4608)] = kernel_3[((((floordiv(blockIdx.x, 7)*589824) + cse_var_1) + threadIdx.x_2) + 294912)]
+        attr [IterVar(threadIdx.x_2, (nullptr), &quot;ThreadIndex&quot;, &quot;threadIdx.x&quot;)] &quot;thread_extent&quot; = 64;
+        kernel.shared_1[(threadIdx.x_2 + 4672)] = kernel_3[(((((floordiv(blockIdx.x, 7)*589824) + (floordiv((threadIdx.x_2 + 4672), 72)*4608)) + cse_var_1) + (floordiv(floormod((threadIdx.x_2 + 64), 72), 3)*3)) + floormod((threadIdx.x_2 + 1), 3))]
+        attr [IterVar(threadIdx.x_2, (nullptr), &quot;ThreadIndex&quot;, &quot;threadIdx.x&quot;)] &quot;thread_extent&quot; = 64;
+        kernel.shared_1[(threadIdx.x_2 + 4736)] = kernel_3[(((((floordiv(blockIdx.x, 7)*589824) + (floordiv((threadIdx.x_2 + 4736), 72)*4608)) + cse_var_1) + (floordiv(floormod((threadIdx.x_2 + 56), 72), 3)*3)) + floormod((threadIdx.x_2 + 2), 3))]
+        attr [IterVar(threadIdx.x_2, (nullptr), &quot;ThreadIndex&quot;, &quot;threadIdx.x&quot;)] &quot;thread_extent&quot; = 64;
+        kernel.shared_1[(threadIdx.x_2 + 4800)] = kernel_3[(((((floordiv(blockIdx.x, 7)*589824) + (floordiv((threadIdx.x_2 + 4800), 72)*4608)) + cse_var_1) + (floormod((floordiv(threadIdx.x_2, 3) + 16), 24)*3)) + floormod(threadIdx.x_2, 3))]
+        attr [IterVar(threadIdx.x_2, (nullptr), &quot;ThreadIndex&quot;, &quot;threadIdx.x&quot;)] &quot;thread_extent&quot; = 64;
+        kernel.shared_1[(threadIdx.x_2 + 4864)] = kernel_3[(((((floordiv(blockIdx.x, 7)*589824) + (floordiv((threadIdx.x_2 + 4864), 72)*4608)) + cse_var_1) + (floordiv(floormod((threadIdx.x_2 + 40), 72), 3)*3)) + floormod((threadIdx.x_2 + 1), 3))]
+        attr [IterVar(threadIdx.x_2, (nullptr), &quot;ThreadIndex&quot;, &quot;threadIdx.x&quot;)] &quot;thread_extent&quot; = 64;
+        kernel.shared_1[(threadIdx.x_2 + 4928)] = kernel_3[(((((floordiv(blockIdx.x, 7)*589824) + (floordiv((threadIdx.x_2 + 4928), 72)*4608)) + cse_var_1) + (floordiv(floormod((threadIdx.x_2 + 32), 72), 3)*3)) + floormod((threadIdx.x_2 + 2), 3))]
+        attr [IterVar(threadIdx.x_2, (nullptr), &quot;ThreadIndex&quot;, &quot;threadIdx.x&quot;)] &quot;thread_extent&quot; = 64;
+        kernel.shared_1[(threadIdx.x_2 + 4992)] = kernel_3[(((((floordiv(blockIdx.x, 7)*589824) + (floordiv((threadIdx.x_2 + 4992), 72)*4608)) + cse_var_1) + (floormod((floordiv(threadIdx.x_2, 3) + 8), 24)*3)) + floormod(threadIdx.x_2, 3))]
+        attr [IterVar(threadIdx.x_2, (nullptr), &quot;ThreadIndex&quot;, &quot;threadIdx.x&quot;)] &quot;thread_extent&quot; = 64;
+        kernel.shared_1[(threadIdx.x_2 + 5056)] = kernel_3[(((((floordiv(blockIdx.x, 7)*589824) + (floordiv((threadIdx.x_2 + 5056), 72)*4608)) + cse_var_1) + (floordiv(floormod((threadIdx.x_2 + 16), 72), 3)*3)) + floormod((threadIdx.x_2 + 1), 3))]
+        attr [IterVar(threadIdx.x_2, (nullptr), &quot;ThreadIndex&quot;, &quot;threadIdx.x&quot;)] &quot;thread_extent&quot; = 64;
+        kernel.shared_1[(threadIdx.x_2 + 5120)] = kernel_3[(((((floordiv(blockIdx.x, 7)*589824) + (floordiv((threadIdx.x_2 + 5120), 72)*4608)) + cse_var_1) + (floordiv(floormod((threadIdx.x_2 + 8), 72), 3)*3)) + floormod((threadIdx.x_2 + 2), 3))]
+        attr [IterVar(threadIdx.x_2, (nullptr), &quot;ThreadIndex&quot;, &quot;threadIdx.x&quot;)] &quot;thread_extent&quot; = 64;
+        kernel.shared_1[(threadIdx.x_2 + 5184)] = kernel_3[((((floordiv(blockIdx.x, 7)*589824) + cse_var_1) + threadIdx.x_2) + 331776)]
+        attr [IterVar(threadIdx.x_2, (nullptr), &quot;ThreadIndex&quot;, &quot;threadIdx.x&quot;)] &quot;thread_extent&quot; = 64;
+        kernel.shared_1[(threadIdx.x_2 + 5248)] = kernel_3[(((((floordiv(blockIdx.x, 7)*589824) + (floordiv((threadIdx.x_2 + 5248), 72)*4608)) + cse_var_1) + (floordiv(floormod((threadIdx.x_2 + 64), 72), 3)*3)) + floormod((threadIdx.x_2 + 1), 3))]
+        attr [IterVar(threadIdx.x_2, (nullptr), &quot;ThreadIndex&quot;, &quot;threadIdx.x&quot;)] &quot;thread_extent&quot; = 64;
+        kernel.shared_1[(threadIdx.x_2 + 5312)] = kernel_3[(((((floordiv(blockIdx.x, 7)*589824) + (floordiv((threadIdx.x_2 + 5312), 72)*4608)) + cse_var_1) + (floordiv(floormod((threadIdx.x_2 + 56), 72), 3)*3)) + floormod((threadIdx.x_2 + 2), 3))]
+        attr [IterVar(threadIdx.x_2, (nullptr), &quot;ThreadIndex&quot;, &quot;threadIdx.x&quot;)] &quot;thread_extent&quot; = 64;
+        kernel.shared_1[(threadIdx.x_2 + 5376)] = kernel_3[(((((floordiv(blockIdx.x, 7)*589824) + (floordiv((threadIdx.x_2 + 5376), 72)*4608)) + cse_var_1) + (floormod((floordiv(threadIdx.x_2, 3) + 16), 24)*3)) + floormod(threadIdx.x_2, 3))]
+        attr [IterVar(threadIdx.x_2, (nullptr), &quot;ThreadIndex&quot;, &quot;threadIdx.x&quot;)] &quot;thread_extent&quot; = 64;
+        kernel.shared_1[(threadIdx.x_2 + 5440)] = kernel_3[(((((floordiv(blockIdx.x, 7)*589824) + (floordiv((threadIdx.x_2 + 5440), 72)*4608)) + cse_var_1) + (floordiv(floormod((threadIdx.x_2 + 40), 72), 3)*3)) + floormod((threadIdx.x_2 + 1), 3))]
+        attr [IterVar(threadIdx.x_2, (nullptr), &quot;ThreadIndex&quot;, &quot;threadIdx.x&quot;)] &quot;thread_extent&quot; = 64;
+        kernel.shared_1[(threadIdx.x_2 + 5504)] = kernel_3[(((((floordiv(blockIdx.x, 7)*589824) + (floordiv((threadIdx.x_2 + 5504), 72)*4608)) + cse_var_1) + (floordiv(floormod((threadIdx.x_2 + 32), 72), 3)*3)) + floormod((threadIdx.x_2 + 2), 3))]
+        attr [IterVar(threadIdx.x_2, (nullptr), &quot;ThreadIndex&quot;, &quot;threadIdx.x&quot;)] &quot;thread_extent&quot; = 64;
+        kernel.shared_1[(threadIdx.x_2 + 5568)] = kernel_3[(((((floordiv(blockIdx.x, 7)*589824) + (floordiv((threadIdx.x_2 + 5568), 72)*4608)) + cse_var_1) + (floormod((floordiv(threadIdx.x_2, 3) + 8), 24)*3)) + floormod(threadIdx.x_2, 3))]
+        attr [IterVar(threadIdx.x_2, (nullptr), &quot;ThreadIndex&quot;, &quot;threadIdx.x&quot;)] &quot;thread_extent&quot; = 64;
+        kernel.shared_1[(threadIdx.x_2 + 5632)] = kernel_3[(((((floordiv(blockIdx.x, 7)*589824) + (floordiv((threadIdx.x_2 + 5632), 72)*4608)) + cse_var_1) + (floordiv(floormod((threadIdx.x_2 + 16), 72), 3)*3)) + floormod((threadIdx.x_2 + 1), 3))]
+        attr [IterVar(threadIdx.x_2, (nullptr), &quot;ThreadIndex&quot;, &quot;threadIdx.x&quot;)] &quot;thread_extent&quot; = 64;
+        kernel.shared_1[(threadIdx.x_2 + 5696)] = kernel_3[(((((floordiv(blockIdx.x, 7)*589824) + (floordiv((threadIdx.x_2 + 5696), 72)*4608)) + cse_var_1) + (floordiv(floormod((threadIdx.x_2 + 8), 72), 3)*3)) + floormod((threadIdx.x_2 + 2), 3))]
+        attr [IterVar(threadIdx.x_2, (nullptr), &quot;ThreadIndex&quot;, &quot;threadIdx.x&quot;)] &quot;thread_extent&quot; = 64;
+        kernel.shared_1[(threadIdx.x_2 + 5760)] = kernel_3[((((floordiv(blockIdx.x, 7)*589824) + cse_var_1) + threadIdx.x_2) + 368640)]
+        attr [IterVar(threadIdx.x_2, (nullptr), &quot;ThreadIndex&quot;, &quot;threadIdx.x&quot;)] &quot;thread_extent&quot; = 64;
+        kernel.shared_1[(threadIdx.x_2 + 5824)] = kernel_3[(((((floordiv(blockIdx.x, 7)*589824) + (floordiv((threadIdx.x_2 + 5824), 72)*4608)) + cse_var_1) + (floordiv(floormod((threadIdx.x_2 + 64), 72), 3)*3)) + floormod((threadIdx.x_2 + 1), 3))]
+        attr [IterVar(threadIdx.x_2, (nullptr), &quot;ThreadIndex&quot;, &quot;threadIdx.x&quot;)] &quot;thread_extent&quot; = 64;
+        kernel.shared_1[(threadIdx.x_2 + 5888)] = kernel_3[(((((floordiv(blockIdx.x, 7)*589824) + (floordiv((threadIdx.x_2 + 5888), 72)*4608)) + cse_var_1) + (floordiv(floormod((threadIdx.x_2 + 56), 72), 3)*3)) + floormod((threadIdx.x_2 + 2), 3))]
+        attr [IterVar(threadIdx.x_2, (nullptr), &quot;ThreadIndex&quot;, &quot;threadIdx.x&quot;)] &quot;thread_extent&quot; = 64;
+        kernel.shared_1[(threadIdx.x_2 + 5952)] = kernel_3[(((((floordiv(blockIdx.x, 7)*589824) + (floordiv((threadIdx.x_2 + 5952), 72)*4608)) + cse_var_1) + (floormod((floordiv(threadIdx.x_2, 3) + 16), 24)*3)) + floormod(threadIdx.x_2, 3))]
+        attr [IterVar(threadIdx.x_2, (nullptr), &quot;ThreadIndex&quot;, &quot;threadIdx.x&quot;)] &quot;thread_extent&quot; = 64;
+        kernel.shared_1[(threadIdx.x_2 + 6016)] = kernel_3[(((((floordiv(blockIdx.x, 7)*589824) + (floordiv((threadIdx.x_2 + 6016), 72)*4608)) + cse_var_1) + (floordiv(floormod((threadIdx.x_2 + 40), 72), 3)*3)) + floormod((threadIdx.x_2 + 1), 3))]
+        attr [IterVar(threadIdx.x_2, (nullptr), &quot;ThreadIndex&quot;, &quot;threadIdx.x&quot;)] &quot;thread_extent&quot; = 64;
+        kernel.shared_1[(threadIdx.x_2 + 6080)] = kernel_3[(((((floordiv(blockIdx.x, 7)*589824) + (floordiv((threadIdx.x_2 + 6080), 72)*4608)) + cse_var_1) + (floordiv(floormod((threadIdx.x_2 + 32), 72), 3)*3)) + floormod((threadIdx.x_2 + 2), 3))]
+        attr [IterVar(threadIdx.x_2, (nullptr), &quot;ThreadIndex&quot;, &quot;threadIdx.x&quot;)] &quot;thread_extent&quot; = 64;
+        kernel.shared_1[(threadIdx.x_2 + 6144)] = kernel_3[(((((floordiv(blockIdx.x, 7)*589824) + (floordiv((threadIdx.x_2 + 6144), 72)*4608)) + cse_var_1) + (floormod((floordiv(threadIdx.x_2, 3) + 8), 24)*3)) + floormod(threadIdx.x_2, 3))]
+        attr [IterVar(threadIdx.x_2, (nullptr), &quot;ThreadIndex&quot;, &quot;threadIdx.x&quot;)] &quot;thread_extent&quot; = 64;
+        kernel.shared_1[(threadIdx.x_2 + 6208)] = kernel_3[(((((floordiv(blockIdx.x, 7)*589824) + (floordiv((threadIdx.x_2 + 6208), 72)*4608)) + cse_var_1) + (floordiv(floormod((threadIdx.x_2 + 16), 72), 3)*3)) + floormod((threadIdx.x_2 + 1), 3))]
+        attr [IterVar(threadIdx.x_2, (nullptr), &quot;ThreadIndex&quot;, &quot;threadIdx.x&quot;)] &quot;thread_extent&quot; = 64;
+        kernel.shared_1[(threadIdx.x_2 + 6272)] = kernel_3[(((((floordiv(blockIdx.x, 7)*589824) + (floordiv((threadIdx.x_2 + 6272), 72)*4608)) + cse_var_1) + (floordiv(floormod((threadIdx.x_2 + 8), 72), 3)*3)) + floormod((threadIdx.x_2 + 2), 3))]
+        attr [IterVar(threadIdx.x_2, (nullptr), &quot;ThreadIndex&quot;, &quot;threadIdx.x&quot;)] &quot;thread_extent&quot; = 64;
+        kernel.shared_1[(threadIdx.x_2 + 6336)] = kernel_3[((((floordiv(blockIdx.x, 7)*589824) + cse_var_1) + threadIdx.x_2) + 405504)]
+        attr [IterVar(threadIdx.x_2, (nullptr), &quot;ThreadIndex&quot;, &quot;threadIdx.x&quot;)] &quot;thread_extent&quot; = 64;
+        kernel.shared_1[(threadIdx.x_2 + 6400)] = kernel_3[(((((floordiv(blockIdx.x, 7)*589824) + (floordiv((threadIdx.x_2 + 6400), 72)*4608)) + cse_var_1) + (floordiv(floormod((threadIdx.x_2 + 64), 72), 3)*3)) + floormod((threadIdx.x_2 + 1), 3))]
+        attr [IterVar(threadIdx.x_2, (nullptr), &quot;ThreadIndex&quot;, &quot;threadIdx.x&quot;)] &quot;thread_extent&quot; = 64;
+        kernel.shared_1[(threadIdx.x_2 + 6464)] = kernel_3[(((((floordiv(blockIdx.x, 7)*589824) + (floordiv((threadIdx.x_2 + 6464), 72)*4608)) + cse_var_1) + (floordiv(floormod((threadIdx.x_2 + 56), 72), 3)*3)) + floormod((threadIdx.x_2 + 2), 3))]
+        attr [IterVar(threadIdx.x_2, (nullptr), &quot;ThreadIndex&quot;, &quot;threadIdx.x&quot;)] &quot;thread_extent&quot; = 64;
+        kernel.shared_1[(threadIdx.x_2 + 6528)] = kernel_3[(((((floordiv(blockIdx.x, 7)*589824) + (floordiv((threadIdx.x_2 + 6528), 72)*4608)) + cse_var_1) + (floormod((floordiv(threadIdx.x_2, 3) + 16), 24)*3)) + floormod(threadIdx.x_2, 3))]
+        attr [IterVar(threadIdx.x_2, (nullptr), &quot;ThreadIndex&quot;, &quot;threadIdx.x&quot;)] &quot;thread_extent&quot; = 64;
+        kernel.shared_1[(threadIdx.x_2 + 6592)] = kernel_3[(((((floordiv(blockIdx.x, 7)*589824) + (floordiv((threadIdx.x_2 + 6592), 72)*4608)) + cse_var_1) + (floordiv(floormod((threadIdx.x_2 + 40), 72), 3)*3)) + floormod((threadIdx.x_2 + 1), 3))]
+        attr [IterVar(threadIdx.x_2, (nullptr), &quot;ThreadIndex&quot;, &quot;threadIdx.x&quot;)] &quot;thread_extent&quot; = 64;
+        kernel.shared_1[(threadIdx.x_2 + 6656)] = kernel_3[(((((floordiv(blockIdx.x, 7)*589824) + (floordiv((threadIdx.x_2 + 6656), 72)*4608)) + cse_var_1) + (floordiv(floormod((threadIdx.x_2 + 32), 72), 3)*3)) + floormod((threadIdx.x_2 + 2), 3))]
+        attr [IterVar(threadIdx.x_2, (nullptr), &quot;ThreadIndex&quot;, &quot;threadIdx.x&quot;)] &quot;thread_extent&quot; = 64;
+        kernel.shared_1[(threadIdx.x_2 + 6720)] = kernel_3[(((((floordiv(blockIdx.x, 7)*589824) + (floordiv((threadIdx.x_2 + 6720), 72)*4608)) + cse_var_1) + (floormod((floordiv(threadIdx.x_2, 3) + 8), 24)*3)) + floormod(threadIdx.x_2, 3))]
+        attr [IterVar(threadIdx.x_2, (nullptr), &quot;ThreadIndex&quot;, &quot;threadIdx.x&quot;)] &quot;thread_extent&quot; = 64;
+        kernel.shared_1[(threadIdx.x_2 + 6784)] = kernel_3[(((((floordiv(blockIdx.x, 7)*589824) + (floordiv((threadIdx.x_2 + 6784), 72)*4608)) + cse_var_1) + (floordiv(floormod((threadIdx.x_2 + 16), 72), 3)*3)) + floormod((threadIdx.x_2 + 1), 3))]
+        attr [IterVar(threadIdx.x_2, (nullptr), &quot;ThreadIndex&quot;, &quot;threadIdx.x&quot;)] &quot;thread_extent&quot; = 64;
+        kernel.shared_1[(threadIdx.x_2 + 6848)] = kernel_3[(((((floordiv(blockIdx.x, 7)*589824) + (floordiv((threadIdx.x_2 + 6848), 72)*4608)) + cse_var_1) + (floordiv(floormod((threadIdx.x_2 + 8), 72), 3)*3)) + floormod((threadIdx.x_2 + 2), 3))]
+        attr [IterVar(threadIdx.x_2, (nullptr), &quot;ThreadIndex&quot;, &quot;threadIdx.x&quot;)] &quot;thread_extent&quot; = 64;
+        kernel.shared_1[(threadIdx.x_2 + 6912)] = kernel_3[((((floordiv(blockIdx.x, 7)*589824) + cse_var_1) + threadIdx.x_2) + 442368)]
+        attr [IterVar(threadIdx.x_2, (nullptr), &quot;ThreadIndex&quot;, &quot;threadIdx.x&quot;)] &quot;thread_extent&quot; = 64;
+        kernel.shared_1[(threadIdx.x_2 + 6976)] = kernel_3[(((((floordiv(blockIdx.x, 7)*589824) + (floordiv((threadIdx.x_2 + 6976), 72)*4608)) + cse_var_1) + (floordiv(floormod((threadIdx.x_2 + 64), 72), 3)*3)) + floormod((threadIdx.x_2 + 1), 3))]
+        attr [IterVar(threadIdx.x_2, (nullptr), &quot;ThreadIndex&quot;, &quot;threadIdx.x&quot;)] &quot;thread_extent&quot; = 64;
+        kernel.shared_1[(threadIdx.x_2 + 7040)] = kernel_3[(((((floordiv(blockIdx.x, 7)*589824) + (floordiv((threadIdx.x_2 + 7040), 72)*4608)) + cse_var_1) + (floordiv(floormod((threadIdx.x_2 + 56), 72), 3)*3)) + floormod((threadIdx.x_2 + 2), 3))]
+        attr [IterVar(threadIdx.x_2, (nullptr), &quot;ThreadIndex&quot;, &quot;threadIdx.x&quot;)] &quot;thread_extent&quot; = 64;
+        kernel.shared_1[(threadIdx.x_2 + 7104)] = kernel_3[(((((floordiv(blockIdx.x, 7)*589824) + (floordiv((threadIdx.x_2 + 7104), 72)*4608)) + cse_var_1) + (floormod((floordiv(threadIdx.x_2, 3) + 16), 24)*3)) + floormod(threadIdx.x_2, 3))]
+        attr [IterVar(threadIdx.x_2, (nullptr), &quot;ThreadIndex&quot;, &quot;threadIdx.x&quot;)] &quot;thread_extent&quot; = 64;
+        kernel.shared_1[(threadIdx.x_2 + 7168)] = kernel_3[(((((floordiv(blockIdx.x, 7)*589824) + (floordiv((threadIdx.x_2 + 7168), 72)*4608)) + cse_var_1) + (floordiv(floormod((threadIdx.x_2 + 40), 72), 3)*3)) + floormod((threadIdx.x_2 + 1), 3))]
+        attr [IterVar(threadIdx.x_2, (nullptr), &quot;ThreadIndex&quot;, &quot;threadIdx.x&quot;)] &quot;thread_extent&quot; = 64;
+        kernel.shared_1[(threadIdx.x_2 + 7232)] = kernel_3[(((((floordiv(blockIdx.x, 7)*589824) + (floordiv((threadIdx.x_2 + 7232), 72)*4608)) + cse_var_1) + (floordiv(floormod((threadIdx.x_2 + 32), 72), 3)*3)) + floormod((threadIdx.x_2 + 2), 3))]
+        attr [IterVar(threadIdx.x_2, (nullptr), &quot;ThreadIndex&quot;, &quot;threadIdx.x&quot;)] &quot;thread_extent&quot; = 64;
+        kernel.shared_1[(threadIdx.x_2 + 7296)] = kernel_3[(((((floordiv(blockIdx.x, 7)*589824) + (floordiv((threadIdx.x_2 + 7296), 72)*4608)) + cse_var_1) + (floormod((floordiv(threadIdx.x_2, 3) + 8), 24)*3)) + floormod(threadIdx.x_2, 3))]
+        attr [IterVar(threadIdx.x_2, (nullptr), &quot;ThreadIndex&quot;, &quot;threadIdx.x&quot;)] &quot;thread_extent&quot; = 64;
+        kernel.shared_1[(threadIdx.x_2 + 7360)] = kernel_3[(((((floordiv(blockIdx.x, 7)*589824) + (floordiv((threadIdx.x_2 + 7360), 72)*4608)) + cse_var_1) + (floordiv(floormod((threadIdx.x_2 + 16), 72), 3)*3)) + floormod((threadIdx.x_2 + 1), 3))]
+        attr [IterVar(threadIdx.x_2, (nullptr), &quot;ThreadIndex&quot;, &quot;threadIdx.x&quot;)] &quot;thread_extent&quot; = 64;
+        kernel.shared_1[(threadIdx.x_2 + 7424)] = kernel_3[(((((floordiv(blockIdx.x, 7)*589824) + (floordiv((threadIdx.x_2 + 7424), 72)*4608)) + cse_var_1) + (floordiv(floormod((threadIdx.x_2 + 8), 72), 3)*3)) + floormod((threadIdx.x_2 + 2), 3))]
+        attr [IterVar(threadIdx.x_2, (nullptr), &quot;ThreadIndex&quot;, &quot;threadIdx.x&quot;)] &quot;thread_extent&quot; = 64;
+        kernel.shared_1[(threadIdx.x_2 + 7488)] = kernel_3[((((floordiv(blockIdx.x, 7)*589824) + cse_var_1) + threadIdx.x_2) + 479232)]
+        attr [IterVar(threadIdx.x_2, (nullptr), &quot;ThreadIndex&quot;, &quot;threadIdx.x&quot;)] &quot;thread_extent&quot; = 64;
+        kernel.shared_1[(threadIdx.x_2 + 7552)] = kernel_3[(((((floordiv(blockIdx.x, 7)*589824) + (floordiv((threadIdx.x_2 + 7552), 72)*4608)) + cse_var_1) + (floordiv(floormod((threadIdx.x_2 + 64), 72), 3)*3)) + floormod((threadIdx.x_2 + 1), 3))]
+        attr [IterVar(threadIdx.x_2, (nullptr), &quot;ThreadIndex&quot;, &quot;threadIdx.x&quot;)] &quot;thread_extent&quot; = 64;
+        kernel.shared_1[(threadIdx.x_2 + 7616)] = kernel_3[(((((floordiv(blockIdx.x, 7)*589824) + (floordiv((threadIdx.x_2 + 7616), 72)*4608)) + cse_var_1) + (floordiv(floormod((threadIdx.x_2 + 56), 72), 3)*3)) + floormod((threadIdx.x_2 + 2), 3))]
+        attr [IterVar(threadIdx.x_2, (nullptr), &quot;ThreadIndex&quot;, &quot;threadIdx.x&quot;)] &quot;thread_extent&quot; = 64;
+        kernel.shared_1[(threadIdx.x_2 + 7680)] = kernel_3[(((((floordiv(blockIdx.x, 7)*589824) + (floordiv((threadIdx.x_2 + 7680), 72)*4608)) + cse_var_1) + (floormod((floordiv(threadIdx.x_2, 3) + 16), 24)*3)) + floormod(threadIdx.x_2, 3))]
+        attr [IterVar(threadIdx.x_2, (nullptr), &quot;ThreadIndex&quot;, &quot;threadIdx.x&quot;)] &quot;thread_extent&quot; = 64;
+        kernel.shared_1[(threadIdx.x_2 + 7744)] = kernel_3[(((((floordiv(blockIdx.x, 7)*589824) + (floordiv((threadIdx.x_2 + 7744), 72)*4608)) + cse_var_1) + (floordiv(floormod((threadIdx.x_2 + 40), 72), 3)*3)) + floormod((threadIdx.x_2 + 1), 3))]
+        attr [IterVar(threadIdx.x_2, (nullptr), &quot;ThreadIndex&quot;, &quot;threadIdx.x&quot;)] &quot;thread_extent&quot; = 64;
+        kernel.shared_1[(threadIdx.x_2 + 7808)] = kernel_3[(((((floordiv(blockIdx.x, 7)*589824) + (floordiv((threadIdx.x_2 + 7808), 72)*4608)) + cse_var_1) + (floordiv(floormod((threadIdx.x_2 + 32), 72), 3)*3)) + floormod((threadIdx.x_2 + 2), 3))]
+        attr [IterVar(threadIdx.x_2, (nullptr), &quot;ThreadIndex&quot;, &quot;threadIdx.x&quot;)] &quot;thread_extent&quot; = 64;
+        kernel.shared_1[(threadIdx.x_2 + 7872)] = kernel_3[(((((floordiv(blockIdx.x, 7)*589824) + (floordiv((threadIdx.x_2 + 7872), 72)*4608)) + cse_var_1) + (floormod((floordiv(threadIdx.x_2, 3) + 8), 24)*3)) + floormod(threadIdx.x_2, 3))]
+        attr [IterVar(threadIdx.x_2, (nullptr), &quot;ThreadIndex&quot;, &quot;threadIdx.x&quot;)] &quot;thread_extent&quot; = 64;
+        kernel.shared_1[(threadIdx.x_2 + 7936)] = kernel_3[(((((floordiv(blockIdx.x, 7)*589824) + (floordiv((threadIdx.x_2 + 7936), 72)*4608)) + cse_var_1) + (floordiv(floormod((threadIdx.x_2 + 16), 72), 3)*3)) + floormod((threadIdx.x_2 + 1), 3))]
+        attr [IterVar(threadIdx.x_2, (nullptr), &quot;ThreadIndex&quot;, &quot;threadIdx.x&quot;)] &quot;thread_extent&quot; = 64;
+        kernel.shared_1[(threadIdx.x_2 + 8000)] = kernel_3[(((((floordiv(blockIdx.x, 7)*589824) + (floordiv((threadIdx.x_2 + 8000), 72)*4608)) + cse_var_1) + (floordiv(floormod((threadIdx.x_2 + 8), 72), 3)*3)) + floormod((threadIdx.x_2 + 2), 3))]
+        attr [IterVar(threadIdx.x_2, (nullptr), &quot;ThreadIndex&quot;, &quot;threadIdx.x&quot;)] &quot;thread_extent&quot; = 64;
+        kernel.shared_1[(threadIdx.x_2 + 8064)] = kernel_3[((((floordiv(blockIdx.x, 7)*589824) + cse_var_1) + threadIdx.x_2) + 516096)]
+        attr [IterVar(threadIdx.x_2, (nullptr), &quot;ThreadIndex&quot;, &quot;threadIdx.x&quot;)] &quot;thread_extent&quot; = 64;
+        kernel.shared_1[(threadIdx.x_2 + 8128)] = kernel_3[(((((floordiv(blockIdx.x, 7)*589824) + (floordiv((threadIdx.x_2 + 8128), 72)*4608)) + cse_var_1) + (floordiv(floormod((threadIdx.x_2 + 64), 72), 3)*3)) + floormod((threadIdx.x_2 + 1), 3))]
+        attr [IterVar(threadIdx.x_2, (nullptr), &quot;ThreadIndex&quot;, &quot;threadIdx.x&quot;)] &quot;thread_extent&quot; = 64;
+        kernel.shared_1[(threadIdx.x_2 + 8192)] = kernel_3[(((((floordiv(blockIdx.x, 7)*589824) + (floordiv((threadIdx.x_2 + 8192), 72)*4608)) + cse_var_1) + (floordiv(floormod((threadIdx.x_2 + 56), 72), 3)*3)) + floormod((threadIdx.x_2 + 2), 3))]
+        attr [IterVar(threadIdx.x_2, (nullptr), &quot;ThreadIndex&quot;, &quot;threadIdx.x&quot;)] &quot;thread_extent&quot; = 64;
+        kernel.shared_1[(threadIdx.x_2 + 8256)] = kernel_3[(((((floordiv(blockIdx.x, 7)*589824) + (floordiv((threadIdx.x_2 + 8256), 72)*4608)) + cse_var_1) + (floormod((floordiv(threadIdx.x_2, 3) + 16), 24)*3)) + floormod(threadIdx.x_2, 3))]
+        attr [IterVar(threadIdx.x_2, (nullptr), &quot;ThreadIndex&quot;, &quot;threadIdx.x&quot;)] &quot;thread_extent&quot; = 64;
+        kernel.shared_1[(threadIdx.x_2 + 8320)] = kernel_3[(((((floordiv(blockIdx.x, 7)*589824) + (floordiv((threadIdx.x_2 + 8320), 72)*4608)) + cse_var_1) + (floordiv(floormod((threadIdx.x_2 + 40), 72), 3)*3)) + floormod((threadIdx.x_2 + 1), 3))]
+        attr [IterVar(threadIdx.x_2, (nullptr), &quot;ThreadIndex&quot;, &quot;threadIdx.x&quot;)] &quot;thread_extent&quot; = 64;
+        kernel.shared_1[(threadIdx.x_2 + 8384)] = kernel_3[(((((floordiv(blockIdx.x, 7)*589824) + (floordiv((threadIdx.x_2 + 8384), 72)*4608)) + cse_var_1) + (floordiv(floormod((threadIdx.x_2 + 32), 72), 3)*3)) + floormod((threadIdx.x_2 + 2), 3))]
+        attr [IterVar(threadIdx.x_2, (nullptr), &quot;ThreadIndex&quot;, &quot;threadIdx.x&quot;)] &quot;thread_extent&quot; = 64;
+        kernel.shared_1[(threadIdx.x_2 + 8448)] = kernel_3[(((((floordiv(blockIdx.x, 7)*589824) + (floordiv((threadIdx.x_2 + 8448), 72)*4608)) + cse_var_1) + (floormod((floordiv(threadIdx.x_2, 3) + 8), 24)*3)) + floormod(threadIdx.x_2, 3))]
+        attr [IterVar(threadIdx.x_2, (nullptr), &quot;ThreadIndex&quot;, &quot;threadIdx.x&quot;)] &quot;thread_extent&quot; = 64;
+        kernel.shared_1[(threadIdx.x_2 + 8512)] = kernel_3[(((((floordiv(blockIdx.x, 7)*589824) + (floordiv((threadIdx.x_2 + 8512), 72)*4608)) + cse_var_1) + (floordiv(floormod((threadIdx.x_2 + 16), 72), 3)*3)) + floormod((threadIdx.x_2 + 1), 3))]
+        attr [IterVar(threadIdx.x_2, (nullptr), &quot;ThreadIndex&quot;, &quot;threadIdx.x&quot;)] &quot;thread_extent&quot; = 64;
+        kernel.shared_1[(threadIdx.x_2 + 8576)] = kernel_3[(((((floordiv(blockIdx.x, 7)*589824) + (floordiv((threadIdx.x_2 + 8576), 72)*4608)) + cse_var_1) + (floordiv(floormod((threadIdx.x_2 + 8), 72), 3)*3)) + floormod((threadIdx.x_2 + 2), 3))]
+        attr [IterVar(threadIdx.x_2, (nullptr), &quot;ThreadIndex&quot;, &quot;threadIdx.x&quot;)] &quot;thread_extent&quot; = 64;
+        kernel.shared_1[(threadIdx.x_2 + 8640)] = kernel_3[((((floordiv(blockIdx.x, 7)*589824) + cse_var_1) + threadIdx.x_2) + 552960)]
+        attr [IterVar(threadIdx.x_2, (nullptr), &quot;ThreadIndex&quot;, &quot;threadIdx.x&quot;)] &quot;thread_extent&quot; = 64;
+        kernel.shared_1[(threadIdx.x_2 + 8704)] = kernel_3[(((((floordiv(blockIdx.x, 7)*589824) + (floordiv((threadIdx.x_2 + 8704), 72)*4608)) + cse_var_1) + (floordiv(floormod((threadIdx.x_2 + 64), 72), 3)*3)) + floormod((threadIdx.x_2 + 1), 3))]
+        attr [IterVar(threadIdx.x_2, (nullptr), &quot;ThreadIndex&quot;, &quot;threadIdx.x&quot;)] &quot;thread_extent&quot; = 64;
+        kernel.shared_1[(threadIdx.x_2 + 8768)] = kernel_3[(((((floordiv(blockIdx.x, 7)*589824) + (floordiv((threadIdx.x_2 + 8768), 72)*4608)) + cse_var_1) + (floordiv(floormod((threadIdx.x_2 + 56), 72), 3)*3)) + floormod((threadIdx.x_2 + 2), 3))]
+        attr [IterVar(threadIdx.x_2, (nullptr), &quot;ThreadIndex&quot;, &quot;threadIdx.x&quot;)] &quot;thread_extent&quot; = 64;
+        kernel.shared_1[(threadIdx.x_2 + 8832)] = kernel_3[(((((floordiv(blockIdx.x, 7)*589824) + (floordiv((threadIdx.x_2 + 8832), 72)*4608)) + cse_var_1) + (floormod((floordiv(threadIdx.x_2, 3) + 16), 24)*3)) + floormod(threadIdx.x_2, 3))]
+        attr [IterVar(threadIdx.x_2, (nullptr), &quot;ThreadIndex&quot;, &quot;threadIdx.x&quot;)] &quot;thread_extent&quot; = 64;
+        kernel.shared_1[(threadIdx.x_2 + 8896)] = kernel_3[(((((floordiv(blockIdx.x, 7)*589824) + (floordiv((threadIdx.x_2 + 8896), 72)*4608)) + cse_var_1) + (floordiv(floormod((threadIdx.x_2 + 40), 72), 3)*3)) + floormod((threadIdx.x_2 + 1), 3))]
+        attr [IterVar(threadIdx.x_2, (nullptr), &quot;ThreadIndex&quot;, &quot;threadIdx.x&quot;)] &quot;thread_extent&quot; = 64;
+        kernel.shared_1[(threadIdx.x_2 + 8960)] = kernel_3[(((((floordiv(blockIdx.x, 7)*589824) + (floordiv((threadIdx.x_2 + 8960), 72)*4608)) + cse_var_1) + (floordiv(floormod((threadIdx.x_2 + 32), 72), 3)*3)) + floormod((threadIdx.x_2 + 2), 3))]
+        attr [IterVar(threadIdx.x_2, (nullptr), &quot;ThreadIndex&quot;, &quot;threadIdx.x&quot;)] &quot;thread_extent&quot; = 64;
+        kernel.shared_1[(threadIdx.x_2 + 9024)] = kernel_3[(((((floordiv(blockIdx.x, 7)*589824) + (floordiv((threadIdx.x_2 + 9024), 72)*4608)) + cse_var_1) + (floormod((floordiv(threadIdx.x_2, 3) + 8), 24)*3)) + floormod(threadIdx.x_2, 3))]
+        attr [IterVar(threadIdx.x_2, (nullptr), &quot;ThreadIndex&quot;, &quot;threadIdx.x&quot;)] &quot;thread_extent&quot; = 64;
+        kernel.shared_1[(threadIdx.x_2 + 9088)] = kernel_3[(((((floordiv(blockIdx.x, 7)*589824) + (floordiv((threadIdx.x_2 + 9088), 72)*4608)) + cse_var_1) + (floordiv(floormod((threadIdx.x_2 + 16), 72), 3)*3)) + floormod((threadIdx.x_2 + 1), 3))]
+        attr [IterVar(threadIdx.x_2, (nullptr), &quot;ThreadIndex&quot;, &quot;threadIdx.x&quot;)] &quot;thread_extent&quot; = 64;
+        kernel.shared_1[(threadIdx.x_2 + 9152)] = kernel_3[(((((floordiv(blockIdx.x, 7)*589824) + (floordiv((threadIdx.x_2 + 9152), 72)*4608)) + cse_var_1) + (floordiv(floormod((threadIdx.x_2 + 8), 72), 3)*3)) + floormod((threadIdx.x_2 + 2), 3))]
+        conv2d_nchw_1[0] = (conv2d_nchw_1[0] + (pad_temp.shared_1[0]*kernel.shared_1[(threadIdx.x*72)]))
+        conv2d_nchw_1[1] = (conv2d_nchw_1[1] + (pad_temp.shared_1[1]*kernel.shared_1[(threadIdx.x*72)]))
+        conv2d_nchw_1[2] = (conv2d_nchw_1[2] + (pad_temp.shared_1[2]*kernel.shared_1[(threadIdx.x*72)]))
+        conv2d_nchw_1[3] = (conv2d_nchw_1[3] + (pad_temp.shared_1[3]*kernel.shared_1[(threadIdx.x*72)]))
+        conv2d_nchw_1[4] = (conv2d_nchw_1[4] + (pad_temp.shared_1[4]*kernel.shared_1[(threadIdx.x*72)]))
+        conv2d_nchw_1[5] = (conv2d_nchw_1[5] + (pad_temp.shared_1[5]*kernel.shared_1[(threadIdx.x*72)]))
+        conv2d_nchw_1[6] = (conv2d_nchw_1[6] + (pad_temp.shared_1[6]*kernel.shared_1[(threadIdx.x*72)]))
+        conv2d_nchw_1[7] = (conv2d_nchw_1[7] + (pad_temp.shared_1[0]*kernel.shared_1[((threadIdx.x*72) + 4608)]))
+        conv2d_nchw_1[8] = (conv2d_nchw_1[8] + (pad_temp.shared_1[1]*kernel.shared_1[((threadIdx.x*72) + 4608)]))
+        conv2d_nchw_1[9] = (conv2d_nchw_1[9] + (pad_temp.shared_1[2]*kernel.shared_1[((threadIdx.x*72) + 4608)]))
+        conv2d_nchw_1[10] = (conv2d_nchw_1[10] + (pad_temp.shared_1[3]*kernel.shared_1[((threadIdx.x*72) + 4608)]))
+        conv2d_nchw_1[11] = (conv2d_nchw_1[11] + (pad_temp.shared_1[4]*kernel.shared_1[((threadIdx.x*72) + 4608)]))
+        conv2d_nchw_1[12] = (conv2d_nchw_1[12] + (pad_temp.shared_1[5]*kernel.shared_1[((threadIdx.x*72) + 4608)]))
+        conv2d_nchw_1[13] = (conv2d_nchw_1[13] + (pad_temp.shared_1[6]*kernel.shared_1[((threadIdx.x*72) + 4608)]))
+        conv2d_nchw_1[0] = (conv2d_nchw_1[0] + (pad_temp.shared_1[1]*kernel.shared_1[((threadIdx.x*72) + 1)]))
+        conv2d_nchw_1[1] = (conv2d_nchw_1[1] + (pad_temp.shared_1[2]*kernel.shared_1[((threadIdx.x*72) + 1)]))
+        conv2d_nchw_1[2] = (conv2d_nchw_1[2] + (pad_temp.shared_1[3]*kernel.shared_1[((threadIdx.x*72) + 1)]))
+        conv2d_nchw_1[3] = (conv2d_nchw_1[3] + (pad_temp.shared_1[4]*kernel.shared_1[((threadIdx.x*72) + 1)]))
+        conv2d_nchw_1[4] = (conv2d_nchw_1[4] + (pad_temp.shared_1[5]*kernel.shared_1[((threadIdx.x*72) + 1)]))
+        conv2d_nchw_1[5] = (conv2d_nchw_1[5] + (pad_temp.shared_1[6]*kernel.shared_1[((threadIdx.x*72) + 1)]))
+        conv2d_nchw_1[6] = (conv2d_nchw_1[6] + (pad_temp.shared_1[7]*kernel.shared_1[((threadIdx.x*72) + 1)]))
+        conv2d_nchw_1[7] = (conv2d_nchw_1[7] + (pad_temp.shared_1[1]*kernel.shared_1[((threadIdx.x*72) + 4609)]))
+        conv2d_nchw_1[8] = (conv2d_nchw_1[8] + (pad_temp.shared_1[2]*kernel.shared_1[((threadIdx.x*72) + 4609)]))
+        conv2d_nchw_1[9] = (conv2d_nchw_1[9] + (pad_temp.shared_1[3]*kernel.shared_1[((threadIdx.x*72) + 4609)]))
+        conv2d_nchw_1[10] = (conv2d_nchw_1[10] + (pad_temp.shared_1[4]*kernel.shared_1[((threadIdx.x*72) + 4609)]))
+        conv2d_nchw_1[11] = (conv2d_nchw_1[11] + (pad_temp.shared_1[5]*kernel.shared_1[((threadIdx.x*72) + 4609)]))
+        conv2d_nchw_1[12] = (conv2d_nchw_1[12] + (pad_temp.shared_1[6]*kernel.shared_1[((threadIdx.x*72) + 4609)]))
+        conv2d_nchw_1[13] = (conv2d_nchw_1[13] + (pad_temp.shared_1[7]*kernel.shared_1[((threadIdx.x*72) + 4609)]))
+        conv2d_nchw_1[0] = (conv2d_nchw_1[0] + (pad_temp.shared_1[2]*kernel.shared_1[((threadIdx.x*72) + 2)]))
+        conv2d_nchw_1[1] = (conv2d_nchw_1[1] + (pad_temp.shared_1[3]*kernel.shared_1[((threadIdx.x*72) + 2)]))
+        conv2d_nchw_1[2] = (conv2d_nchw_1[2] + (pad_temp.shared_1[4]*kernel.shared_1[((threadIdx.x*72) + 2)]))
+        conv2d_nchw_1[3] = (conv2d_nchw_1[3] + (pad_temp.shared_1[5]*kernel.shared_1[((threadIdx.x*72) + 2)]))
+        conv2d_nchw_1[4] = (conv2d_nchw_1[4] + (pad_temp.shared_1[6]*kernel.shared_1[((threadIdx.x*72) + 2)]))
+        conv2d_nchw_1[5] = (conv2d_nchw_1[5] + (pad_temp.shared_1[7]*kernel.shared_1[((threadIdx.x*72) + 2)]))
+        conv2d_nchw_1[6] = (conv2d_nchw_1[6] + (pad_temp.shared_1[8]*kernel.shared_1[((threadIdx.x*72) + 2)]))
+        conv2d_nchw_1[7] = (conv2d_nchw_1[7] + (pad_temp.shared_1[2]*kernel.shared_1[((threadIdx.x*72) + 4610)]))
+        conv2d_nchw_1[8] = (conv2d_nchw_1[8] + (pad_temp.shared_1[3]*kernel.shared_1[((threadIdx.x*72) + 4610)]))
+        conv2d_nchw_1[9] = (conv2d_nchw_1[9] + (pad_temp.shared_1[4]*kernel.shared_1[((threadIdx.x*72) + 4610)]))
+        conv2d_nchw_1[10] = (conv2d_nchw_1[10] + (pad_temp.shared_1[5]*kernel.shared_1[((threadIdx.x*72) + 4610)]))
+        conv2d_nchw_1[11] = (conv2d_nchw_1[11] + (pad_temp.shared_1[6]*kernel.shared_1[((threadIdx.x*72) + 4610)]))
+        conv2d_nchw_1[12] = (conv2d_nchw_1[12] + (pad_temp.shared_1[7]*kernel.shared_1[((threadIdx.x*72) + 4610)]))
+        conv2d_nchw_1[13] = (conv2d_nchw_1[13] + (pad_temp.shared_1[8]*kernel.shared_1[((threadIdx.x*72) + 4610)]))
+        conv2d_nchw_1[0] = (conv2d_nchw_1[0] + (pad_temp.shared_1[9]*kernel.shared_1[((threadIdx.x*72) + 3)]))
+        conv2d_nchw_1[1] = (conv2d_nchw_1[1] + (pad_temp.shared_1[10]*kernel.shared_1[((threadIdx.x*72) + 3)]))
+        conv2d_nchw_1[2] = (conv2d_nchw_1[2] + (pad_temp.shared_1[11]*kernel.shared_1[((threadIdx.x*72) + 3)]))
+        conv2d_nchw_1[3] = (conv2d_nchw_1[3] + (pad_temp.shared_1[12]*kernel.shared_1[((threadIdx.x*72) + 3)]))
+        conv2d_nchw_1[4] = (conv2d_nchw_1[4] + (pad_temp.shared_1[13]*kernel.shared_1[((threadIdx.x*72) + 3)]))
+        conv2d_nchw_1[5] = (conv2d_nchw_1[5] + (pad_temp.shared_1[14]*kernel.shared_1[((threadIdx.x*72) + 3)]))
+        conv2d_nchw_1[6] = (conv2d_nchw_1[6] + (pad_temp.shared_1[15]*kernel.shared_1[((threadIdx.x*72) + 3)]))
+        conv2d_nchw_1[7] = (conv2d_nchw_1[7] + (pad_temp.shared_1[9]*kernel.shared_1[((threadIdx.x*72) + 4611)]))
+        conv2d_nchw_1[8] = (conv2d_nchw_1[8] + (pad_temp.shared_1[10]*kernel.shared_1[((threadIdx.x*72) + 4611)]))
+        conv2d_nchw_1[9] = (conv2d_nchw_1[9] + (pad_temp.shared_1[11]*kernel.shared_1[((threadIdx.x*72) + 4611)]))
+        conv2d_nchw_1[10] = (conv2d_nchw_1[10] + (pad_temp.shared_1[12]*kernel.shared_1[((threadIdx.x*72) + 4611)]))
+        conv2d_nchw_1[11] = (conv2d_nchw_1[11] + (pad_temp.shared_1[13]*kernel.shared_1[((threadIdx.x*72) + 4611)]))
+        conv2d_nchw_1[12] = (conv2d_nchw_1[12] + (pad_temp.shared_1[14]*kernel.shared_1[((threadIdx.x*72) + 4611)]))
+        conv2d_nchw_1[13] = (conv2d_nchw_1[13] + (pad_temp.shared_1[15]*kernel.shared_1[((threadIdx.x*72) + 4611)]))
+        conv2d_nchw_1[0] = (conv2d_nchw_1[0] + (pad_temp.shared_1[10]*kernel.shared_1[((threadIdx.x*72) + 4)]))
+        conv2d_nchw_1[1] = (conv2d_nchw_1[1] + (pad_temp.shared_1[11]*kernel.shared_1[((threadIdx.x*72) + 4)]))
+        conv2d_nchw_1[2] = (conv2d_nchw_1[2] + (pad_temp.shared_1[12]*kernel.shared_1[((threadIdx.x*72) + 4)]))
+        conv2d_nchw_1[3] = (conv2d_nchw_1[3] + (pad_temp.shared_1[13]*kernel.shared_1[((threadIdx.x*72) + 4)]))
+        conv2d_nchw_1[4] = (conv2d_nchw_1[4] + (pad_temp.shared_1[14]*kernel.shared_1[((threadIdx.x*72) + 4)]))
+        conv2d_nchw_1[5] = (conv2d_nchw_1[5] + (pad_temp.shared_1[15]*kernel.shared_1[((threadIdx.x*72) + 4)]))
+        conv2d_nchw_1[6] = (conv2d_nchw_1[6] + (pad_temp.shared_1[16]*kernel.shared_1[((threadIdx.x*72) + 4)]))
+        conv2d_nchw_1[7] = (conv2d_nchw_1[7] + (pad_temp.shared_1[10]*kernel.shared_1[((threadIdx.x*72) + 4612)]))
+        conv2d_nchw_1[8] = (conv2d_nchw_1[8] + (pad_temp.shared_1[11]*kernel.shared_1[((threadIdx.x*72) + 4612)]))
+        conv2d_nchw_1[9] = (conv2d_nchw_1[9] + (pad_temp.shared_1[12]*kernel.shared_1[((threadIdx.x*72) + 4612)]))
+        conv2d_nchw_1[10] = (conv2d_nchw_1[10] + (pad_temp.shared_1[13]*kernel.shared_1[((threadIdx.x*72) + 4612)]))
+        conv2d_nchw_1[11] = (conv2d_nchw_1[11] + (pad_temp.shared_1[14]*kernel.shared_1[((threadIdx.x*72) + 4612)]))
+        conv2d_nchw_1[12] = (conv2d_nchw_1[12] + (pad_temp.shared_1[15]*kernel.shared_1[((threadIdx.x*72) + 4612)]))
+        conv2d_nchw_1[13] = (conv2d_nchw_1[13] + (pad_temp.shared_1[16]*kernel.shared_1[((threadIdx.x*72) + 4612)]))
+        conv2d_nchw_1[0] = (conv2d_nchw_1[0] + (pad_temp.shared_1[11]*kernel.shared_1[((threadIdx.x*72) + 5)]))
+        conv2d_nchw_1[1] = (conv2d_nchw_1[1] + (pad_temp.shared_1[12]*kernel.shared_1[((threadIdx.x*72) + 5)]))
+        conv2d_nchw_1[2] = (conv2d_nchw_1[2] + (pad_temp.shared_1[13]*kernel.shared_1[((threadIdx.x*72) + 5)]))
+        conv2d_nchw_1[3] = (conv2d_nchw_1[3] + (pad_temp.shared_1[14]*kernel.shared_1[((threadIdx.x*72) + 5)]))
+        conv2d_nchw_1[4] = (conv2d_nchw_1[4] + (pad_temp.shared_1[15]*kernel.shared_1[((threadIdx.x*72) + 5)]))
+        conv2d_nchw_1[5] = (conv2d_nchw_1[5] + (pad_temp.shared_1[16]*kernel.shared_1[((threadIdx.x*72) + 5)]))
+        conv2d_nchw_1[6] = (conv2d_nchw_1[6] + (pad_temp.shared_1[17]*kernel.shared_1[((threadIdx.x*72) + 5)]))
+        conv2d_nchw_1[7] = (conv2d_nchw_1[7] + (pad_temp.shared_1[11]*kernel.shared_1[((threadIdx.x*72) + 4613)]))
+        conv2d_nchw_1[8] = (conv2d_nchw_1[8] + (pad_temp.shared_1[12]*kernel.shared_1[((threadIdx.x*72) + 4613)]))
+        conv2d_nchw_1[9] = (conv2d_nchw_1[9] + (pad_temp.shared_1[13]*kernel.shared_1[((threadIdx.x*72) + 4613)]))
+        conv2d_nchw_1[10] = (conv2d_nchw_1[10] + (pad_temp.shared_1[14]*kernel.shared_1[((threadIdx.x*72) + 4613)]))
+        conv2d_nchw_1[11] = (conv2d_nchw_1[11] + (pad_temp.shared_1[15]*kernel.shared_1[((threadIdx.x*72) + 4613)]))
+        conv2d_nchw_1[12] = (conv2d_nchw_1[12] + (pad_temp.shared_1[16]*kernel.shared_1[((threadIdx.x*72) + 4613)]))
+        conv2d_nchw_1[13] = (conv2d_nchw_1[13] + (pad_temp.shared_1[17]*kernel.shared_1[((threadIdx.x*72) + 4613)]))
+        conv2d_nchw_1[0] = (conv2d_nchw_1[0] + (pad_temp.shared_1[18]*kernel.shared_1[((threadIdx.x*72) + 6)]))
+        conv2d_nchw_1[1] = (conv2d_nchw_1[1] + (pad_temp.shared_1[19]*kernel.shared_1[((threadIdx.x*72) + 6)]))
+        conv2d_nchw_1[2] = (conv2d_nchw_1[2] + (pad_temp.shared_1[20]*kernel.shared_1[((threadIdx.x*72) + 6)]))
+        conv2d_nchw_1[3] = (conv2d_nchw_1[3] + (pad_temp.shared_1[21]*kernel.shared_1[((threadIdx.x*72) + 6)]))
+        conv2d_nchw_1[4] = (conv2d_nchw_1[4] + (pad_temp.shared_1[22]*kernel.shared_1[((threadIdx.x*72) + 6)]))
+        conv2d_nchw_1[5] = (conv2d_nchw_1[5] + (pad_temp.shared_1[23]*kernel.shared_1[((threadIdx.x*72) + 6)]))
+        conv2d_nchw_1[6] = (conv2d_nchw_1[6] + (pad_temp.shared_1[24]*kernel.shared_1[((threadIdx.x*72) + 6)]))
+        conv2d_nchw_1[7] = (conv2d_nchw_1[7] + (pad_temp.shared_1[18]*kernel.shared_1[((threadIdx.x*72) + 4614)]))
+        conv2d_nchw_1[8] = (conv2d_nchw_1[8] + (pad_temp.shared_1[19]*kernel.shared_1[((threadIdx.x*72) + 4614)]))
+        conv2d_nchw_1[9] = (conv2d_nchw_1[9] + (pad_temp.shared_1[20]*kernel.shared_1[((threadIdx.x*72) + 4614)]))
+        conv2d_nchw_1[10] = (conv2d_nchw_1[10] + (pad_temp.shared_1[21]*kernel.shared_1[((threadIdx.x*72) + 4614)]))
+        conv2d_nchw_1[11] = (conv2d_nchw_1[11] + (pad_temp.shared_1[22]*kernel.shared_1[((threadIdx.x*72) + 4614)]))
+        conv2d_nchw_1[12] = (conv2d_nchw_1[12] + (pad_temp.shared_1[23]*kernel.shared_1[((threadIdx.x*72) + 4614)]))
+        conv2d_nchw_1[13] = (conv2d_nchw_1[13] + (pad_temp.shared_1[24]*kernel.shared_1[((threadIdx.x*72) + 4614)]))
+        conv2d_nchw_1[0] = (conv2d_nchw_1[0] + (pad_temp.shared_1[19]*kernel.shared_1[((threadIdx.x*72) + 7)]))
+        conv2d_nchw_1[1] = (conv2d_nchw_1[1] + (pad_temp.shared_1[20]*kernel.shared_1[((threadIdx.x*72) + 7)]))
+        conv2d_nchw_1[2] = (conv2d_nchw_1[2] + (pad_temp.shared_1[21]*kernel.shared_1[((threadIdx.x*72) + 7)]))
+        conv2d_nchw_1[3] = (conv2d_nchw_1[3] + (pad_temp.shared_1[22]*kernel.shared_1[((threadIdx.x*72) + 7)]))
+        conv2d_nchw_1[4] = (conv2d_nchw_1[4] + (pad_temp.shared_1[23]*kernel.shared_1[((threadIdx.x*72) + 7)]))
+        conv2d_nchw_1[5] = (conv2d_nchw_1[5] + (pad_temp.shared_1[24]*kernel.shared_1[((threadIdx.x*72) + 7)]))
+        conv2d_nchw_1[6] = (conv2d_nchw_1[6] + (pad_temp.shared_1[25]*kernel.shared_1[((threadIdx.x*72) + 7)]))
+        conv2d_nchw_1[7] = (conv2d_nchw_1[7] + (pad_temp.shared_1[19]*kernel.shared_1[((threadIdx.x*72) + 4615)]))
+        conv2d_nchw_1[8] = (conv2d_nchw_1[8] + (pad_temp.shared_1[20]*kernel.shared_1[((threadIdx.x*72) + 4615)]))
+        conv2d_nchw_1[9] = (conv2d_nchw_1[9] + (pad_temp.shared_1[21]*kernel.shared_1[((threadIdx.x*72) + 4615)]))
+        conv2d_nchw_1[10] = (conv2d_nchw_1[10] + (pad_temp.shared_1[22]*kernel.shared_1[((threadIdx.x*72) + 4615)]))
+        conv2d_nchw_1[11] = (conv2d_nchw_1[11] + (pad_temp.shared_1[23]*kernel.shared_1[((threadIdx.x*72) + 4615)]))
+        conv2d_nchw_1[12] = (conv2d_nchw_1[12] + (pad_temp.shared_1[24]*kernel.shared_1[((threadIdx.x*72) + 4615)]))
+        conv2d_nchw_1[13] = (conv2d_nchw_1[13] + (pad_temp.shared_1[25]*kernel.shared_1[((threadIdx.x*72) + 4615)]))
+        conv2d_nchw_1[0] = (conv2d_nchw_1[0] + (pad_temp.shared_1[20]*kernel.shared_1[((threadIdx.x*72) + 8)]))
+        conv2d_nchw_1[1] = (conv2d_nchw_1[1] + (pad_temp.shared_1[21]*kernel.shared_1[((threadIdx.x*72) + 8)]))
+        conv2d_nchw_1[2] = (conv2d_nchw_1[2] + (pad_temp.shared_1[22]*kernel.shared_1[((threadIdx.x*72) + 8)]))
+        conv2d_nchw_1[3] = (conv2d_nchw_1[3] + (pad_temp.shared_1[23]*kernel.shared_1[((threadIdx.x*72) + 8)]))
+        conv2d_nchw_1[4] = (conv2d_nchw_1[4] + (pad_temp.shared_1[24]*kernel.shared_1[((threadIdx.x*72) + 8)]))
+        conv2d_nchw_1[5] = (conv2d_nchw_1[5] + (pad_temp.shared_1[25]*kernel.shared_1[((threadIdx.x*72) + 8)]))
+        conv2d_nchw_1[6] = (conv2d_nchw_1[6] + (pad_temp.shared_1[26]*kernel.shared_1[((threadIdx.x*72) + 8)]))
+        conv2d_nchw_1[7] = (conv2d_nchw_1[7] + (pad_temp.shared_1[20]*kernel.shared_1[((threadIdx.x*72) + 4616)]))
+        conv2d_nchw_1[8] = (conv2d_nchw_1[8] + (pad_temp.shared_1[21]*kernel.shared_1[((threadIdx.x*72) + 4616)]))
+        conv2d_nchw_1[9] = (conv2d_nchw_1[9] + (pad_temp.shared_1[22]*kernel.shared_1[((threadIdx.x*72) + 4616)]))
+        conv2d_nchw_1[10] = (conv2d_nchw_1[10] + (pad_temp.shared_1[23]*kernel.shared_1[((threadIdx.x*72) + 4616)]))
+        conv2d_nchw_1[11] = (conv2d_nchw_1[11] + (pad_temp.shared_1[24]*kernel.shared_1[((threadIdx.x*72) + 4616)]))
+        conv2d_nchw_1[12] = (conv2d_nchw_1[12] + (pad_temp.shared_1[25]*kernel.shared_1[((threadIdx.x*72) + 4616)]))
+        conv2d_nchw_1[13] = (conv2d_nchw_1[13] + (pad_temp.shared_1[26]*kernel.shared_1[((threadIdx.x*72) + 4616)]))
+        conv2d_nchw_1[0] = (conv2d_nchw_1[0] + (pad_temp.shared_1[27]*kernel.shared_1[((threadIdx.x*72) + 9)]))
+        conv2d_nchw_1[1] = (conv2d_nchw_1[1] + (pad_temp.shared_1[28]*kernel.shared_1[((threadIdx.x*72) + 9)]))
+        conv2d_nchw_1[2] = (conv2d_nchw_1[2] + (pad_temp.shared_1[29]*kernel.shared_1[((threadIdx.x*72) + 9)]))
+        conv2d_nchw_1[3] = (conv2d_nchw_1[3] + (pad_temp.shared_1[30]*kernel.shared_1[((threadIdx.x*72) + 9)]))
+        conv2d_nchw_1[4] = (conv2d_nchw_1[4] + (pad_temp.shared_1[31]*kernel.shared_1[((threadIdx.x*72) + 9)]))
+        conv2d_nchw_1[5] = (conv2d_nchw_1[5] + (pad_temp.shared_1[32]*kernel.shared_1[((threadIdx.x*72) + 9)]))
+        conv2d_nchw_1[6] = (conv2d_nchw_1[6] + (pad_temp.shared_1[33]*kernel.shared_1[((threadIdx.x*72) + 9)]))
+        conv2d_nchw_1[7] = (conv2d_nchw_1[7] + (pad_temp.shared_1[27]*kernel.shared_1[((threadIdx.x*72) + 4617)]))
+        conv2d_nchw_1[8] = (conv2d_nchw_1[8] + (pad_temp.shared_1[28]*kernel.shared_1[((threadIdx.x*72) + 4617)]))
+        conv2d_nchw_1[9] = (conv2d_nchw_1[9] + (pad_temp.shared_1[29]*kernel.shared_1[((threadIdx.x*72) + 4617)]))
+        conv2d_nchw_1[10] = (conv2d_nchw_1[10] + (pad_temp.shared_1[30]*kernel.shared_1[((threadIdx.x*72) + 4617)]))
+        conv2d_nchw_1[11] = (conv2d_nchw_1[11] + (pad_temp.shared_1[31]*kernel.shared_1[((threadIdx.x*72) + 4617)]))
+        conv2d_nchw_1[12] = (conv2d_nchw_1[12] + (pad_temp.shared_1[32]*kernel.shared_1[((threadIdx.x*72) + 4617)]))
+        conv2d_nchw_1[13] = (conv2d_nchw_1[13] + (pad_temp.shared_1[33]*kernel.shared_1[((threadIdx.x*72) + 4617)]))
+        conv2d_nchw_1[0] = (conv2d_nchw_1[0] + (pad_temp.shared_1[28]*kernel.shared_1[((threadIdx.x*72) + 10)]))
+        conv2d_nchw_1[1] = (conv2d_nchw_1[1] + (pad_temp.shared_1[29]*kernel.shared_1[((threadIdx.x*72) + 10)]))
+        conv2d_nchw_1[2] = (conv2d_nchw_1[2] + (pad_temp.shared_1[30]*kernel.shared_1[((threadIdx.x*72) + 10)]))
+        conv2d_nchw_1[3] = (conv2d_nchw_1[3] + (pad_temp.shared_1[31]*kernel.shared_1[((threadIdx.x*72) + 10)]))
+        conv2d_nchw_1[4] = (conv2d_nchw_1[4] + (pad_temp.shared_1[32]*kernel.shared_1[((threadIdx.x*72) + 10)]))
+        conv2d_nchw_1[5] = (conv2d_nchw_1[5] + (pad_temp.shared_1[33]*kernel.shared_1[((threadIdx.x*72) + 10)]))
+        conv2d_nchw_1[6] = (conv2d_nchw_1[6] + (pad_temp.shared_1[34]*kernel.shared_1[((threadIdx.x*72) + 10)]))
+        conv2d_nchw_1[7] = (conv2d_nchw_1[7] + (pad_temp.shared_1[28]*kernel.shared_1[((threadIdx.x*72) + 4618)]))
+        conv2d_nchw_1[8] = (conv2d_nchw_1[8] + (pad_temp.shared_1[29]*kernel.shared_1[((threadIdx.x*72) + 4618)]))
+        conv2d_nchw_1[9] = (conv2d_nchw_1[9] + (pad_temp.shared_1[30]*kernel.shared_1[((threadIdx.x*72) + 4618)]))
+        conv2d_nchw_1[10] = (conv2d_nchw_1[10] + (pad_temp.shared_1[31]*kernel.shared_1[((threadIdx.x*72) + 4618)]))
+        conv2d_nchw_1[11] = (conv2d_nchw_1[11] + (pad_temp.shared_1[32]*kernel.shared_1[((threadIdx.x*72) + 4618)]))
+        conv2d_nchw_1[12] = (conv2d_nchw_1[12] + (pad_temp.shared_1[33]*kernel.shared_1[((threadIdx.x*72) + 4618)]))
+        conv2d_nchw_1[13] = (conv2d_nchw_1[13] + (pad_temp.shared_1[34]*kernel.shared_1[((threadIdx.x*72) + 4618)]))
+        conv2d_nchw_1[0] = (conv2d_nchw_1[0] + (pad_temp.shared_1[29]*kernel.shared_1[((threadIdx.x*72) + 11)]))
+        conv2d_nchw_1[1] = (conv2d_nchw_1[1] + (pad_temp.shared_1[30]*kernel.shared_1[((threadIdx.x*72) + 11)]))
+        conv2d_nchw_1[2] = (conv2d_nchw_1[2] + (pad_temp.shared_1[31]*kernel.shared_1[((threadIdx.x*72) + 11)]))
+        conv2d_nchw_1[3] = (conv2d_nchw_1[3] + (pad_temp.shared_1[32]*kernel.shared_1[((threadIdx.x*72) + 11)]))
+        conv2d_nchw_1[4] = (conv2d_nchw_1[4] + (pad_temp.shared_1[33]*kernel.shared_1[((threadIdx.x*72) + 11)]))
+        conv2d_nchw_1[5] = (conv2d_nchw_1[5] + (pad_temp.shared_1[34]*kernel.shared_1[((threadIdx.x*72) + 11)]))
+        conv2d_nchw_1[6] = (conv2d_nchw_1[6] + (pad_temp.shared_1[35]*kernel.shared_1[((threadIdx.x*72) + 11)]))
+        conv2d_nchw_1[7] = (conv2d_nchw_1[7] + (pad_temp.shared_1[29]*kernel.shared_1[((threadIdx.x*72) + 4619)]))
+        conv2d_nchw_1[8] = (conv2d_nchw_1[8] + (pad_temp.shared_1[30]*kernel.shared_1[((threadIdx.x*72) + 4619)]))
+        conv2d_nchw_1[9] = (conv2d_nchw_1[9] + (pad_temp.shared_1[31]*kernel.shared_1[((threadIdx.x*72) + 4619)]))
+        conv2d_nchw_1[10] = (conv2d_nchw_1[10] + (pad_temp.shared_1[32]*kernel.shared_1[((threadIdx.x*72) + 4619)]))
+        conv2d_nchw_1[11] = (conv2d_nchw_1[11] + (pad_temp.shared_1[33]*kernel.shared_1[((threadIdx.x*72) + 4619)]))
+        conv2d_nchw_1[12] = (conv2d_nchw_1[12] + (pad_temp.shared_1[34]*kernel.shared_1[((threadIdx.x*72) + 4619)]))
+        conv2d_nchw_1[13] = (conv2d_nchw_1[13] + (pad_temp.shared_1[35]*kernel.shared_1[((threadIdx.x*72) + 4619)]))
+        conv2d_nchw_1[0] = (conv2d_nchw_1[0] + (pad_temp.shared_1[36]*kernel.shared_1[((threadIdx.x*72) + 12)]))
+        conv2d_nchw_1[1] = (conv2d_nchw_1[1] + (pad_temp.shared_1[37]*kernel.shared_1[((threadIdx.x*72) + 12)]))
+        conv2d_nchw_1[2] = (conv2d_nchw_1[2] + (pad_temp.shared_1[38]*kernel.shared_1[((threadIdx.x*72) + 12)]))
+        conv2d_nchw_1[3] = (conv2d_nchw_1[3] + (pad_temp.shared_1[39]*kernel.shared_1[((threadIdx.x*72) + 12)]))
+        conv2d_nchw_1[4] = (conv2d_nchw_1[4] + (pad_temp.shared_1[40]*kernel.shared_1[((threadIdx.x*72) + 12)]))
+        conv2d_nchw_1[5] = (conv2d_nchw_1[5] + (pad_temp.shared_1[41]*kernel.shared_1[((threadIdx.x*72) + 12)]))
+        conv2d_nchw_1[6] = (conv2d_nchw_1[6] + (pad_temp.shared_1[42]*kernel.shared_1[((threadIdx.x*72) + 12)]))
+        conv2d_nchw_1[7] = (conv2d_nchw_1[7] + (pad_temp.shared_1[36]*kernel.shared_1[((threadIdx.x*72) + 4620)]))
+        conv2d_nchw_1[8] = (conv2d_nchw_1[8] + (pad_temp.shared_1[37]*kernel.shared_1[((threadIdx.x*72) + 4620)]))
+        conv2d_nchw_1[9] = (conv2d_nchw_1[9] + (pad_temp.shared_1[38]*kernel.shared_1[((threadIdx.x*72) + 4620)]))
+        conv2d_nchw_1[10] = (conv2d_nchw_1[10] + (pad_temp.shared_1[39]*kernel.shared_1[((threadIdx.x*72) + 4620)]))
+        conv2d_nchw_1[11] = (conv2d_nchw_1[11] + (pad_temp.shared_1[40]*kernel.shared_1[((threadIdx.x*72) + 4620)]))
+        conv2d_nchw_1[12] = (conv2d_nchw_1[12] + (pad_temp.shared_1[41]*kernel.shared_1[((threadIdx.x*72) + 4620)]))
+        conv2d_nchw_1[13] = (conv2d_nchw_1[13] + (pad_temp.shared_1[42]*kernel.shared_1[((threadIdx.x*72) + 4620)]))
+        conv2d_nchw_1[0] = (conv2d_nchw_1[0] + (pad_temp.shared_1[37]*kernel.shared_1[((threadIdx.x*72) + 13)]))
+        conv2d_nchw_1[1] = (conv2d_nchw_1[1] + (pad_temp.shared_1[38]*kernel.shared_1[((threadIdx.x*72) + 13)]))
+        conv2d_nchw_1[2] = (conv2d_nchw_1[2] + (pad_temp.shared_1[39]*kernel.shared_1[((threadIdx.x*72) + 13)]))
+        conv2d_nchw_1[3] = (conv2d_nchw_1[3] + (pad_temp.shared_1[40]*kernel.shared_1[((threadIdx.x*72) + 13)]))
+        conv2d_nchw_1[4] = (conv2d_nchw_1[4] + (pad_temp.shared_1[41]*kernel.shared_1[((threadIdx.x*72) + 13)]))
+        conv2d_nchw_1[5] = (conv2d_nchw_1[5] + (pad_temp.shared_1[42]*kernel.shared_1[((threadIdx.x*72) + 13)]))
+        conv2d_nchw_1[6] = (conv2d_nchw_1[6] + (pad_temp.shared_1[43]*kernel.shared_1[((threadIdx.x*72) + 13)]))
+        conv2d_nchw_1[7] = (conv2d_nchw_1[7] + (pad_temp.shared_1[37]*kernel.shared_1[((threadIdx.x*72) + 4621)]))
+        conv2d_nchw_1[8] = (conv2d_nchw_1[8] + (pad_temp.shared_1[38]*kernel.shared_1[((threadIdx.x*72) + 4621)]))
+        conv2d_nchw_1[9] = (conv2d_nchw_1[9] + (pad_temp.shared_1[39]*kernel.shared_1[((threadIdx.x*72) + 4621)]))
+        conv2d_nchw_1[10] = (conv2d_nchw_1[10] + (pad_temp.shared_1[40]*kernel.shared_1[((threadIdx.x*72) + 4621)]))
+        conv2d_nchw_1[11] = (conv2d_nchw_1[11] + (pad_temp.shared_1[41]*kernel.shared_1[((threadIdx.x*72) + 4621)]))
+        conv2d_nchw_1[12] = (conv2d_nchw_1[12] + (pad_temp.shared_1[42]*kernel.shared_1[((threadIdx.x*72) + 4621)]))
+        conv2d_nchw_1[13] = (conv2d_nchw_1[13] + (pad_temp.shared_1[43]*kernel.shared_1[((threadIdx.x*72) + 4621)]))
+        conv2d_nchw_1[0] = (conv2d_nchw_1[0] + (pad_temp.shared_1[38]*kernel.shared_1[((threadIdx.x*72) + 14)]))
+        conv2d_nchw_1[1] = (conv2d_nchw_1[1] + (pad_temp.shared_1[39]*kernel.shared_1[((threadIdx.x*72) + 14)]))
+        conv2d_nchw_1[2] = (conv2d_nchw_1[2] + (pad_temp.shared_1[40]*kernel.shared_1[((threadIdx.x*72) + 14)]))
+        conv2d_nchw_1[3] = (conv2d_nchw_1[3] + (pad_temp.shared_1[41]*kernel.shared_1[((threadIdx.x*72) + 14)]))
+        conv2d_nchw_1[4] = (conv2d_nchw_1[4] + (pad_temp.shared_1[42]*kernel.shared_1[((threadIdx.x*72) + 14)]))
+        conv2d_nchw_1[5] = (conv2d_nchw_1[5] + (pad_temp.shared_1[43]*kernel.shared_1[((threadIdx.x*72) + 14)]))
+        conv2d_nchw_1[6] = (conv2d_nchw_1[6] + (pad_temp.shared_1[44]*kernel.shared_1[((threadIdx.x*72) + 14)]))
+        conv2d_nchw_1[7] = (conv2d_nchw_1[7] + (pad_temp.shared_1[38]*kernel.shared_1[((threadIdx.x*72) + 4622)]))
+        conv2d_nchw_1[8] = (conv2d_nchw_1[8] + (pad_temp.shared_1[39]*kernel.shared_1[((threadIdx.x*72) + 4622)]))
+        conv2d_nchw_1[9] = (conv2d_nchw_1[9] + (pad_temp.shared_1[40]*kernel.shared_1[((threadIdx.x*72) + 4622)]))
+        conv2d_nchw_1[10] = (conv2d_nchw_1[10] + (pad_temp.shared_1[41]*kernel.shared_1[((threadIdx.x*72) + 4622)]))
+        conv2d_nchw_1[11] = (conv2d_nchw_1[11] + (pad_temp.shared_1[42]*kernel.shared_1[((threadIdx.x*72) + 4622)]))
+        conv2d_nchw_1[12] = (conv2d_nchw_1[12] + (pad_temp.shared_1[43]*kernel.shared_1[((threadIdx.x*72) + 4622)]))
+        conv2d_nchw_1[13] = (conv2d_nchw_1[13] + (pad_temp.shared_1[44]*kernel.shared_1[((threadIdx.x*72) + 4622)]))
+        conv2d_nchw_1[0] = (conv2d_nchw_1[0] + (pad_temp.shared_1[45]*kernel.shared_1[((threadIdx.x*72) + 15)]))
+        conv2d_nchw_1[1] = (conv2d_nchw_1[1] + (pad_temp.shared_1[46]*kernel.shared_1[((threadIdx.x*72) + 15)]))
+        conv2d_nchw_1[2] = (conv2d_nchw_1[2] + (pad_temp.shared_1[47]*kernel.shared_1[((threadIdx.x*72) + 15)]))
+        conv2d_nchw_1[3] = (conv2d_nchw_1[3] + (pad_temp.shared_1[48]*kernel.shared_1[((threadIdx.x*72) + 15)]))
+        conv2d_nchw_1[4] = (conv2d_nchw_1[4] + (pad_temp.shared_1[49]*kernel.shared_1[((threadIdx.x*72) + 15)]))
+        conv2d_nchw_1[5] = (conv2d_nchw_1[5] + (pad_temp.shared_1[50]*kernel.shared_1[((threadIdx.x*72) + 15)]))
+        conv2d_nchw_1[6] = (conv2d_nchw_1[6] + (pad_temp.shared_1[51]*kernel.shared_1[((threadIdx.x*72) + 15)]))
+        conv2d_nchw_1[7] = (conv2d_nchw_1[7] + (pad_temp.shared_1[45]*kernel.shared_1[((threadIdx.x*72) + 4623)]))
+        conv2d_nchw_1[8] = (conv2d_nchw_1[8] + (pad_temp.shared_1[46]*kernel.shared_1[((threadIdx.x*72) + 4623)]))
+        conv2d_nchw_1[9] = (conv2d_nchw_1[9] + (pad_temp.shared_1[47]*kernel.shared_1[((threadIdx.x*72) + 4623)]))
+        conv2d_nchw_1[10] = (conv2d_nchw_1[10] + (pad_temp.shared_1[48]*kernel.shared_1[((threadIdx.x*72) + 4623)]))
+        conv2d_nchw_1[11] = (conv2d_nchw_1[11] + (pad_temp.shared_1[49]*kernel.shared_1[((threadIdx.x*72) + 4623)]))
+        conv2d_nchw_1[12] = (conv2d_nchw_1[12] + (pad_temp.shared_1[50]*kernel.shared_1[((threadIdx.x*72) + 4623)]))
+        conv2d_nchw_1[13] = (conv2d_nchw_1[13] + (pad_temp.shared_1[51]*kernel.shared_1[((threadIdx.x*72) + 4623)]))
+        conv2d_nchw_1[0] = (conv2d_nchw_1[0] + (pad_temp.shared_1[46]*kernel.shared_1[((threadIdx.x*72) + 16)]))
+        conv2d_nchw_1[1] = (conv2d_nchw_1[1] + (pad_temp.shared_1[47]*kernel.shared_1[((threadIdx.x*72) + 16)]))
+        conv2d_nchw_1[2] = (conv2d_nchw_1[2] + (pad_temp.shared_1[48]*kernel.shared_1[((threadIdx.x*72) + 16)]))
+        conv2d_nchw_1[3] = (conv2d_nchw_1[3] + (pad_temp.shared_1[49]*kernel.shared_1[((threadIdx.x*72) + 16)]))
+        conv2d_nchw_1[4] = (conv2d_nchw_1[4] + (pad_temp.shared_1[50]*kernel.shared_1[((threadIdx.x*72) + 16)]))
+        conv2d_nchw_1[5] = (conv2d_nchw_1[5] + (pad_temp.shared_1[51]*kernel.shared_1[((threadIdx.x*72) + 16)]))
+        conv2d_nchw_1[6] = (conv2d_nchw_1[6] + (pad_temp.shared_1[52]*kernel.shared_1[((threadIdx.x*72) + 16)]))
+        conv2d_nchw_1[7] = (conv2d_nchw_1[7] + (pad_temp.shared_1[46]*kernel.shared_1[((threadIdx.x*72) + 4624)]))
+        conv2d_nchw_1[8] = (conv2d_nchw_1[8] + (pad_temp.shared_1[47]*kernel.shared_1[((threadIdx.x*72) + 4624)]))
+        conv2d_nchw_1[9] = (conv2d_nchw_1[9] + (pad_temp.shared_1[48]*kernel.shared_1[((threadIdx.x*72) + 4624)]))
+        conv2d_nchw_1[10] = (conv2d_nchw_1[10] + (pad_temp.shared_1[49]*kernel.shared_1[((threadIdx.x*72) + 4624)]))
+        conv2d_nchw_1[11] = (conv2d_nchw_1[11] + (pad_temp.shared_1[50]*kernel.shared_1[((threadIdx.x*72) + 4624)]))
+        conv2d_nchw_1[12] = (conv2d_nchw_1[12] + (pad_temp.shared_1[51]*kernel.shared_1[((threadIdx.x*72) + 4624)]))
+        conv2d_nchw_1[13] = (conv2d_nchw_1[13] + (pad_temp.shared_1[52]*kernel.shared_1[((threadIdx.x*72) + 4624)]))
+        conv2d_nchw_1[0] = (conv2d_nchw_1[0] + (pad_temp.shared_1[47]*kernel.shared_1[((threadIdx.x*72) + 17)]))
+        conv2d_nchw_1[1] = (conv2d_nchw_1[1] + (pad_temp.shared_1[48]*kernel.shared_1[((threadIdx.x*72) + 17)]))
+        conv2d_nchw_1[2] = (conv2d_nchw_1[2] + (pad_temp.shared_1[49]*kernel.shared_1[((threadIdx.x*72) + 17)]))
+        conv2d_nchw_1[3] = (conv2d_nchw_1[3] + (pad_temp.shared_1[50]*kernel.shared_1[((threadIdx.x*72) + 17)]))
+        conv2d_nchw_1[4] = (conv2d_nchw_1[4] + (pad_temp.shared_1[51]*kernel.shared_1[((threadIdx.x*72) + 17)]))
+        conv2d_nchw_1[5] = (conv2d_nchw_1[5] + (pad_temp.shared_1[52]*kernel.shared_1[((threadIdx.x*72) + 17)]))
+        conv2d_nchw_1[6] = (conv2d_nchw_1[6] + (pad_temp.shared_1[53]*kernel.shared_1[((threadIdx.x*72) + 17)]))
+        conv2d_nchw_1[7] = (conv2d_nchw_1[7] + (pad_temp.shared_1[47]*kernel.shared_1[((threadIdx.x*72) + 4625)]))
+        conv2d_nchw_1[8] = (conv2d_nchw_1[8] + (pad_temp.shared_1[48]*kernel.shared_1[((threadIdx.x*72) + 4625)]))
+        conv2d_nchw_1[9] = (conv2d_nchw_1[9] + (pad_temp.shared_1[49]*kernel.shared_1[((threadIdx.x*72) + 4625)]))
+        conv2d_nchw_1[10] = (conv2d_nchw_1[10] + (pad_temp.shared_1[50]*kernel.shared_1[((threadIdx.x*72) + 4625)]))
+        conv2d_nchw_1[11] = (conv2d_nchw_1[11] + (pad_temp.shared_1[51]*kernel.shared_1[((threadIdx.x*72) + 4625)]))
+        conv2d_nchw_1[12] = (conv2d_nchw_1[12] + (pad_temp.shared_1[52]*kernel.shared_1[((threadIdx.x*72) + 4625)]))
+        conv2d_nchw_1[13] = (conv2d_nchw_1[13] + (pad_temp.shared_1[53]*kernel.shared_1[((threadIdx.x*72) + 4625)]))
+        conv2d_nchw_1[0] = (conv2d_nchw_1[0] + (pad_temp.shared_1[54]*kernel.shared_1[((threadIdx.x*72) + 18)]))
+        conv2d_nchw_1[1] = (conv2d_nchw_1[1] + (pad_temp.shared_1[55]*kernel.shared_1[((threadIdx.x*72) + 18)]))
+        conv2d_nchw_1[2] = (conv2d_nchw_1[2] + (pad_temp.shared_1[56]*kernel.shared_1[((threadIdx.x*72) + 18)]))
+        conv2d_nchw_1[3] = (conv2d_nchw_1[3] + (pad_temp.shared_1[57]*kernel.shared_1[((threadIdx.x*72) + 18)]))
+        conv2d_nchw_1[4] = (conv2d_nchw_1[4] + (pad_temp.shared_1[58]*kernel.shared_1[((threadIdx.x*72) + 18)]))
+        conv2d_nchw_1[5] = (conv2d_nchw_1[5] + (pad_temp.shared_1[59]*kernel.shared_1[((threadIdx.x*72) + 18)]))
+        conv2d_nchw_1[6] = (conv2d_nchw_1[6] + (pad_temp.shared_1[60]*kernel.shared_1[((threadIdx.x*72) + 18)]))
+        conv2d_nchw_1[7] = (conv2d_nchw_1[7] + (pad_temp.shared_1[54]*kernel.shared_1[((threadIdx.x*72) + 4626)]))
+        conv2d_nchw_1[8] = (conv2d_nchw_1[8] + (pad_temp.shared_1[55]*kernel.shared_1[((threadIdx.x*72) + 4626)]))
+        conv2d_nchw_1[9] = (conv2d_nchw_1[9] + (pad_temp.shared_1[56]*kernel.shared_1[((threadIdx.x*72) + 4626)]))
+        conv2d_nchw_1[10] = (conv2d_nchw_1[10] + (pad_temp.shared_1[57]*kernel.shared_1[((threadIdx.x*72) + 4626)]))
+        conv2d_nchw_1[11] = (conv2d_nchw_1[11] + (pad_temp.shared_1[58]*kernel.shared_1[((threadIdx.x*72) + 4626)]))
+        conv2d_nchw_1[12] = (conv2d_nchw_1[12] + (pad_temp.shared_1[59]*kernel.shared_1[((threadIdx.x*72) + 4626)]))
+        conv2d_nchw_1[13] = (conv2d_nchw_1[13] + (pad_temp.shared_1[60]*kernel.shared_1[((threadIdx.x*72) + 4626)]))
+        conv2d_nchw_1[0] = (conv2d_nchw_1[0] + (pad_temp.shared_1[55]*kernel.shared_1[((threadIdx.x*72) + 19)]))
+        conv2d_nchw_1[1] = (conv2d_nchw_1[1] + (pad_temp.shared_1[56]*kernel.shared_1[((threadIdx.x*72) + 19)]))
+        conv2d_nchw_1[2] = (conv2d_nchw_1[2] + (pad_temp.shared_1[57]*kernel.shared_1[((threadIdx.x*72) + 19)]))
+        conv2d_nchw_1[3] = (conv2d_nchw_1[3] + (pad_temp.shared_1[58]*kernel.shared_1[((threadIdx.x*72) + 19)]))
+        conv2d_nchw_1[4] = (conv2d_nchw_1[4] + (pad_temp.shared_1[59]*kernel.shared_1[((threadIdx.x*72) + 19)]))
+        conv2d_nchw_1[5] = (conv2d_nchw_1[5] + (pad_temp.shared_1[60]*kernel.shared_1[((threadIdx.x*72) + 19)]))
+        conv2d_nchw_1[6] = (conv2d_nchw_1[6] + (pad_temp.shared_1[61]*kernel.shared_1[((threadIdx.x*72) + 19)]))
+        conv2d_nchw_1[7] = (conv2d_nchw_1[7] + (pad_temp.shared_1[55]*kernel.shared_1[((threadIdx.x*72) + 4627)]))
+        conv2d_nchw_1[8] = (conv2d_nchw_1[8] + (pad_temp.shared_1[56]*kernel.shared_1[((threadIdx.x*72) + 4627)]))
+        conv2d_nchw_1[9] = (conv2d_nchw_1[9] + (pad_temp.shared_1[57]*kernel.shared_1[((threadIdx.x*72) + 4627)]))
+        conv2d_nchw_1[10] = (conv2d_nchw_1[10] + (pad_temp.shared_1[58]*kernel.shared_1[((threadIdx.x*72) + 4627)]))
+        conv2d_nchw_1[11] = (conv2d_nchw_1[11] + (pad_temp.shared_1[59]*kernel.shared_1[((threadIdx.x*72) + 4627)]))
+        conv2d_nchw_1[12] = (conv2d_nchw_1[12] + (pad_temp.shared_1[60]*kernel.shared_1[((threadIdx.x*72) + 4627)]))
+        conv2d_nchw_1[13] = (conv2d_nchw_1[13] + (pad_temp.shared_1[61]*kernel.shared_1[((threadIdx.x*72) + 4627)]))
+        conv2d_nchw_1[0] = (conv2d_nchw_1[0] + (pad_temp.shared_1[56]*kernel.shared_1[((threadIdx.x*72) + 20)]))
+        conv2d_nchw_1[1] = (conv2d_nchw_1[1] + (pad_temp.shared_1[57]*kernel.shared_1[((threadIdx.x*72) + 20)]))
+        conv2d_nchw_1[2] = (conv2d_nchw_1[2] + (pad_temp.shared_1[58]*kernel.shared_1[((threadIdx.x*72) + 20)]))
+        conv2d_nchw_1[3] = (conv2d_nchw_1[3] + (pad_temp.shared_1[59]*kernel.shared_1[((threadIdx.x*72) + 20)]))
+        conv2d_nchw_1[4] = (conv2d_nchw_1[4] + (pad_temp.shared_1[60]*kernel.shared_1[((threadIdx.x*72) + 20)]))
+        conv2d_nchw_1[5] = (conv2d_nchw_1[5] + (pad_temp.shared_1[61]*kernel.shared_1[((threadIdx.x*72) + 20)]))
+        conv2d_nchw_1[6] = (conv2d_nchw_1[6] + (pad_temp.shared_1[62]*kernel.shared_1[((threadIdx.x*72) + 20)]))
+        conv2d_nchw_1[7] = (conv2d_nchw_1[7] + (pad_temp.shared_1[56]*kernel.shared_1[((threadIdx.x*72) + 4628)]))
+        conv2d_nchw_1[8] = (conv2d_nchw_1[8] + (pad_temp.shared_1[57]*kernel.shared_1[((threadIdx.x*72) + 4628)]))
+        conv2d_nchw_1[9] = (conv2d_nchw_1[9] + (pad_temp.shared_1[58]*kernel.shared_1[((threadIdx.x*72) + 4628)]))
+        conv2d_nchw_1[10] = (conv2d_nchw_1[10] + (pad_temp.shared_1[59]*kernel.shared_1[((threadIdx.x*72) + 4628)]))
+        conv2d_nchw_1[11] = (conv2d_nchw_1[11] + (pad_temp.shared_1[60]*kernel.shared_1[((threadIdx.x*72) + 4628)]))
+        conv2d_nchw_1[12] = (conv2d_nchw_1[12] + (pad_temp.shared_1[61]*kernel.shared_1[((threadIdx.x*72) + 4628)]))
+        conv2d_nchw_1[13] = (conv2d_nchw_1[13] + (pad_temp.shared_1[62]*kernel.shared_1[((threadIdx.x*72) + 4628)]))
+        conv2d_nchw_1[0] = (conv2d_nchw_1[0] + (pad_temp.shared_1[63]*kernel.shared_1[((threadIdx.x*72) + 21)]))
+        conv2d_nchw_1[1] = (conv2d_nchw_1[1] + (pad_temp.shared_1[64]*kernel.shared_1[((threadIdx.x*72) + 21)]))
+        conv2d_nchw_1[2] = (conv2d_nchw_1[2] + (pad_temp.shared_1[65]*kernel.shared_1[((threadIdx.x*72) + 21)]))
+        conv2d_nchw_1[3] = (conv2d_nchw_1[3] + (pad_temp.shared_1[66]*kernel.shared_1[((threadIdx.x*72) + 21)]))
+        conv2d_nchw_1[4] = (conv2d_nchw_1[4] + (pad_temp.shared_1[67]*kernel.shared_1[((threadIdx.x*72) + 21)]))
+        conv2d_nchw_1[5] = (conv2d_nchw_1[5] + (pad_temp.shared_1[68]*kernel.shared_1[((threadIdx.x*72) + 21)]))
+        conv2d_nchw_1[6] = (conv2d_nchw_1[6] + (pad_temp.shared_1[69]*kernel.shared_1[((threadIdx.x*72) + 21)]))
+        conv2d_nchw_1[7] = (conv2d_nchw_1[7] + (pad_temp.shared_1[63]*kernel.shared_1[((threadIdx.x*72) + 4629)]))
+        conv2d_nchw_1[8] = (conv2d_nchw_1[8] + (pad_temp.shared_1[64]*kernel.shared_1[((threadIdx.x*72) + 4629)]))
+        conv2d_nchw_1[9] = (conv2d_nchw_1[9] + (pad_temp.shared_1[65]*kernel.shared_1[((threadIdx.x*72) + 4629)]))
+        conv2d_nchw_1[10] = (conv2d_nchw_1[10] + (pad_temp.shared_1[66]*kernel.shared_1[((threadIdx.x*72) + 4629)]))
+        conv2d_nchw_1[11] = (conv2d_nchw_1[11] + (pad_temp.shared_1[67]*kernel.shared_1[((threadIdx.x*72) + 4629)]))
+        conv2d_nchw_1[12] = (conv2d_nchw_1[12] + (pad_temp.shared_1[68]*kernel.shared_1[((threadIdx.x*72) + 4629)]))
+        conv2d_nchw_1[13] = (conv2d_nchw_1[13] + (pad_temp.shared_1[69]*kernel.shared_1[((threadIdx.x*72) + 4629)]))
+        conv2d_nchw_1[0] = (conv2d_nchw_1[0] + (pad_temp.shared_1[64]*kernel.shared_1[((threadIdx.x*72) + 22)]))
+        conv2d_nchw_1[1] = (conv2d_nchw_1[1] + (pad_temp.shared_1[65]*kernel.shared_1[((threadIdx.x*72) + 22)]))
+        conv2d_nchw_1[2] = (conv2d_nchw_1[2] + (pad_temp.shared_1[66]*kernel.shared_1[((threadIdx.x*72) + 22)]))
+        conv2d_nchw_1[3] = (conv2d_nchw_1[3] + (pad_temp.shared_1[67]*kernel.shared_1[((threadIdx.x*72) + 22)]))
+        conv2d_nchw_1[4] = (conv2d_nchw_1[4] + (pad_temp.shared_1[68]*kernel.shared_1[((threadIdx.x*72) + 22)]))
+        conv2d_nchw_1[5] = (conv2d_nchw_1[5] + (pad_temp.shared_1[69]*kernel.shared_1[((threadIdx.x*72) + 22)]))
+        conv2d_nchw_1[6] = (conv2d_nchw_1[6] + (pad_temp.shared_1[70]*kernel.shared_1[((threadIdx.x*72) + 22)]))
+        conv2d_nchw_1[7] = (conv2d_nchw_1[7] + (pad_temp.shared_1[64]*kernel.shared_1[((threadIdx.x*72) + 4630)]))
+        conv2d_nchw_1[8] = (conv2d_nchw_1[8] + (pad_temp.shared_1[65]*kernel.shared_1[((threadIdx.x*72) + 4630)]))
+        conv2d_nchw_1[9] = (conv2d_nchw_1[9] + (pad_temp.shared_1[66]*kernel.shared_1[((threadIdx.x*72) + 4630)]))
+        conv2d_nchw_1[10] = (conv2d_nchw_1[10] + (pad_temp.shared_1[67]*kernel.shared_1[((threadIdx.x*72) + 4630)]))
+        conv2d_nchw_1[11] = (conv2d_nchw_1[11] + (pad_temp.shared_1[68]*kernel.shared_1[((threadIdx.x*72) + 4630)]))
+        conv2d_nchw_1[12] = (conv2d_nchw_1[12] + (pad_temp.shared_1[69]*kernel.shared_1[((threadIdx.x*72) + 4630)]))
+        conv2d_nchw_1[13] = (conv2d_nchw_1[13] + (pad_temp.shared_1[70]*kernel.shared_1[((threadIdx.x*72) + 4630)]))
+        conv2d_nchw_1[0] = (conv2d_nchw_1[0] + (pad_temp.shared_1[65]*kernel.shared_1[((threadIdx.x*72) + 23)]))
+        conv2d_nchw_1[1] = (conv2d_nchw_1[1] + (pad_temp.shared_1[66]*kernel.shared_1[((threadIdx.x*72) + 23)]))
+        conv2d_nchw_1[2] = (conv2d_nchw_1[2] + (pad_temp.shared_1[67]*kernel.shared_1[((threadIdx.x*72) + 23)]))
+        conv2d_nchw_1[3] = (conv2d_nchw_1[3] + (pad_temp.shared_1[68]*kernel.shared_1[((threadIdx.x*72) + 23)]))
+        conv2d_nchw_1[4] = (conv2d_nchw_1[4] + (pad_temp.shared_1[69]*kernel.shared_1[((threadIdx.x*72) + 23)]))
+        conv2d_nchw_1[5] = (conv2d_nchw_1[5] + (pad_temp.shared_1[70]*kernel.shared_1[((threadIdx.x*72) + 23)]))
+        conv2d_nchw_1[6] = (conv2d_nchw_1[6] + (pad_temp.shared_1[71]*kernel.shared_1[((threadIdx.x*72) + 23)]))
+        conv2d_nchw_1[7] = (conv2d_nchw_1[7] + (pad_temp.shared_1[65]*kernel.shared_1[((threadIdx.x*72) + 4631)]))
+        conv2d_nchw_1[8] = (conv2d_nchw_1[8] + (pad_temp.shared_1[66]*kernel.shared_1[((threadIdx.x*72) + 4631)]))
+        conv2d_nchw_1[9] = (conv2d_nchw_1[9] + (pad_temp.shared_1[67]*kernel.shared_1[((threadIdx.x*72) + 4631)]))
+        conv2d_nchw_1[10] = (conv2d_nchw_1[10] + (pad_temp.shared_1[68]*kernel.shared_1[((threadIdx.x*72) + 4631)]))
+        conv2d_nchw_1[11] = (conv2d_nchw_1[11] + (pad_temp.shared_1[69]*kernel.shared_1[((threadIdx.x*72) + 4631)]))
+        conv2d_nchw_1[12] = (conv2d_nchw_1[12] + (pad_temp.shared_1[70]*kernel.shared_1[((threadIdx.x*72) + 4631)]))
+        conv2d_nchw_1[13] = (conv2d_nchw_1[13] + (pad_temp.shared_1[71]*kernel.shared_1[((threadIdx.x*72) + 4631)]))
+        conv2d_nchw_1[0] = (conv2d_nchw_1[0] + (pad_temp.shared_1[72]*kernel.shared_1[((threadIdx.x*72) + 24)]))
+        conv2d_nchw_1[1] = (conv2d_nchw_1[1] + (pad_temp.shared_1[73]*kernel.shared_1[((threadIdx.x*72) + 24)]))
+        conv2d_nchw_1[2] = (conv2d_nchw_1[2] + (pad_temp.shared_1[74]*kernel.shared_1[((threadIdx.x*72) + 24)]))
+        conv2d_nchw_1[3] = (conv2d_nchw_1[3] + (pad_temp.shared_1[75]*kernel.shared_1[((threadIdx.x*72) + 24)]))
+        conv2d_nchw_1[4] = (conv2d_nchw_1[4] + (pad_temp.shared_1[76]*kernel.shared_1[((threadIdx.x*72) + 24)]))
+        conv2d_nchw_1[5] = (conv2d_nchw_1[5] + (pad_temp.shared_1[77]*kernel.shared_1[((threadIdx.x*72) + 24)]))
+        conv2d_nchw_1[6] = (conv2d_nchw_1[6] + (pad_temp.shared_1[78]*kernel.shared_1[((threadIdx.x*72) + 24)]))
+        conv2d_nchw_1[7] = (conv2d_nchw_1[7] + (pad_temp.shared_1[72]*kernel.shared_1[((threadIdx.x*72) + 4632)]))
+        conv2d_nchw_1[8] = (conv2d_nchw_1[8] + (pad_temp.shared_1[73]*kernel.shared_1[((threadIdx.x*72) + 4632)]))
+        conv2d_nchw_1[9] = (conv2d_nchw_1[9] + (pad_temp.shared_1[74]*kernel.shared_1[((threadIdx.x*72) + 4632)]))
+        conv2d_nchw_1[10] = (conv2d_nchw_1[10] + (pad_temp.shared_1[75]*kernel.shared_1[((threadIdx.x*72) + 4632)]))
+        conv2d_nchw_1[11] = (conv2d_nchw_1[11] + (pad_temp.shared_1[76]*kernel.shared_1[((threadIdx.x*72) + 4632)]))
+        conv2d_nchw_1[12] = (conv2d_nchw_1[12] + (pad_temp.shared_1[77]*kernel.shared_1[((threadIdx.x*72) + 4632)]))
+        conv2d_nchw_1[13] = (conv2d_nchw_1[13] + (pad_temp.shared_1[78]*kernel.shared_1[((threadIdx.x*72) + 4632)]))
+        conv2d_nchw_1[0] = (conv2d_nchw_1[0] + (pad_temp.shared_1[73]*kernel.shared_1[((threadIdx.x*72) + 25)]))
+        conv2d_nchw_1[1] = (conv2d_nchw_1[1] + (pad_temp.shared_1[74]*kernel.shared_1[((threadIdx.x*72) + 25)]))
+        conv2d_nchw_1[2] = (conv2d_nchw_1[2] + (pad_temp.shared_1[75]*kernel.shared_1[((threadIdx.x*72) + 25)]))
+        conv2d_nchw_1[3] = (conv2d_nchw_1[3] + (pad_temp.shared_1[76]*kernel.shared_1[((threadIdx.x*72) + 25)]))
+        conv2d_nchw_1[4] = (conv2d_nchw_1[4] + (pad_temp.shared_1[77]*kernel.shared_1[((threadIdx.x*72) + 25)]))
+        conv2d_nchw_1[5] = (conv2d_nchw_1[5] + (pad_temp.shared_1[78]*kernel.shared_1[((threadIdx.x*72) + 25)]))
+        conv2d_nchw_1[6] = (conv2d_nchw_1[6] + (pad_temp.shared_1[79]*kernel.shared_1[((threadIdx.x*72) + 25)]))
+        conv2d_nchw_1[7] = (conv2d_nchw_1[7] + (pad_temp.shared_1[73]*kernel.shared_1[((threadIdx.x*72) + 4633)]))
+        conv2d_nchw_1[8] = (conv2d_nchw_1[8] + (pad_temp.shared_1[74]*kernel.shared_1[((threadIdx.x*72) + 4633)]))
+        conv2d_nchw_1[9] = (conv2d_nchw_1[9] + (pad_temp.shared_1[75]*kernel.shared_1[((threadIdx.x*72) + 4633)]))
+        conv2d_nchw_1[10] = (conv2d_nchw_1[10] + (pad_temp.shared_1[76]*kernel.shared_1[((threadIdx.x*72) + 4633)]))
+        conv2d_nchw_1[11] = (conv2d_nchw_1[11] + (pad_temp.shared_1[77]*kernel.shared_1[((threadIdx.x*72) + 4633)]))
+        conv2d_nchw_1[12] = (conv2d_nchw_1[12] + (pad_temp.shared_1[78]*kernel.shared_1[((threadIdx.x*72) + 4633)]))
+        conv2d_nchw_1[13] = (conv2d_nchw_1[13] + (pad_temp.shared_1[79]*kernel.shared_1[((threadIdx.x*72) + 4633)]))
+        conv2d_nchw_1[0] = (conv2d_nchw_1[0] + (pad_temp.shared_1[74]*kernel.shared_1[((threadIdx.x*72) + 26)]))
+        conv2d_nchw_1[1] = (conv2d_nchw_1[1] + (pad_temp.shared_1[75]*kernel.shared_1[((threadIdx.x*72) + 26)]))
+        conv2d_nchw_1[2] = (conv2d_nchw_1[2] + (pad_temp.shared_1[76]*kernel.shared_1[((threadIdx.x*72) + 26)]))
+        conv2d_nchw_1[3] = (conv2d_nchw_1[3] + (pad_temp.shared_1[77]*kernel.shared_1[((threadIdx.x*72) + 26)]))
+        conv2d_nchw_1[4] = (conv2d_nchw_1[4] + (pad_temp.shared_1[78]*kernel.shared_1[((threadIdx.x*72) + 26)]))
+        conv2d_nchw_1[5] = (conv2d_nchw_1[5] + (pad_temp.shared_1[79]*kernel.shared_1[((threadIdx.x*72) + 26)]))
+        conv2d_nchw_1[6] = (conv2d_nchw_1[6] + (pad_temp.shared_1[80]*kernel.shared_1[((threadIdx.x*72) + 26)]))
+        conv2d_nchw_1[7] = (conv2d_nchw_1[7] + (pad_temp.shared_1[74]*kernel.shared_1[((threadIdx.x*72) + 4634)]))
+        conv2d_nchw_1[8] = (conv2d_nchw_1[8] + (pad_temp.shared_1[75]*kernel.shared_1[((threadIdx.x*72) + 4634)]))
+        conv2d_nchw_1[9] = (conv2d_nchw_1[9] + (pad_temp.shared_1[76]*kernel.shared_1[((threadIdx.x*72) + 4634)]))
+        conv2d_nchw_1[10] = (conv2d_nchw_1[10] + (pad_temp.shared_1[77]*kernel.shared_1[((threadIdx.x*72) + 4634)]))
+        conv2d_nchw_1[11] = (conv2d_nchw_1[11] + (pad_temp.shared_1[78]*kernel.shared_1[((threadIdx.x*72) + 4634)]))
+        conv2d_nchw_1[12] = (conv2d_nchw_1[12] + (pad_temp.shared_1[79]*kernel.shared_1[((threadIdx.x*72) + 4634)]))
+        conv2d_nchw_1[13] = (conv2d_nchw_1[13] + (pad_temp.shared_1[80]*kernel.shared_1[((threadIdx.x*72) + 4634)]))
+        conv2d_nchw_1[0] = (conv2d_nchw_1[0] + (pad_temp.shared_1[81]*kernel.shared_1[((threadIdx.x*72) + 27)]))
+        conv2d_nchw_1[1] = (conv2d_nchw_1[1] + (pad_temp.shared_1[82]*kernel.shared_1[((threadIdx.x*72) + 27)]))
+        conv2d_nchw_1[2] = (conv2d_nchw_1[2] + (pad_temp.shared_1[83]*kernel.shared_1[((threadIdx.x*72) + 27)]))
+        conv2d_nchw_1[3] = (conv2d_nchw_1[3] + (pad_temp.shared_1[84]*kernel.shared_1[((threadIdx.x*72) + 27)]))
+        conv2d_nchw_1[4] = (conv2d_nchw_1[4] + (pad_temp.shared_1[85]*kernel.shared_1[((threadIdx.x*72) + 27)]))
+        conv2d_nchw_1[5] = (conv2d_nchw_1[5] + (pad_temp.shared_1[86]*kernel.shared_1[((threadIdx.x*72) + 27)]))
+        conv2d_nchw_1[6] = (conv2d_nchw_1[6] + (pad_temp.shared_1[87]*kernel.shared_1[((threadIdx.x*72) + 27)]))
+        conv2d_nchw_1[7] = (conv2d_nchw_1[7] + (pad_temp.shared_1[81]*kernel.shared_1[((threadIdx.x*72) + 4635)]))
+        conv2d_nchw_1[8] = (conv2d_nchw_1[8] + (pad_temp.shared_1[82]*kernel.shared_1[((threadIdx.x*72) + 4635)]))
+        conv2d_nchw_1[9] = (conv2d_nchw_1[9] + (pad_temp.shared_1[83]*kernel.shared_1[((threadIdx.x*72) + 4635)]))
+        conv2d_nchw_1[10] = (conv2d_nchw_1[10] + (pad_temp.shared_1[84]*kernel.shared_1[((threadIdx.x*72) + 4635)]))
+        conv2d_nchw_1[11] = (conv2d_nchw_1[11] + (pad_temp.shared_1[85]*kernel.shared_1[((threadIdx.x*72) + 4635)]))
+        conv2d_nchw_1[12] = (conv2d_nchw_1[12] + (pad_temp.shared_1[86]*kernel.shared_1[((threadIdx.x*72) + 4635)]))
+        conv2d_nchw_1[13] = (conv2d_nchw_1[13] + (pad_temp.shared_1[87]*kernel.shared_1[((threadIdx.x*72) + 4635)]))
+        conv2d_nchw_1[0] = (conv2d_nchw_1[0] + (pad_temp.shared_1[82]*kernel.shared_1[((threadIdx.x*72) + 28)]))
+        conv2d_nchw_1[1] = (conv2d_nchw_1[1] + (pad_temp.shared_1[83]*kernel.shared_1[((threadIdx.x*72) + 28)]))
+        conv2d_nchw_1[2] = (conv2d_nchw_1[2] + (pad_temp.shared_1[84]*kernel.shared_1[((threadIdx.x*72) + 28)]))
+        conv2d_nchw_1[3] = (conv2d_nchw_1[3] + (pad_temp.shared_1[85]*kernel.shared_1[((threadIdx.x*72) + 28)]))
+        conv2d_nchw_1[4] = (conv2d_nchw_1[4] + (pad_temp.shared_1[86]*kernel.shared_1[((threadIdx.x*72) + 28)]))
+        conv2d_nchw_1[5] = (conv2d_nchw_1[5] + (pad_temp.shared_1[87]*kernel.shared_1[((threadIdx.x*72) + 28)]))
+        conv2d_nchw_1[6] = (conv2d_nchw_1[6] + (pad_temp.shared_1[88]*kernel.shared_1[((threadIdx.x*72) + 28)]))
+        conv2d_nchw_1[7] = (conv2d_nchw_1[7] + (pad_temp.shared_1[82]*kernel.shared_1[((threadIdx.x*72) + 4636)]))
+        conv2d_nchw_1[8] = (conv2d_nchw_1[8] + (pad_temp.shared_1[83]*kernel.shared_1[((threadIdx.x*72) + 4636)]))
+        conv2d_nchw_1[9] = (conv2d_nchw_1[9] + (pad_temp.shared_1[84]*kernel.shared_1[((threadIdx.x*72) + 4636)]))
+        conv2d_nchw_1[10] = (conv2d_nchw_1[10] + (pad_temp.shared_1[85]*kernel.shared_1[((threadIdx.x*72) + 4636)]))
+        conv2d_nchw_1[11] = (conv2d_nchw_1[11] + (pad_temp.shared_1[86]*kernel.shared_1[((threadIdx.x*72) + 4636)]))
+        conv2d_nchw_1[12] = (conv2d_nchw_1[12] + (pad_temp.shared_1[87]*kernel.shared_1[((threadIdx.x*72) + 4636)]))
+        conv2d_nchw_1[13] = (conv2d_nchw_1[13] + (pad_temp.shared_1[88]*kernel.shared_1[((threadIdx.x*72) + 4636)]))
+        conv2d_nchw_1[0] = (conv2d_nchw_1[0] + (pad_temp.shared_1[83]*kernel.shared_1[((threadIdx.x*72) + 29)]))
+        conv2d_nchw_1[1] = (conv2d_nchw_1[1] + (pad_temp.shared_1[84]*kernel.shared_1[((threadIdx.x*72) + 29)]))
+        conv2d_nchw_1[2] = (conv2d_nchw_1[2] + (pad_temp.shared_1[85]*kernel.shared_1[((threadIdx.x*72) + 29)]))
+        conv2d_nchw_1[3] = (conv2d_nchw_1[3] + (pad_temp.shared_1[86]*kernel.shared_1[((threadIdx.x*72) + 29)]))
+        conv2d_nchw_1[4] = (conv2d_nchw_1[4] + (pad_temp.shared_1[87]*kernel.shared_1[((threadIdx.x*72) + 29)]))
+        conv2d_nchw_1[5] = (conv2d_nchw_1[5] + (pad_temp.shared_1[88]*kernel.shared_1[((threadIdx.x*72) + 29)]))
+        conv2d_nchw_1[6] = (conv2d_nchw_1[6] + (pad_temp.shared_1[89]*kernel.shared_1[((threadIdx.x*72) + 29)]))
+        conv2d_nchw_1[7] = (conv2d_nchw_1[7] + (pad_temp.shared_1[83]*kernel.shared_1[((threadIdx.x*72) + 4637)]))
+        conv2d_nchw_1[8] = (conv2d_nchw_1[8] + (pad_temp.shared_1[84]*kernel.shared_1[((threadIdx.x*72) + 4637)]))
+        conv2d_nchw_1[9] = (conv2d_nchw_1[9] + (pad_temp.shared_1[85]*kernel.shared_1[((threadIdx.x*72) + 4637)]))
+        conv2d_nchw_1[10] = (conv2d_nchw_1[10] + (pad_temp.shared_1[86]*kernel.shared_1[((threadIdx.x*72) + 4637)]))
+        conv2d_nchw_1[11] = (conv2d_nchw_1[11] + (pad_temp.shared_1[87]*kernel.shared_1[((threadIdx.x*72) + 4637)]))
+        conv2d_nchw_1[12] = (conv2d_nchw_1[12] + (pad_temp.shared_1[88]*kernel.shared_1[((threadIdx.x*72) + 4637)]))
+        conv2d_nchw_1[13] = (conv2d_nchw_1[13] + (pad_temp.shared_1[89]*kernel.shared_1[((threadIdx.x*72) + 4637)]))
+        conv2d_nchw_1[0] = (conv2d_nchw_1[0] + (pad_temp.shared_1[90]*kernel.shared_1[((threadIdx.x*72) + 30)]))
+        conv2d_nchw_1[1] = (conv2d_nchw_1[1] + (pad_temp.shared_1[91]*kernel.shared_1[((threadIdx.x*72) + 30)]))
+        conv2d_nchw_1[2] = (conv2d_nchw_1[2] + (pad_temp.shared_1[92]*kernel.shared_1[((threadIdx.x*72) + 30)]))
+        conv2d_nchw_1[3] = (conv2d_nchw_1[3] + (pad_temp.shared_1[93]*kernel.shared_1[((threadIdx.x*72) + 30)]))
+        conv2d_nchw_1[4] = (conv2d_nchw_1[4] + (pad_temp.shared_1[94]*kernel.shared_1[((threadIdx.x*72) + 30)]))
+        conv2d_nchw_1[5] = (conv2d_nchw_1[5] + (pad_temp.shared_1[95]*kernel.shared_1[((threadIdx.x*72) + 30)]))
+        conv2d_nchw_1[6] = (conv2d_nchw_1[6] + (pad_temp.shared_1[96]*kernel.shared_1[((threadIdx.x*72) + 30)]))
+        conv2d_nchw_1[7] = (conv2d_nchw_1[7] + (pad_temp.shared_1[90]*kernel.shared_1[((threadIdx.x*72) + 4638)]))
+        conv2d_nchw_1[8] = (conv2d_nchw_1[8] + (pad_temp.shared_1[91]*kernel.shared_1[((threadIdx.x*72) + 4638)]))
+        conv2d_nchw_1[9] = (conv2d_nchw_1[9] + (pad_temp.shared_1[92]*kernel.shared_1[((threadIdx.x*72) + 4638)]))
+        conv2d_nchw_1[10] = (conv2d_nchw_1[10] + (pad_temp.shared_1[93]*kernel.shared_1[((threadIdx.x*72) + 4638)]))
+        conv2d_nchw_1[11] = (conv2d_nchw_1[11] + (pad_temp.shared_1[94]*kernel.shared_1[((threadIdx.x*72) + 4638)]))
+        conv2d_nchw_1[12] = (conv2d_nchw_1[12] + (pad_temp.shared_1[95]*kernel.shared_1[((threadIdx.x*72) + 4638)]))
+        conv2d_nchw_1[13] = (conv2d_nchw_1[13] + (pad_temp.shared_1[96]*kernel.shared_1[((threadIdx.x*72) + 4638)]))
+        conv2d_nchw_1[0] = (conv2d_nchw_1[0] + (pad_temp.shared_1[91]*kernel.shared_1[((threadIdx.x*72) + 31)]))
+        conv2d_nchw_1[1] = (conv2d_nchw_1[1] + (pad_temp.shared_1[92]*kernel.shared_1[((threadIdx.x*72) + 31)]))
+        conv2d_nchw_1[2] = (conv2d_nchw_1[2] + (pad_temp.shared_1[93]*kernel.shared_1[((threadIdx.x*72) + 31)]))
+        conv2d_nchw_1[3] = (conv2d_nchw_1[3] + (pad_temp.shared_1[94]*kernel.shared_1[((threadIdx.x*72) + 31)]))
+        conv2d_nchw_1[4] = (conv2d_nchw_1[4] + (pad_temp.shared_1[95]*kernel.shared_1[((threadIdx.x*72) + 31)]))
+        conv2d_nchw_1[5] = (conv2d_nchw_1[5] + (pad_temp.shared_1[96]*kernel.shared_1[((threadIdx.x*72) + 31)]))
+        conv2d_nchw_1[6] = (conv2d_nchw_1[6] + (pad_temp.shared_1[97]*kernel.shared_1[((threadIdx.x*72) + 31)]))
+        conv2d_nchw_1[7] = (conv2d_nchw_1[7] + (pad_temp.shared_1[91]*kernel.shared_1[((threadIdx.x*72) + 4639)]))
+        conv2d_nchw_1[8] = (conv2d_nchw_1[8] + (pad_temp.shared_1[92]*kernel.shared_1[((threadIdx.x*72) + 4639)]))
+        conv2d_nchw_1[9] = (conv2d_nchw_1[9] + (pad_temp.shared_1[93]*kernel.shared_1[((threadIdx.x*72) + 4639)]))
+        conv2d_nchw_1[10] = (conv2d_nchw_1[10] + (pad_temp.shared_1[94]*kernel.shared_1[((threadIdx.x*72) + 4639)]))
+        conv2d_nchw_1[11] = (conv2d_nchw_1[11] + (pad_temp.shared_1[95]*kernel.shared_1[((threadIdx.x*72) + 4639)]))
+        conv2d_nchw_1[12] = (conv2d_nchw_1[12] + (pad_temp.shared_1[96]*kernel.shared_1[((threadIdx.x*72) + 4639)]))
+        conv2d_nchw_1[13] = (conv2d_nchw_1[13] + (pad_temp.shared_1[97]*kernel.shared_1[((threadIdx.x*72) + 4639)]))
+        conv2d_nchw_1[0] = (conv2d_nchw_1[0] + (pad_temp.shared_1[92]*kernel.shared_1[((threadIdx.x*72) + 32)]))
+        conv2d_nchw_1[1] = (conv2d_nchw_1[1] + (pad_temp.shared_1[93]*kernel.shared_1[((threadIdx.x*72) + 32)]))
+        conv2d_nchw_1[2] = (conv2d_nchw_1[2] + (pad_temp.shared_1[94]*kernel.shared_1[((threadIdx.x*72) + 32)]))
+        conv2d_nchw_1[3] = (conv2d_nchw_1[3] + (pad_temp.shared_1[95]*kernel.shared_1[((threadIdx.x*72) + 32)]))
+        conv2d_nchw_1[4] = (conv2d_nchw_1[4] + (pad_temp.shared_1[96]*kernel.shared_1[((threadIdx.x*72) + 32)]))
+        conv2d_nchw_1[5] = (conv2d_nchw_1[5] + (pad_temp.shared_1[97]*kernel.shared_1[((threadIdx.x*72) + 32)]))
+        conv2d_nchw_1[6] = (conv2d_nchw_1[6] + (pad_temp.shared_1[98]*kernel.shared_1[((threadIdx.x*72) + 32)]))
+        conv2d_nchw_1[7] = (conv2d_nchw_1[7] + (pad_temp.shared_1[92]*kernel.shared_1[((threadIdx.x*72) + 4640)]))
+        conv2d_nchw_1[8] = (conv2d_nchw_1[8] + (pad_temp.shared_1[93]*kernel.shared_1[((threadIdx.x*72) + 4640)]))
+        conv2d_nchw_1[9] = (conv2d_nchw_1[9] + (pad_temp.shared_1[94]*kernel.shared_1[((threadIdx.x*72) + 4640)]))
+        conv2d_nchw_1[10] = (conv2d_nchw_1[10] + (pad_temp.shared_1[95]*kernel.shared_1[((threadIdx.x*72) + 4640)]))
+        conv2d_nchw_1[11] = (conv2d_nchw_1[11] + (pad_temp.shared_1[96]*kernel.shared_1[((threadIdx.x*72) + 4640)]))
+        conv2d_nchw_1[12] = (conv2d_nchw_1[12] + (pad_temp.shared_1[97]*kernel.shared_1[((threadIdx.x*72) + 4640)]))
+        conv2d_nchw_1[13] = (conv2d_nchw_1[13] + (pad_temp.shared_1[98]*kernel.shared_1[((threadIdx.x*72) + 4640)]))
+        conv2d_nchw_1[0] = (conv2d_nchw_1[0] + (pad_temp.shared_1[99]*kernel.shared_1[((threadIdx.x*72) + 33)]))
+        conv2d_nchw_1[1] = (conv2d_nchw_1[1] + (pad_temp.shared_1[100]*kernel.shared_1[((threadIdx.x*72) + 33)]))
+        conv2d_nchw_1[2] = (conv2d_nchw_1[2] + (pad_temp.shared_1[101]*kernel.shared_1[((threadIdx.x*72) + 33)]))
+        conv2d_nchw_1[3] = (conv2d_nchw_1[3] + (pad_temp.shared_1[102]*kernel.shared_1[((threadIdx.x*72) + 33)]))
+        conv2d_nchw_1[4] = (conv2d_nchw_1[4] + (pad_temp.shared_1[103]*kernel.shared_1[((threadIdx.x*72) + 33)]))
+        conv2d_nchw_1[5] = (conv2d_nchw_1[5] + (pad_temp.shared_1[104]*kernel.shared_1[((threadIdx.x*72) + 33)]))
+        conv2d_nchw_1[6] = (conv2d_nchw_1[6] + (pad_temp.shared_1[105]*kernel.shared_1[((threadIdx.x*72) + 33)]))
+        conv2d_nchw_1[7] = (conv2d_nchw_1[7] + (pad_temp.shared_1[99]*kernel.shared_1[((threadIdx.x*72) + 4641)]))
+        conv2d_nchw_1[8] = (conv2d_nchw_1[8] + (pad_temp.shared_1[100]*kernel.shared_1[((threadIdx.x*72) + 4641)]))
+        conv2d_nchw_1[9] = (conv2d_nchw_1[9] + (pad_temp.shared_1[101]*kernel.shared_1[((threadIdx.x*72) + 4641)]))
+        conv2d_nchw_1[10] = (conv2d_nchw_1[10] + (pad_temp.shared_1[102]*kernel.shared_1[((threadIdx.x*72) + 4641)]))
+        conv2d_nchw_1[11] = (conv2d_nchw_1[11] + (pad_temp.shared_1[103]*kernel.shared_1[((threadIdx.x*72) + 4641)]))
+        conv2d_nchw_1[12] = (conv2d_nchw_1[12] + (pad_temp.shared_1[104]*kernel.shared_1[((threadIdx.x*72) + 4641)]))
+        conv2d_nchw_1[13] = (conv2d_nchw_1[13] + (pad_temp.shared_1[105]*kernel.shared_1[((threadIdx.x*72) + 4641)]))
+        conv2d_nchw_1[0] = (conv2d_nchw_1[0] + (pad_temp.shared_1[100]*kernel.shared_1[((threadIdx.x*72) + 34)]))
+        conv2d_nchw_1[1] = (conv2d_nchw_1[1] + (pad_temp.shared_1[101]*kernel.shared_1[((threadIdx.x*72) + 34)]))
+        conv2d_nchw_1[2] = (conv2d_nchw_1[2] + (pad_temp.shared_1[102]*kernel.shared_1[((threadIdx.x*72) + 34)]))
+        conv2d_nchw_1[3] = (conv2d_nchw_1[3] + (pad_temp.shared_1[103]*kernel.shared_1[((threadIdx.x*72) + 34)]))
+        conv2d_nchw_1[4] = (conv2d_nchw_1[4] + (pad_temp.shared_1[104]*kernel.shared_1[((threadIdx.x*72) + 34)]))
+        conv2d_nchw_1[5] = (conv2d_nchw_1[5] + (pad_temp.shared_1[105]*kernel.shared_1[((threadIdx.x*72) + 34)]))
+        conv2d_nchw_1[6] = (conv2d_nchw_1[6] + (pad_temp.shared_1[106]*kernel.shared_1[((threadIdx.x*72) + 34)]))
+        conv2d_nchw_1[7] = (conv2d_nchw_1[7] + (pad_temp.shared_1[100]*kernel.shared_1[((threadIdx.x*72) + 4642)]))
+        conv2d_nchw_1[8] = (conv2d_nchw_1[8] + (pad_temp.shared_1[101]*kernel.shared_1[((threadIdx.x*72) + 4642)]))
+        conv2d_nchw_1[9] = (conv2d_nchw_1[9] + (pad_temp.shared_1[102]*kernel.shared_1[((threadIdx.x*72) + 4642)]))
+        conv2d_nchw_1[10] = (conv2d_nchw_1[10] + (pad_temp.shared_1[103]*kernel.shared_1[((threadIdx.x*72) + 4642)]))
+        conv2d_nchw_1[11] = (conv2d_nchw_1[11] + (pad_temp.shared_1[104]*kernel.shared_1[((threadIdx.x*72) + 4642)]))
+        conv2d_nchw_1[12] = (conv2d_nchw_1[12] + (pad_temp.shared_1[105]*kernel.shared_1[((threadIdx.x*72) + 4642)]))
+        conv2d_nchw_1[13] = (conv2d_nchw_1[13] + (pad_temp.shared_1[106]*kernel.shared_1[((threadIdx.x*72) + 4642)]))
+        conv2d_nchw_1[0] = (conv2d_nchw_1[0] + (pad_temp.shared_1[101]*kernel.shared_1[((threadIdx.x*72) + 35)]))
+        conv2d_nchw_1[1] = (conv2d_nchw_1[1] + (pad_temp.shared_1[102]*kernel.shared_1[((threadIdx.x*72) + 35)]))
+        conv2d_nchw_1[2] = (conv2d_nchw_1[2] + (pad_temp.shared_1[103]*kernel.shared_1[((threadIdx.x*72) + 35)]))
+        conv2d_nchw_1[3] = (conv2d_nchw_1[3] + (pad_temp.shared_1[104]*kernel.shared_1[((threadIdx.x*72) + 35)]))
+        conv2d_nchw_1[4] = (conv2d_nchw_1[4] + (pad_temp.shared_1[105]*kernel.shared_1[((threadIdx.x*72) + 35)]))
+        conv2d_nchw_1[5] = (conv2d_nchw_1[5] + (pad_temp.shared_1[106]*kernel.shared_1[((threadIdx.x*72) + 35)]))
+        conv2d_nchw_1[6] = (conv2d_nchw_1[6] + (pad_temp.shared_1[107]*kernel.shared_1[((threadIdx.x*72) + 35)]))
+        conv2d_nchw_1[7] = (conv2d_nchw_1[7] + (pad_temp.shared_1[101]*kernel.shared_1[((threadIdx.x*72) + 4643)]))
+        conv2d_nchw_1[8] = (conv2d_nchw_1[8] + (pad_temp.shared_1[102]*kernel.shared_1[((threadIdx.x*72) + 4643)]))
+        conv2d_nchw_1[9] = (conv2d_nchw_1[9] + (pad_temp.shared_1[103]*kernel.shared_1[((threadIdx.x*72) + 4643)]))
+        conv2d_nchw_1[10] = (conv2d_nchw_1[10] + (pad_temp.shared_1[104]*kernel.shared_1[((threadIdx.x*72) + 4643)]))
+        conv2d_nchw_1[11] = (conv2d_nchw_1[11] + (pad_temp.shared_1[105]*kernel.shared_1[((threadIdx.x*72) + 4643)]))
+        conv2d_nchw_1[12] = (conv2d_nchw_1[12] + (pad_temp.shared_1[106]*kernel.shared_1[((threadIdx.x*72) + 4643)]))
+        conv2d_nchw_1[13] = (conv2d_nchw_1[13] + (pad_temp.shared_1[107]*kernel.shared_1[((threadIdx.x*72) + 4643)]))
+        conv2d_nchw_1[0] = (conv2d_nchw_1[0] + (pad_temp.shared_1[108]*kernel.shared_1[((threadIdx.x*72) + 36)]))
+        conv2d_nchw_1[1] = (conv2d_nchw_1[1] + (pad_temp.shared_1[109]*kernel.shared_1[((threadIdx.x*72) + 36)]))
+        conv2d_nchw_1[2] = (conv2d_nchw_1[2] + (pad_temp.shared_1[110]*kernel.shared_1[((threadIdx.x*72) + 36)]))
+        conv2d_nchw_1[3] = (conv2d_nchw_1[3] + (pad_temp.shared_1[111]*kernel.shared_1[((threadIdx.x*72) + 36)]))
+        conv2d_nchw_1[4] = (conv2d_nchw_1[4] + (pad_temp.shared_1[112]*kernel.shared_1[((threadIdx.x*72) + 36)]))
+        conv2d_nchw_1[5] = (conv2d_nchw_1[5] + (pad_temp.shared_1[113]*kernel.shared_1[((threadIdx.x*72) + 36)]))
+        conv2d_nchw_1[6] = (conv2d_nchw_1[6] + (pad_temp.shared_1[114]*kernel.shared_1[((threadIdx.x*72) + 36)]))
+        conv2d_nchw_1[7] = (conv2d_nchw_1[7] + (pad_temp.shared_1[108]*kernel.shared_1[((threadIdx.x*72) + 4644)]))
+        conv2d_nchw_1[8] = (conv2d_nchw_1[8] + (pad_temp.shared_1[109]*kernel.shared_1[((threadIdx.x*72) + 4644)]))
+        conv2d_nchw_1[9] = (conv2d_nchw_1[9] + (pad_temp.shared_1[110]*kernel.shared_1[((threadIdx.x*72) + 4644)]))
+        conv2d_nchw_1[10] = (conv2d_nchw_1[10] + (pad_temp.shared_1[111]*kernel.shared_1[((threadIdx.x*72) + 4644)]))
+        conv2d_nchw_1[11] = (conv2d_nchw_1[11] + (pad_temp.shared_1[112]*kernel.shared_1[((threadIdx.x*72) + 4644)]))
+        conv2d_nchw_1[12] = (conv2d_nchw_1[12] + (pad_temp.shared_1[113]*kernel.shared_1[((threadIdx.x*72) + 4644)]))
+        conv2d_nchw_1[13] = (conv2d_nchw_1[13] + (pad_temp.shared_1[114]*kernel.shared_1[((threadIdx.x*72) + 4644)]))
+        conv2d_nchw_1[0] = (conv2d_nchw_1[0] + (pad_temp.shared_1[109]*kernel.shared_1[((threadIdx.x*72) + 37)]))
+        conv2d_nchw_1[1] = (conv2d_nchw_1[1] + (pad_temp.shared_1[110]*kernel.shared_1[((threadIdx.x*72) + 37)]))
+        conv2d_nchw_1[2] = (conv2d_nchw_1[2] + (pad_temp.shared_1[111]*kernel.shared_1[((threadIdx.x*72) + 37)]))
+        conv2d_nchw_1[3] = (conv2d_nchw_1[3] + (pad_temp.shared_1[112]*kernel.shared_1[((threadIdx.x*72) + 37)]))
+        conv2d_nchw_1[4] = (conv2d_nchw_1[4] + (pad_temp.shared_1[113]*kernel.shared_1[((threadIdx.x*72) + 37)]))
+        conv2d_nchw_1[5] = (conv2d_nchw_1[5] + (pad_temp.shared_1[114]*kernel.shared_1[((threadIdx.x*72) + 37)]))
+        conv2d_nchw_1[6] = (conv2d_nchw_1[6] + (pad_temp.shared_1[115]*kernel.shared_1[((threadIdx.x*72) + 37)]))
+        conv2d_nchw_1[7] = (conv2d_nchw_1[7] + (pad_temp.shared_1[109]*kernel.shared_1[((threadIdx.x*72) + 4645)]))
+        conv2d_nchw_1[8] = (conv2d_nchw_1[8] + (pad_temp.shared_1[110]*kernel.shared_1[((threadIdx.x*72) + 4645)]))
+        conv2d_nchw_1[9] = (conv2d_nchw_1[9] + (pad_temp.shared_1[111]*kernel.shared_1[((threadIdx.x*72) + 4645)]))
+        conv2d_nchw_1[10] = (conv2d_nchw_1[10] + (pad_temp.shared_1[112]*kernel.shared_1[((threadIdx.x*72) + 4645)]))
+        conv2d_nchw_1[11] = (conv2d_nchw_1[11] + (pad_temp.shared_1[113]*kernel.shared_1[((threadIdx.x*72) + 4645)]))
+        conv2d_nchw_1[12] = (conv2d_nchw_1[12] + (pad_temp.shared_1[114]*kernel.shared_1[((threadIdx.x*72) + 4645)]))
+        conv2d_nchw_1[13] = (conv2d_nchw_1[13] + (pad_temp.shared_1[115]*kernel.shared_1[((threadIdx.x*72) + 4645)]))
+        conv2d_nchw_1[0] = (conv2d_nchw_1[0] + (pad_temp.shared_1[110]*kernel.shared_1[((threadIdx.x*72) + 38)]))
+        conv2d_nchw_1[1] = (conv2d_nchw_1[1] + (pad_temp.shared_1[111]*kernel.shared_1[((threadIdx.x*72) + 38)]))
+        conv2d_nchw_1[2] = (conv2d_nchw_1[2] + (pad_temp.shared_1[112]*kernel.shared_1[((threadIdx.x*72) + 38)]))
+        conv2d_nchw_1[3] = (conv2d_nchw_1[3] + (pad_temp.shared_1[113]*kernel.shared_1[((threadIdx.x*72) + 38)]))
+        conv2d_nchw_1[4] = (conv2d_nchw_1[4] + (pad_temp.shared_1[114]*kernel.shared_1[((threadIdx.x*72) + 38)]))
+        conv2d_nchw_1[5] = (conv2d_nchw_1[5] + (pad_temp.shared_1[115]*kernel.shared_1[((threadIdx.x*72) + 38)]))
+        conv2d_nchw_1[6] = (conv2d_nchw_1[6] + (pad_temp.shared_1[116]*kernel.shared_1[((threadIdx.x*72) + 38)]))
+        conv2d_nchw_1[7] = (conv2d_nchw_1[7] + (pad_temp.shared_1[110]*kernel.shared_1[((threadIdx.x*72) + 4646)]))
+        conv2d_nchw_1[8] = (conv2d_nchw_1[8] + (pad_temp.shared_1[111]*kernel.shared_1[((threadIdx.x*72) + 4646)]))
+        conv2d_nchw_1[9] = (conv2d_nchw_1[9] + (pad_temp.shared_1[112]*kernel.shared_1[((threadIdx.x*72) + 4646)]))
+        conv2d_nchw_1[10] = (conv2d_nchw_1[10] + (pad_temp.shared_1[113]*kernel.shared_1[((threadIdx.x*72) + 4646)]))
+        conv2d_nchw_1[11] = (conv2d_nchw_1[11] + (pad_temp.shared_1[114]*kernel.shared_1[((threadIdx.x*72) + 4646)]))
+        conv2d_nchw_1[12] = (conv2d_nchw_1[12] + (pad_temp.shared_1[115]*kernel.shared_1[((threadIdx.x*72) + 4646)]))
+        conv2d_nchw_1[13] = (conv2d_nchw_1[13] + (pad_temp.shared_1[116]*kernel.shared_1[((threadIdx.x*72) + 4646)]))
+        conv2d_nchw_1[0] = (conv2d_nchw_1[0] + (pad_temp.shared_1[117]*kernel.shared_1[((threadIdx.x*72) + 39)]))
+        conv2d_nchw_1[1] = (conv2d_nchw_1[1] + (pad_temp.shared_1[118]*kernel.shared_1[((threadIdx.x*72) + 39)]))
+        conv2d_nchw_1[2] = (conv2d_nchw_1[2] + (pad_temp.shared_1[119]*kernel.shared_1[((threadIdx.x*72) + 39)]))
+        conv2d_nchw_1[3] = (conv2d_nchw_1[3] + (pad_temp.shared_1[120]*kernel.shared_1[((threadIdx.x*72) + 39)]))
+        conv2d_nchw_1[4] = (conv2d_nchw_1[4] + (pad_temp.shared_1[121]*kernel.shared_1[((threadIdx.x*72) + 39)]))
+        conv2d_nchw_1[5] = (conv2d_nchw_1[5] + (pad_temp.shared_1[122]*kernel.shared_1[((threadIdx.x*72) + 39)]))
+        conv2d_nchw_1[6] = (conv2d_nchw_1[6] + (pad_temp.shared_1[123]*kernel.shared_1[((threadIdx.x*72) + 39)]))
+        conv2d_nchw_1[7] = (conv2d_nchw_1[7] + (pad_temp.shared_1[117]*kernel.shared_1[((threadIdx.x*72) + 4647)]))
+        conv2d_nchw_1[8] = (conv2d_nchw_1[8] + (pad_temp.shared_1[118]*kernel.shared_1[((threadIdx.x*72) + 4647)]))
+        conv2d_nchw_1[9] = (conv2d_nchw_1[9] + (pad_temp.shared_1[119]*kernel.shared_1[((threadIdx.x*72) + 4647)]))
+        conv2d_nchw_1[10] = (conv2d_nchw_1[10] + (pad_temp.shared_1[120]*kernel.shared_1[((threadIdx.x*72) + 4647)]))
+        conv2d_nchw_1[11] = (conv2d_nchw_1[11] + (pad_temp.shared_1[121]*kernel.shared_1[((threadIdx.x*72) + 4647)]))
+        conv2d_nchw_1[12] = (conv2d_nchw_1[12] + (pad_temp.shared_1[122]*kernel.shared_1[((threadIdx.x*72) + 4647)]))
+        conv2d_nchw_1[13] = (conv2d_nchw_1[13] + (pad_temp.shared_1[123]*kernel.shared_1[((threadIdx.x*72) + 4647)]))
+        conv2d_nchw_1[0] = (conv2d_nchw_1[0] + (pad_temp.shared_1[118]*kernel.shared_1[((threadIdx.x*72) + 40)]))
+        conv2d_nchw_1[1] = (conv2d_nchw_1[1] + (pad_temp.shared_1[119]*kernel.shared_1[((threadIdx.x*72) + 40)]))
+        conv2d_nchw_1[2] = (conv2d_nchw_1[2] + (pad_temp.shared_1[120]*kernel.shared_1[((threadIdx.x*72) + 40)]))
+        conv2d_nchw_1[3] = (conv2d_nchw_1[3] + (pad_temp.shared_1[121]*kernel.shared_1[((threadIdx.x*72) + 40)]))
+        conv2d_nchw_1[4] = (conv2d_nchw_1[4] + (pad_temp.shared_1[122]*kernel.shared_1[((threadIdx.x*72) + 40)]))
+        conv2d_nchw_1[5] = (conv2d_nchw_1[5] + (pad_temp.shared_1[123]*kernel.shared_1[((threadIdx.x*72) + 40)]))
+        conv2d_nchw_1[6] = (conv2d_nchw_1[6] + (pad_temp.shared_1[124]*kernel.shared_1[((threadIdx.x*72) + 40)]))
+        conv2d_nchw_1[7] = (conv2d_nchw_1[7] + (pad_temp.shared_1[118]*kernel.shared_1[((threadIdx.x*72) + 4648)]))
+        conv2d_nchw_1[8] = (conv2d_nchw_1[8] + (pad_temp.shared_1[119]*kernel.shared_1[((threadIdx.x*72) + 4648)]))
+        conv2d_nchw_1[9] = (conv2d_nchw_1[9] + (pad_temp.shared_1[120]*kernel.shared_1[((threadIdx.x*72) + 4648)]))
+        conv2d_nchw_1[10] = (conv2d_nchw_1[10] + (pad_temp.shared_1[121]*kernel.shared_1[((threadIdx.x*72) + 4648)]))
+        conv2d_nchw_1[11] = (conv2d_nchw_1[11] + (pad_temp.shared_1[122]*kernel.shared_1[((threadIdx.x*72) + 4648)]))
+        conv2d_nchw_1[12] = (conv2d_nchw_1[12] + (pad_temp.shared_1[123]*kernel.shared_1[((threadIdx.x*72) + 4648)]))
+        conv2d_nchw_1[13] = (conv2d_nchw_1[13] + (pad_temp.shared_1[124]*kernel.shared_1[((threadIdx.x*72) + 4648)]))
+        conv2d_nchw_1[0] = (conv2d_nchw_1[0] + (pad_temp.shared_1[119]*kernel.shared_1[((threadIdx.x*72) + 41)]))
+        conv2d_nchw_1[1] = (conv2d_nchw_1[1] + (pad_temp.shared_1[120]*kernel.shared_1[((threadIdx.x*72) + 41)]))
+        conv2d_nchw_1[2] = (conv2d_nchw_1[2] + (pad_temp.shared_1[121]*kernel.shared_1[((threadIdx.x*72) + 41)]))
+        conv2d_nchw_1[3] = (conv2d_nchw_1[3] + (pad_temp.shared_1[122]*kernel.shared_1[((threadIdx.x*72) + 41)]))
+        conv2d_nchw_1[4] = (conv2d_nchw_1[4] + (pad_temp.shared_1[123]*kernel.shared_1[((threadIdx.x*72) + 41)]))
+        conv2d_nchw_1[5] = (conv2d_nchw_1[5] + (pad_temp.shared_1[124]*kernel.shared_1[((threadIdx.x*72) + 41)]))
+        conv2d_nchw_1[6] = (conv2d_nchw_1[6] + (pad_temp.shared_1[125]*kernel.shared_1[((threadIdx.x*72) + 41)]))
+        conv2d_nchw_1[7] = (conv2d_nchw_1[7] + (pad_temp.shared_1[119]*kernel.shared_1[((threadIdx.x*72) + 4649)]))
+        conv2d_nchw_1[8] = (conv2d_nchw_1[8] + (pad_temp.shared_1[120]*kernel.shared_1[((threadIdx.x*72) + 4649)]))
+        conv2d_nchw_1[9] = (conv2d_nchw_1[9] + (pad_temp.shared_1[121]*kernel.shared_1[((threadIdx.x*72) + 4649)]))
+        conv2d_nchw_1[10] = (conv2d_nchw_1[10] + (pad_temp.shared_1[122]*kernel.shared_1[((threadIdx.x*72) + 4649)]))
+        conv2d_nchw_1[11] = (conv2d_nchw_1[11] + (pad_temp.shared_1[123]*kernel.shared_1[((threadIdx.x*72) + 4649)]))
+        conv2d_nchw_1[12] = (conv2d_nchw_1[12] + (pad_temp.shared_1[124]*kernel.shared_1[((threadIdx.x*72) + 4649)]))
+        conv2d_nchw_1[13] = (conv2d_nchw_1[13] + (pad_temp.shared_1[125]*kernel.shared_1[((threadIdx.x*72) + 4649)]))
+        conv2d_nchw_1[0] = (conv2d_nchw_1[0] + (pad_temp.shared_1[126]*kernel.shared_1[((threadIdx.x*72) + 42)]))
+        conv2d_nchw_1[1] = (conv2d_nchw_1[1] + (pad_temp.shared_1[127]*kernel.shared_1[((threadIdx.x*72) + 42)]))
+        conv2d_nchw_1[2] = (conv2d_nchw_1[2] + (pad_temp.shared_1[128]*kernel.shared_1[((threadIdx.x*72) + 42)]))
+        conv2d_nchw_1[3] = (conv2d_nchw_1[3] + (pad_temp.shared_1[129]*kernel.shared_1[((threadIdx.x*72) + 42)]))
+        conv2d_nchw_1[4] = (conv2d_nchw_1[4] + (pad_temp.shared_1[130]*kernel.shared_1[((threadIdx.x*72) + 42)]))
+        conv2d_nchw_1[5] = (conv2d_nchw_1[5] + (pad_temp.shared_1[131]*kernel.shared_1[((threadIdx.x*72) + 42)]))
+        conv2d_nchw_1[6] = (conv2d_nchw_1[6] + (pad_temp.shared_1[132]*kernel.shared_1[((threadIdx.x*72) + 42)]))
+        conv2d_nchw_1[7] = (conv2d_nchw_1[7] + (pad_temp.shared_1[126]*kernel.shared_1[((threadIdx.x*72) + 4650)]))
+        conv2d_nchw_1[8] = (conv2d_nchw_1[8] + (pad_temp.shared_1[127]*kernel.shared_1[((threadIdx.x*72) + 4650)]))
+        conv2d_nchw_1[9] = (conv2d_nchw_1[9] + (pad_temp.shared_1[128]*kernel.shared_1[((threadIdx.x*72) + 4650)]))
+        conv2d_nchw_1[10] = (conv2d_nchw_1[10] + (pad_temp.shared_1[129]*kernel.shared_1[((threadIdx.x*72) + 4650)]))
+        conv2d_nchw_1[11] = (conv2d_nchw_1[11] + (pad_temp.shared_1[130]*kernel.shared_1[((threadIdx.x*72) + 4650)]))
+        conv2d_nchw_1[12] = (conv2d_nchw_1[12] + (pad_temp.shared_1[131]*kernel.shared_1[((threadIdx.x*72) + 4650)]))
+        conv2d_nchw_1[13] = (conv2d_nchw_1[13] + (pad_temp.shared_1[132]*kernel.shared_1[((threadIdx.x*72) + 4650)]))
+        conv2d_nchw_1[0] = (conv2d_nchw_1[0] + (pad_temp.shared_1[127]*kernel.shared_1[((threadIdx.x*72) + 43)]))
+        conv2d_nchw_1[1] = (conv2d_nchw_1[1] + (pad_temp.shared_1[128]*kernel.shared_1[((threadIdx.x*72) + 43)]))
+        conv2d_nchw_1[2] = (conv2d_nchw_1[2] + (pad_temp.shared_1[129]*kernel.shared_1[((threadIdx.x*72) + 43)]))
+        conv2d_nchw_1[3] = (conv2d_nchw_1[3] + (pad_temp.shared_1[130]*kernel.shared_1[((threadIdx.x*72) + 43)]))
+        conv2d_nchw_1[4] = (conv2d_nchw_1[4] + (pad_temp.shared_1[131]*kernel.shared_1[((threadIdx.x*72) + 43)]))
+        conv2d_nchw_1[5] = (conv2d_nchw_1[5] + (pad_temp.shared_1[132]*kernel.shared_1[((threadIdx.x*72) + 43)]))
+        conv2d_nchw_1[6] = (conv2d_nchw_1[6] + (pad_temp.shared_1[133]*kernel.shared_1[((threadIdx.x*72) + 43)]))
+        conv2d_nchw_1[7] = (conv2d_nchw_1[7] + (pad_temp.shared_1[127]*kernel.shared_1[((threadIdx.x*72) + 4651)]))
+        conv2d_nchw_1[8] = (conv2d_nchw_1[8] + (pad_temp.shared_1[128]*kernel.shared_1[((threadIdx.x*72) + 4651)]))
+        conv2d_nchw_1[9] = (conv2d_nchw_1[9] + (pad_temp.shared_1[129]*kernel.shared_1[((threadIdx.x*72) + 4651)]))
+        conv2d_nchw_1[10] = (conv2d_nchw_1[10] + (pad_temp.shared_1[130]*kernel.shared_1[((threadIdx.x*72) + 4651)]))
+        conv2d_nchw_1[11] = (conv2d_nchw_1[11] + (pad_temp.shared_1[131]*kernel.shared_1[((threadIdx.x*72) + 4651)]))
+        conv2d_nchw_1[12] = (conv2d_nchw_1[12] + (pad_temp.shared_1[132]*kernel.shared_1[((threadIdx.x*72) + 4651)]))
+        conv2d_nchw_1[13] = (conv2d_nchw_1[13] + (pad_temp.shared_1[133]*kernel.shared_1[((threadIdx.x*72) + 4651)]))
+        conv2d_nchw_1[0] = (conv2d_nchw_1[0] + (pad_temp.shared_1[128]*kernel.shared_1[((threadIdx.x*72) + 44)]))
+        conv2d_nchw_1[1] = (conv2d_nchw_1[1] + (pad_temp.shared_1[129]*kernel.shared_1[((threadIdx.x*72) + 44)]))
+        conv2d_nchw_1[2] = (conv2d_nchw_1[2] + (pad_temp.shared_1[130]*kernel.shared_1[((threadIdx.x*72) + 44)]))
+        conv2d_nchw_1[3] = (conv2d_nchw_1[3] + (pad_temp.shared_1[131]*kernel.shared_1[((threadIdx.x*72) + 44)]))
+        conv2d_nchw_1[4] = (conv2d_nchw_1[4] + (pad_temp.shared_1[132]*kernel.shared_1[((threadIdx.x*72) + 44)]))
+        conv2d_nchw_1[5] = (conv2d_nchw_1[5] + (pad_temp.shared_1[133]*kernel.shared_1[((threadIdx.x*72) + 44)]))
+        conv2d_nchw_1[6] = (conv2d_nchw_1[6] + (pad_temp.shared_1[134]*kernel.shared_1[((threadIdx.x*72) + 44)]))
+        conv2d_nchw_1[7] = (conv2d_nchw_1[7] + (pad_temp.shared_1[128]*kernel.shared_1[((threadIdx.x*72) + 4652)]))
+        conv2d_nchw_1[8] = (conv2d_nchw_1[8] + (pad_temp.shared_1[129]*kernel.shared_1[((threadIdx.x*72) + 4652)]))
+        conv2d_nchw_1[9] = (conv2d_nchw_1[9] + (pad_temp.shared_1[130]*kernel.shared_1[((threadIdx.x*72) + 4652)]))
+        conv2d_nchw_1[10] = (conv2d_nchw_1[10] + (pad_temp.shared_1[131]*kernel.shared_1[((threadIdx.x*72) + 4652)]))
+        conv2d_nchw_1[11] = (conv2d_nchw_1[11] + (pad_temp.shared_1[132]*kernel.shared_1[((threadIdx.x*72) + 4652)]))
+        conv2d_nchw_1[12] = (conv2d_nchw_1[12] + (pad_temp.shared_1[133]*kernel.shared_1[((threadIdx.x*72) + 4652)]))
+        conv2d_nchw_1[13] = (conv2d_nchw_1[13] + (pad_temp.shared_1[134]*kernel.shared_1[((threadIdx.x*72) + 4652)]))
+        conv2d_nchw_1[0] = (conv2d_nchw_1[0] + (pad_temp.shared_1[135]*kernel.shared_1[((threadIdx.x*72) + 45)]))
+        conv2d_nchw_1[1] = (conv2d_nchw_1[1] + (pad_temp.shared_1[136]*kernel.shared_1[((threadIdx.x*72) + 45)]))
+        conv2d_nchw_1[2] = (conv2d_nchw_1[2] + (pad_temp.shared_1[137]*kernel.shared_1[((threadIdx.x*72) + 45)]))
+        conv2d_nchw_1[3] = (conv2d_nchw_1[3] + (pad_temp.shared_1[138]*kernel.shared_1[((threadIdx.x*72) + 45)]))
+        conv2d_nchw_1[4] = (conv2d_nchw_1[4] + (pad_temp.shared_1[139]*kernel.shared_1[((threadIdx.x*72) + 45)]))
+        conv2d_nchw_1[5] = (conv2d_nchw_1[5] + (pad_temp.shared_1[140]*kernel.shared_1[((threadIdx.x*72) + 45)]))
+        conv2d_nchw_1[6] = (conv2d_nchw_1[6] + (pad_temp.shared_1[141]*kernel.shared_1[((threadIdx.x*72) + 45)]))
+        conv2d_nchw_1[7] = (conv2d_nchw_1[7] + (pad_temp.shared_1[135]*kernel.shared_1[((threadIdx.x*72) + 4653)]))
+        conv2d_nchw_1[8] = (conv2d_nchw_1[8] + (pad_temp.shared_1[136]*kernel.shared_1[((threadIdx.x*72) + 4653)]))
+        conv2d_nchw_1[9] = (conv2d_nchw_1[9] + (pad_temp.shared_1[137]*kernel.shared_1[((threadIdx.x*72) + 4653)]))
+        conv2d_nchw_1[10] = (conv2d_nchw_1[10] + (pad_temp.shared_1[138]*kernel.shared_1[((threadIdx.x*72) + 4653)]))
+        conv2d_nchw_1[11] = (conv2d_nchw_1[11] + (pad_temp.shared_1[139]*kernel.shared_1[((threadIdx.x*72) + 4653)]))
+        conv2d_nchw_1[12] = (conv2d_nchw_1[12] + (pad_temp.shared_1[140]*kernel.shared_1[((threadIdx.x*72) + 4653)]))
+        conv2d_nchw_1[13] = (conv2d_nchw_1[13] + (pad_temp.shared_1[141]*kernel.shared_1[((threadIdx.x*72) + 4653)]))
+        conv2d_nchw_1[0] = (conv2d_nchw_1[0] + (pad_temp.shared_1[136]*kernel.shared_1[((threadIdx.x*72) + 46)]))
+        conv2d_nchw_1[1] = (conv2d_nchw_1[1] + (pad_temp.shared_1[137]*kernel.shared_1[((threadIdx.x*72) + 46)]))
+        conv2d_nchw_1[2] = (conv2d_nchw_1[2] + (pad_temp.shared_1[138]*kernel.shared_1[((threadIdx.x*72) + 46)]))
+        conv2d_nchw_1[3] = (conv2d_nchw_1[3] + (pad_temp.shared_1[139]*kernel.shared_1[((threadIdx.x*72) + 46)]))
+        conv2d_nchw_1[4] = (conv2d_nchw_1[4] + (pad_temp.shared_1[140]*kernel.shared_1[((threadIdx.x*72) + 46)]))
+        conv2d_nchw_1[5] = (conv2d_nchw_1[5] + (pad_temp.shared_1[141]*kernel.shared_1[((threadIdx.x*72) + 46)]))
+        conv2d_nchw_1[6] = (conv2d_nchw_1[6] + (pad_temp.shared_1[142]*kernel.shared_1[((threadIdx.x*72) + 46)]))
+        conv2d_nchw_1[7] = (conv2d_nchw_1[7] + (pad_temp.shared_1[136]*kernel.shared_1[((threadIdx.x*72) + 4654)]))
+        conv2d_nchw_1[8] = (conv2d_nchw_1[8] + (pad_temp.shared_1[137]*kernel.shared_1[((threadIdx.x*72) + 4654)]))
+        conv2d_nchw_1[9] = (conv2d_nchw_1[9] + (pad_temp.shared_1[138]*kernel.shared_1[((threadIdx.x*72) + 4654)]))
+        conv2d_nchw_1[10] = (conv2d_nchw_1[10] + (pad_temp.shared_1[139]*kernel.shared_1[((threadIdx.x*72) + 4654)]))
+        conv2d_nchw_1[11] = (conv2d_nchw_1[11] + (pad_temp.shared_1[140]*kernel.shared_1[((threadIdx.x*72) + 4654)]))
+        conv2d_nchw_1[12] = (conv2d_nchw_1[12] + (pad_temp.shared_1[141]*kernel.shared_1[((threadIdx.x*72) + 4654)]))
+        conv2d_nchw_1[13] = (conv2d_nchw_1[13] + (pad_temp.shared_1[142]*kernel.shared_1[((threadIdx.x*72) + 4654)]))
+        conv2d_nchw_1[0] = (conv2d_nchw_1[0] + (pad_temp.shared_1[137]*kernel.shared_1[((threadIdx.x*72) + 47)]))
+        conv2d_nchw_1[1] = (conv2d_nchw_1[1] + (pad_temp.shared_1[138]*kernel.shared_1[((threadIdx.x*72) + 47)]))
+        conv2d_nchw_1[2] = (conv2d_nchw_1[2] + (pad_temp.shared_1[139]*kernel.shared_1[((threadIdx.x*72) + 47)]))
+        conv2d_nchw_1[3] = (conv2d_nchw_1[3] + (pad_temp.shared_1[140]*kernel.shared_1[((threadIdx.x*72) + 47)]))
+        conv2d_nchw_1[4] = (conv2d_nchw_1[4] + (pad_temp.shared_1[141]*kernel.shared_1[((threadIdx.x*72) + 47)]))
+        conv2d_nchw_1[5] = (conv2d_nchw_1[5] + (pad_temp.shared_1[142]*kernel.shared_1[((threadIdx.x*72) + 47)]))
+        conv2d_nchw_1[6] = (conv2d_nchw_1[6] + (pad_temp.shared_1[143]*kernel.shared_1[((threadIdx.x*72) + 47)]))
+        conv2d_nchw_1[7] = (conv2d_nchw_1[7] + (pad_temp.shared_1[137]*kernel.shared_1[((threadIdx.x*72) + 4655)]))
+        conv2d_nchw_1[8] = (conv2d_nchw_1[8] + (pad_temp.shared_1[138]*kernel.shared_1[((threadIdx.x*72) + 4655)]))
+        conv2d_nchw_1[9] = (conv2d_nchw_1[9] + (pad_temp.shared_1[139]*kernel.shared_1[((threadIdx.x*72) + 4655)]))
+        conv2d_nchw_1[10] = (conv2d_nchw_1[10] + (pad_temp.shared_1[140]*kernel.shared_1[((threadIdx.x*72) + 4655)]))
+        conv2d_nchw_1[11] = (conv2d_nchw_1[11] + (pad_temp.shared_1[141]*kernel.shared_1[((threadIdx.x*72) + 4655)]))
+        conv2d_nchw_1[12] = (conv2d_nchw_1[12] + (pad_temp.shared_1[142]*kernel.shared_1[((threadIdx.x*72) + 4655)]))
+        conv2d_nchw_1[13] = (conv2d_nchw_1[13] + (pad_temp.shared_1[143]*kernel.shared_1[((threadIdx.x*72) + 4655)]))
+        conv2d_nchw_1[0] = (conv2d_nchw_1[0] + (pad_temp.shared_1[144]*kernel.shared_1[((threadIdx.x*72) + 48)]))
+        conv2d_nchw_1[1] = (conv2d_nchw_1[1] + (pad_temp.shared_1[145]*kernel.shared_1[((threadIdx.x*72) + 48)]))
+        conv2d_nchw_1[2] = (conv2d_nchw_1[2] + (pad_temp.shared_1[146]*kernel.shared_1[((threadIdx.x*72) + 48)]))
+        conv2d_nchw_1[3] = (conv2d_nchw_1[3] + (pad_temp.shared_1[147]*kernel.shared_1[((threadIdx.x*72) + 48)]))
+        conv2d_nchw_1[4] = (conv2d_nchw_1[4] + (pad_temp.shared_1[148]*kernel.shared_1[((threadIdx.x*72) + 48)]))
+        conv2d_nchw_1[5] = (conv2d_nchw_1[5] + (pad_temp.shared_1[149]*kernel.shared_1[((threadIdx.x*72) + 48)]))
+        conv2d_nchw_1[6] = (conv2d_nchw_1[6] + (pad_temp.shared_1[150]*kernel.shared_1[((threadIdx.x*72) + 48)]))
+        conv2d_nchw_1[7] = (conv2d_nchw_1[7] + (pad_temp.shared_1[144]*kernel.shared_1[((threadIdx.x*72) + 4656)]))
+        conv2d_nchw_1[8] = (conv2d_nchw_1[8] + (pad_temp.shared_1[145]*kernel.shared_1[((threadIdx.x*72) + 4656)]))
+        conv2d_nchw_1[9] = (conv2d_nchw_1[9] + (pad_temp.shared_1[146]*kernel.shared_1[((threadIdx.x*72) + 4656)]))
+        conv2d_nchw_1[10] = (conv2d_nchw_1[10] + (pad_temp.shared_1[147]*kernel.shared_1[((threadIdx.x*72) + 4656)]))
+        conv2d_nchw_1[11] = (conv2d_nchw_1[11] + (pad_temp.shared_1[148]*kernel.shared_1[((threadIdx.x*72) + 4656)]))
+        conv2d_nchw_1[12] = (conv2d_nchw_1[12] + (pad_temp.shared_1[149]*kernel.shared_1[((threadIdx.x*72) + 4656)]))
+        conv2d_nchw_1[13] = (conv2d_nchw_1[13] + (pad_temp.shared_1[150]*kernel.shared_1[((threadIdx.x*72) + 4656)]))
+        conv2d_nchw_1[0] = (conv2d_nchw_1[0] + (pad_temp.shared_1[145]*kernel.shared_1[((threadIdx.x*72) + 49)]))
+        conv2d_nchw_1[1] = (conv2d_nchw_1[1] + (pad_temp.shared_1[146]*kernel.shared_1[((threadIdx.x*72) + 49)]))
+        conv2d_nchw_1[2] = (conv2d_nchw_1[2] + (pad_temp.shared_1[147]*kernel.shared_1[((threadIdx.x*72) + 49)]))
+        conv2d_nchw_1[3] = (conv2d_nchw_1[3] + (pad_temp.shared_1[148]*kernel.shared_1[((threadIdx.x*72) + 49)]))
+        conv2d_nchw_1[4] = (conv2d_nchw_1[4] + (pad_temp.shared_1[149]*kernel.shared_1[((threadIdx.x*72) + 49)]))
+        conv2d_nchw_1[5] = (conv2d_nchw_1[5] + (pad_temp.shared_1[150]*kernel.shared_1[((threadIdx.x*72) + 49)]))
+        conv2d_nchw_1[6] = (conv2d_nchw_1[6] + (pad_temp.shared_1[151]*kernel.shared_1[((threadIdx.x*72) + 49)]))
+        conv2d_nchw_1[7] = (conv2d_nchw_1[7] + (pad_temp.shared_1[145]*kernel.shared_1[((threadIdx.x*72) + 4657)]))
+        conv2d_nchw_1[8] = (conv2d_nchw_1[8] + (pad_temp.shared_1[146]*kernel.shared_1[((threadIdx.x*72) + 4657)]))
+        conv2d_nchw_1[9] = (conv2d_nchw_1[9] + (pad_temp.shared_1[147]*kernel.shared_1[((threadIdx.x*72) + 4657)]))
+        conv2d_nchw_1[10] = (conv2d_nchw_1[10] + (pad_temp.shared_1[148]*kernel.shared_1[((threadIdx.x*72) + 4657)]))
+        conv2d_nchw_1[11] = (conv2d_nchw_1[11] + (pad_temp.shared_1[149]*kernel.shared_1[((threadIdx.x*72) + 4657)]))
+        conv2d_nchw_1[12] = (conv2d_nchw_1[12] + (pad_temp.shared_1[150]*kernel.shared_1[((threadIdx.x*72) + 4657)]))
+        conv2d_nchw_1[13] = (conv2d_nchw_1[13] + (pad_temp.shared_1[151]*kernel.shared_1[((threadIdx.x*72) + 4657)]))
+        conv2d_nchw_1[0] = (conv2d_nchw_1[0] + (pad_temp.shared_1[146]*kernel.shared_1[((threadIdx.x*72) + 50)]))
+        conv2d_nchw_1[1] = (conv2d_nchw_1[1] + (pad_temp.shared_1[147]*kernel.shared_1[((threadIdx.x*72) + 50)]))
+        conv2d_nchw_1[2] = (conv2d_nchw_1[2] + (pad_temp.shared_1[148]*kernel.shared_1[((threadIdx.x*72) + 50)]))
+        conv2d_nchw_1[3] = (conv2d_nchw_1[3] + (pad_temp.shared_1[149]*kernel.shared_1[((threadIdx.x*72) + 50)]))
+        conv2d_nchw_1[4] = (conv2d_nchw_1[4] + (pad_temp.shared_1[150]*kernel.shared_1[((threadIdx.x*72) + 50)]))
+        conv2d_nchw_1[5] = (conv2d_nchw_1[5] + (pad_temp.shared_1[151]*kernel.shared_1[((threadIdx.x*72) + 50)]))
+        conv2d_nchw_1[6] = (conv2d_nchw_1[6] + (pad_temp.shared_1[152]*kernel.shared_1[((threadIdx.x*72) + 50)]))
+        conv2d_nchw_1[7] = (conv2d_nchw_1[7] + (pad_temp.shared_1[146]*kernel.shared_1[((threadIdx.x*72) + 4658)]))
+        conv2d_nchw_1[8] = (conv2d_nchw_1[8] + (pad_temp.shared_1[147]*kernel.shared_1[((threadIdx.x*72) + 4658)]))
+        conv2d_nchw_1[9] = (conv2d_nchw_1[9] + (pad_temp.shared_1[148]*kernel.shared_1[((threadIdx.x*72) + 4658)]))
+        conv2d_nchw_1[10] = (conv2d_nchw_1[10] + (pad_temp.shared_1[149]*kernel.shared_1[((threadIdx.x*72) + 4658)]))
+        conv2d_nchw_1[11] = (conv2d_nchw_1[11] + (pad_temp.shared_1[150]*kernel.shared_1[((threadIdx.x*72) + 4658)]))
+        conv2d_nchw_1[12] = (conv2d_nchw_1[12] + (pad_temp.shared_1[151]*kernel.shared_1[((threadIdx.x*72) + 4658)]))
+        conv2d_nchw_1[13] = (conv2d_nchw_1[13] + (pad_temp.shared_1[152]*kernel.shared_1[((threadIdx.x*72) + 4658)]))
+        conv2d_nchw_1[0] = (conv2d_nchw_1[0] + (pad_temp.shared_1[153]*kernel.shared_1[((threadIdx.x*72) + 51)]))
+        conv2d_nchw_1[1] = (conv2d_nchw_1[1] + (pad_temp.shared_1[154]*kernel.shared_1[((threadIdx.x*72) + 51)]))
+        conv2d_nchw_1[2] = (conv2d_nchw_1[2] + (pad_temp.shared_1[155]*kernel.shared_1[((threadIdx.x*72) + 51)]))
+        conv2d_nchw_1[3] = (conv2d_nchw_1[3] + (pad_temp.shared_1[156]*kernel.shared_1[((threadIdx.x*72) + 51)]))
+        conv2d_nchw_1[4] = (conv2d_nchw_1[4] + (pad_temp.shared_1[157]*kernel.shared_1[((threadIdx.x*72) + 51)]))
+        conv2d_nchw_1[5] = (conv2d_nchw_1[5] + (pad_temp.shared_1[158]*kernel.shared_1[((threadIdx.x*72) + 51)]))
+        conv2d_nchw_1[6] = (conv2d_nchw_1[6] + (pad_temp.shared_1[159]*kernel.shared_1[((threadIdx.x*72) + 51)]))
+        conv2d_nchw_1[7] = (conv2d_nchw_1[7] + (pad_temp.shared_1[153]*kernel.shared_1[((threadIdx.x*72) + 4659)]))
+        conv2d_nchw_1[8] = (conv2d_nchw_1[8] + (pad_temp.shared_1[154]*kernel.shared_1[((threadIdx.x*72) + 4659)]))
+        conv2d_nchw_1[9] = (conv2d_nchw_1[9] + (pad_temp.shared_1[155]*kernel.shared_1[((threadIdx.x*72) + 4659)]))
+        conv2d_nchw_1[10] = (conv2d_nchw_1[10] + (pad_temp.shared_1[156]*kernel.shared_1[((threadIdx.x*72) + 4659)]))
+        conv2d_nchw_1[11] = (conv2d_nchw_1[11] + (pad_temp.shared_1[157]*kernel.shared_1[((threadIdx.x*72) + 4659)]))
+        conv2d_nchw_1[12] = (conv2d_nchw_1[12] + (pad_temp.shared_1[158]*kernel.shared_1[((threadIdx.x*72) + 4659)]))
+        conv2d_nchw_1[13] = (conv2d_nchw_1[13] + (pad_temp.shared_1[159]*kernel.shared_1[((threadIdx.x*72) + 4659)]))
+        conv2d_nchw_1[0] = (conv2d_nchw_1[0] + (pad_temp.shared_1[154]*kernel.shared_1[((threadIdx.x*72) + 52)]))
+        conv2d_nchw_1[1] = (conv2d_nchw_1[1] + (pad_temp.shared_1[155]*kernel.shared_1[((threadIdx.x*72) + 52)]))
+        conv2d_nchw_1[2] = (conv2d_nchw_1[2] + (pad_temp.shared_1[156]*kernel.shared_1[((threadIdx.x*72) + 52)]))
+        conv2d_nchw_1[3] = (conv2d_nchw_1[3] + (pad_temp.shared_1[157]*kernel.shared_1[((threadIdx.x*72) + 52)]))
+        conv2d_nchw_1[4] = (conv2d_nchw_1[4] + (pad_temp.shared_1[158]*kernel.shared_1[((threadIdx.x*72) + 52)]))
+        conv2d_nchw_1[5] = (conv2d_nchw_1[5] + (pad_temp.shared_1[159]*kernel.shared_1[((threadIdx.x*72) + 52)]))
+        conv2d_nchw_1[6] = (conv2d_nchw_1[6] + (pad_temp.shared_1[160]*kernel.shared_1[((threadIdx.x*72) + 52)]))
+        conv2d_nchw_1[7] = (conv2d_nchw_1[7] + (pad_temp.shared_1[154]*kernel.shared_1[((threadIdx.x*72) + 4660)]))
+        conv2d_nchw_1[8] = (conv2d_nchw_1[8] + (pad_temp.shared_1[155]*kernel.shared_1[((threadIdx.x*72) + 4660)]))
+        conv2d_nchw_1[9] = (conv2d_nchw_1[9] + (pad_temp.shared_1[156]*kernel.shared_1[((threadIdx.x*72) + 4660)]))
+        conv2d_nchw_1[10] = (conv2d_nchw_1[10] + (pad_temp.shared_1[157]*kernel.shared_1[((threadIdx.x*72) + 4660)]))
+        conv2d_nchw_1[11] = (conv2d_nchw_1[11] + (pad_temp.shared_1[158]*kernel.shared_1[((threadIdx.x*72) + 4660)]))
+        conv2d_nchw_1[12] = (conv2d_nchw_1[12] + (pad_temp.shared_1[159]*kernel.shared_1[((threadIdx.x*72) + 4660)]))
+        conv2d_nchw_1[13] = (conv2d_nchw_1[13] + (pad_temp.shared_1[160]*kernel.shared_1[((threadIdx.x*72) + 4660)]))
+        conv2d_nchw_1[0] = (conv2d_nchw_1[0] + (pad_temp.shared_1[155]*kernel.shared_1[((threadIdx.x*72) + 53)]))
+        conv2d_nchw_1[1] = (conv2d_nchw_1[1] + (pad_temp.shared_1[156]*kernel.shared_1[((threadIdx.x*72) + 53)]))
+        conv2d_nchw_1[2] = (conv2d_nchw_1[2] + (pad_temp.shared_1[157]*kernel.shared_1[((threadIdx.x*72) + 53)]))
+        conv2d_nchw_1[3] = (conv2d_nchw_1[3] + (pad_temp.shared_1[158]*kernel.shared_1[((threadIdx.x*72) + 53)]))
+        conv2d_nchw_1[4] = (conv2d_nchw_1[4] + (pad_temp.shared_1[159]*kernel.shared_1[((threadIdx.x*72) + 53)]))
+        conv2d_nchw_1[5] = (conv2d_nchw_1[5] + (pad_temp.shared_1[160]*kernel.shared_1[((threadIdx.x*72) + 53)]))
+        conv2d_nchw_1[6] = (conv2d_nchw_1[6] + (pad_temp.shared_1[161]*kernel.shared_1[((threadIdx.x*72) + 53)]))
+        conv2d_nchw_1[7] = (conv2d_nchw_1[7] + (pad_temp.shared_1[155]*kernel.shared_1[((threadIdx.x*72) + 4661)]))
+        conv2d_nchw_1[8] = (conv2d_nchw_1[8] + (pad_temp.shared_1[156]*kernel.shared_1[((threadIdx.x*72) + 4661)]))
+        conv2d_nchw_1[9] = (conv2d_nchw_1[9] + (pad_temp.shared_1[157]*kernel.shared_1[((threadIdx.x*72) + 4661)]))
+        conv2d_nchw_1[10] = (conv2d_nchw_1[10] + (pad_temp.shared_1[158]*kernel.shared_1[((threadIdx.x*72) + 4661)]))
+        conv2d_nchw_1[11] = (conv2d_nchw_1[11] + (pad_temp.shared_1[159]*kernel.shared_1[((threadIdx.x*72) + 4661)]))
+        conv2d_nchw_1[12] = (conv2d_nchw_1[12] + (pad_temp.shared_1[160]*kernel.shared_1[((threadIdx.x*72) + 4661)]))
+        conv2d_nchw_1[13] = (conv2d_nchw_1[13] + (pad_temp.shared_1[161]*kernel.shared_1[((threadIdx.x*72) + 4661)]))
+        conv2d_nchw_1[0] = (conv2d_nchw_1[0] + (pad_temp.shared_1[162]*kernel.shared_1[((threadIdx.x*72) + 54)]))
+        conv2d_nchw_1[1] = (conv2d_nchw_1[1] + (pad_temp.shared_1[163]*kernel.shared_1[((threadIdx.x*72) + 54)]))
+        conv2d_nchw_1[2] = (conv2d_nchw_1[2] + (pad_temp.shared_1[164]*kernel.shared_1[((threadIdx.x*72) + 54)]))
+        conv2d_nchw_1[3] = (conv2d_nchw_1[3] + (pad_temp.shared_1[165]*kernel.shared_1[((threadIdx.x*72) + 54)]))
+        conv2d_nchw_1[4] = (conv2d_nchw_1[4] + (pad_temp.shared_1[166]*kernel.shared_1[((threadIdx.x*72) + 54)]))
+        conv2d_nchw_1[5] = (conv2d_nchw_1[5] + (pad_temp.shared_1[167]*kernel.shared_1[((threadIdx.x*72) + 54)]))
+        conv2d_nchw_1[6] = (conv2d_nchw_1[6] + (pad_temp.shared_1[168]*kernel.shared_1[((threadIdx.x*72) + 54)]))
+        conv2d_nchw_1[7] = (conv2d_nchw_1[7] + (pad_temp.shared_1[162]*kernel.shared_1[((threadIdx.x*72) + 4662)]))
+        conv2d_nchw_1[8] = (conv2d_nchw_1[8] + (pad_temp.shared_1[163]*kernel.shared_1[((threadIdx.x*72) + 4662)]))
+        conv2d_nchw_1[9] = (conv2d_nchw_1[9] + (pad_temp.shared_1[164]*kernel.shared_1[((threadIdx.x*72) + 4662)]))
+        conv2d_nchw_1[10] = (conv2d_nchw_1[10] + (pad_temp.shared_1[165]*kernel.shared_1[((threadIdx.x*72) + 4662)]))
+        conv2d_nchw_1[11] = (conv2d_nchw_1[11] + (pad_temp.shared_1[166]*kernel.shared_1[((threadIdx.x*72) + 4662)]))
+        conv2d_nchw_1[12] = (conv2d_nchw_1[12] + (pad_temp.shared_1[167]*kernel.shared_1[((threadIdx.x*72) + 4662)]))
+        conv2d_nchw_1[13] = (conv2d_nchw_1[13] + (pad_temp.shared_1[168]*kernel.shared_1[((threadIdx.x*72) + 4662)]))
+        conv2d_nchw_1[0] = (conv2d_nchw_1[0] + (pad_temp.shared_1[163]*kernel.shared_1[((threadIdx.x*72) + 55)]))
+        conv2d_nchw_1[1] = (conv2d_nchw_1[1] + (pad_temp.shared_1[164]*kernel.shared_1[((threadIdx.x*72) + 55)]))
+        conv2d_nchw_1[2] = (conv2d_nchw_1[2] + (pad_temp.shared_1[165]*kernel.shared_1[((threadIdx.x*72) + 55)]))
+        conv2d_nchw_1[3] = (conv2d_nchw_1[3] + (pad_temp.shared_1[166]*kernel.shared_1[((threadIdx.x*72) + 55)]))
+        conv2d_nchw_1[4] = (conv2d_nchw_1[4] + (pad_temp.shared_1[167]*kernel.shared_1[((threadIdx.x*72) + 55)]))
+        conv2d_nchw_1[5] = (conv2d_nchw_1[5] + (pad_temp.shared_1[168]*kernel.shared_1[((threadIdx.x*72) + 55)]))
+        conv2d_nchw_1[6] = (conv2d_nchw_1[6] + (pad_temp.shared_1[169]*kernel.shared_1[((threadIdx.x*72) + 55)]))
+        conv2d_nchw_1[7] = (conv2d_nchw_1[7] + (pad_temp.shared_1[163]*kernel.shared_1[((threadIdx.x*72) + 4663)]))
+        conv2d_nchw_1[8] = (conv2d_nchw_1[8] + (pad_temp.shared_1[164]*kernel.shared_1[((threadIdx.x*72) + 4663)]))
+        conv2d_nchw_1[9] = (conv2d_nchw_1[9] + (pad_temp.shared_1[165]*kernel.shared_1[((threadIdx.x*72) + 4663)]))
+        conv2d_nchw_1[10] = (conv2d_nchw_1[10] + (pad_temp.shared_1[166]*kernel.shared_1[((threadIdx.x*72) + 4663)]))
+        conv2d_nchw_1[11] = (conv2d_nchw_1[11] + (pad_temp.shared_1[167]*kernel.shared_1[((threadIdx.x*72) + 4663)]))
+        conv2d_nchw_1[12] = (conv2d_nchw_1[12] + (pad_temp.shared_1[168]*kernel.shared_1[((threadIdx.x*72) + 4663)]))
+        conv2d_nchw_1[13] = (conv2d_nchw_1[13] + (pad_temp.shared_1[169]*kernel.shared_1[((threadIdx.x*72) + 4663)]))
+        conv2d_nchw_1[0] = (conv2d_nchw_1[0] + (pad_temp.shared_1[164]*kernel.shared_1[((threadIdx.x*72) + 56)]))
+        conv2d_nchw_1[1] = (conv2d_nchw_1[1] + (pad_temp.shared_1[165]*kernel.shared_1[((threadIdx.x*72) + 56)]))
+        conv2d_nchw_1[2] = (conv2d_nchw_1[2] + (pad_temp.shared_1[166]*kernel.shared_1[((threadIdx.x*72) + 56)]))
+        conv2d_nchw_1[3] = (conv2d_nchw_1[3] + (pad_temp.shared_1[167]*kernel.shared_1[((threadIdx.x*72) + 56)]))
+        conv2d_nchw_1[4] = (conv2d_nchw_1[4] + (pad_temp.shared_1[168]*kernel.shared_1[((threadIdx.x*72) + 56)]))
+        conv2d_nchw_1[5] = (conv2d_nchw_1[5] + (pad_temp.shared_1[169]*kernel.shared_1[((threadIdx.x*72) + 56)]))
+        conv2d_nchw_1[6] = (conv2d_nchw_1[6] + (pad_temp.shared_1[170]*kernel.shared_1[((threadIdx.x*72) + 56)]))
+        conv2d_nchw_1[7] = (conv2d_nchw_1[7] + (pad_temp.shared_1[164]*kernel.shared_1[((threadIdx.x*72) + 4664)]))
+        conv2d_nchw_1[8] = (conv2d_nchw_1[8] + (pad_temp.shared_1[165]*kernel.shared_1[((threadIdx.x*72) + 4664)]))
+        conv2d_nchw_1[9] = (conv2d_nchw_1[9] + (pad_temp.shared_1[166]*kernel.shared_1[((threadIdx.x*72) + 4664)]))
+        conv2d_nchw_1[10] = (conv2d_nchw_1[10] + (pad_temp.shared_1[167]*kernel.shared_1[((threadIdx.x*72) + 4664)]))
+        conv2d_nchw_1[11] = (conv2d_nchw_1[11] + (pad_temp.shared_1[168]*kernel.shared_1[((threadIdx.x*72) + 4664)]))
+        conv2d_nchw_1[12] = (conv2d_nchw_1[12] + (pad_temp.shared_1[169]*kernel.shared_1[((threadIdx.x*72) + 4664)]))
+        conv2d_nchw_1[13] = (conv2d_nchw_1[13] + (pad_temp.shared_1[170]*kernel.shared_1[((threadIdx.x*72) + 4664)]))
+        conv2d_nchw_1[0] = (conv2d_nchw_1[0] + (pad_temp.shared_1[171]*kernel.shared_1[((threadIdx.x*72) + 57)]))
+        conv2d_nchw_1[1] = (conv2d_nchw_1[1] + (pad_temp.shared_1[172]*kernel.shared_1[((threadIdx.x*72) + 57)]))
+        conv2d_nchw_1[2] = (conv2d_nchw_1[2] + (pad_temp.shared_1[173]*kernel.shared_1[((threadIdx.x*72) + 57)]))
+        conv2d_nchw_1[3] = (conv2d_nchw_1[3] + (pad_temp.shared_1[174]*kernel.shared_1[((threadIdx.x*72) + 57)]))
+        conv2d_nchw_1[4] = (conv2d_nchw_1[4] + (pad_temp.shared_1[175]*kernel.shared_1[((threadIdx.x*72) + 57)]))
+        conv2d_nchw_1[5] = (conv2d_nchw_1[5] + (pad_temp.shared_1[176]*kernel.shared_1[((threadIdx.x*72) + 57)]))
+        conv2d_nchw_1[6] = (conv2d_nchw_1[6] + (pad_temp.shared_1[177]*kernel.shared_1[((threadIdx.x*72) + 57)]))
+        conv2d_nchw_1[7] = (conv2d_nchw_1[7] + (pad_temp.shared_1[171]*kernel.shared_1[((threadIdx.x*72) + 4665)]))
+        conv2d_nchw_1[8] = (conv2d_nchw_1[8] + (pad_temp.shared_1[172]*kernel.shared_1[((threadIdx.x*72) + 4665)]))
+        conv2d_nchw_1[9] = (conv2d_nchw_1[9] + (pad_temp.shared_1[173]*kernel.shared_1[((threadIdx.x*72) + 4665)]))
+        conv2d_nchw_1[10] = (conv2d_nchw_1[10] + (pad_temp.shared_1[174]*kernel.shared_1[((threadIdx.x*72) + 4665)]))
+        conv2d_nchw_1[11] = (conv2d_nchw_1[11] + (pad_temp.shared_1[175]*kernel.shared_1[((threadIdx.x*72) + 4665)]))
+        conv2d_nchw_1[12] = (conv2d_nchw_1[12] + (pad_temp.shared_1[176]*kernel.shared_1[((threadIdx.x*72) + 4665)]))
+        conv2d_nchw_1[13] = (conv2d_nchw_1[13] + (pad_temp.shared_1[177]*kernel.shared_1[((threadIdx.x*72) + 4665)]))
+        conv2d_nchw_1[0] = (conv2d_nchw_1[0] + (pad_temp.shared_1[172]*kernel.shared_1[((threadIdx.x*72) + 58)]))
+        conv2d_nchw_1[1] = (conv2d_nchw_1[1] + (pad_temp.shared_1[173]*kernel.shared_1[((threadIdx.x*72) + 58)]))
+        conv2d_nchw_1[2] = (conv2d_nchw_1[2] + (pad_temp.shared_1[174]*kernel.shared_1[((threadIdx.x*72) + 58)]))
+        conv2d_nchw_1[3] = (conv2d_nchw_1[3] + (pad_temp.shared_1[175]*kernel.shared_1[((threadIdx.x*72) + 58)]))
+        conv2d_nchw_1[4] = (conv2d_nchw_1[4] + (pad_temp.shared_1[176]*kernel.shared_1[((threadIdx.x*72) + 58)]))
+        conv2d_nchw_1[5] = (conv2d_nchw_1[5] + (pad_temp.shared_1[177]*kernel.shared_1[((threadIdx.x*72) + 58)]))
+        conv2d_nchw_1[6] = (conv2d_nchw_1[6] + (pad_temp.shared_1[178]*kernel.shared_1[((threadIdx.x*72) + 58)]))
+        conv2d_nchw_1[7] = (conv2d_nchw_1[7] + (pad_temp.shared_1[172]*kernel.shared_1[((threadIdx.x*72) + 4666)]))
+        conv2d_nchw_1[8] = (conv2d_nchw_1[8] + (pad_temp.shared_1[173]*kernel.shared_1[((threadIdx.x*72) + 4666)]))
+        conv2d_nchw_1[9] = (conv2d_nchw_1[9] + (pad_temp.shared_1[174]*kernel.shared_1[((threadIdx.x*72) + 4666)]))
+        conv2d_nchw_1[10] = (conv2d_nchw_1[10] + (pad_temp.shared_1[175]*kernel.shared_1[((threadIdx.x*72) + 4666)]))
+        conv2d_nchw_1[11] = (conv2d_nchw_1[11] + (pad_temp.shared_1[176]*kernel.shared_1[((threadIdx.x*72) + 4666)]))
+        conv2d_nchw_1[12] = (conv2d_nchw_1[12] + (pad_temp.shared_1[177]*kernel.shared_1[((threadIdx.x*72) + 4666)]))
+        conv2d_nchw_1[13] = (conv2d_nchw_1[13] + (pad_temp.shared_1[178]*kernel.shared_1[((threadIdx.x*72) + 4666)]))
+        conv2d_nchw_1[0] = (conv2d_nchw_1[0] + (pad_temp.shared_1[173]*kernel.shared_1[((threadIdx.x*72) + 59)]))
+        conv2d_nchw_1[1] = (conv2d_nchw_1[1] + (pad_temp.shared_1[174]*kernel.shared_1[((threadIdx.x*72) + 59)]))
+        conv2d_nchw_1[2] = (conv2d_nchw_1[2] + (pad_temp.shared_1[175]*kernel.shared_1[((threadIdx.x*72) + 59)]))
+        conv2d_nchw_1[3] = (conv2d_nchw_1[3] + (pad_temp.shared_1[176]*kernel.shared_1[((threadIdx.x*72) + 59)]))
+        conv2d_nchw_1[4] = (conv2d_nchw_1[4] + (pad_temp.shared_1[177]*kernel.shared_1[((threadIdx.x*72) + 59)]))
+        conv2d_nchw_1[5] = (conv2d_nchw_1[5] + (pad_temp.shared_1[178]*kernel.shared_1[((threadIdx.x*72) + 59)]))
+        conv2d_nchw_1[6] = (conv2d_nchw_1[6] + (pad_temp.shared_1[179]*kernel.shared_1[((threadIdx.x*72) + 59)]))
+        conv2d_nchw_1[7] = (conv2d_nchw_1[7] + (pad_temp.shared_1[173]*kernel.shared_1[((threadIdx.x*72) + 4667)]))
+        conv2d_nchw_1[8] = (conv2d_nchw_1[8] + (pad_temp.shared_1[174]*kernel.shared_1[((threadIdx.x*72) + 4667)]))
+        conv2d_nchw_1[9] = (conv2d_nchw_1[9] + (pad_temp.shared_1[175]*kernel.shared_1[((threadIdx.x*72) + 4667)]))
+        conv2d_nchw_1[10] = (conv2d_nchw_1[10] + (pad_temp.shared_1[176]*kernel.shared_1[((threadIdx.x*72) + 4667)]))
+        conv2d_nchw_1[11] = (conv2d_nchw_1[11] + (pad_temp.shared_1[177]*kernel.shared_1[((threadIdx.x*72) + 4667)]))
+        conv2d_nchw_1[12] = (conv2d_nchw_1[12] + (pad_temp.shared_1[178]*kernel.shared_1[((threadIdx.x*72) + 4667)]))
+        conv2d_nchw_1[13] = (conv2d_nchw_1[13] + (pad_temp.shared_1[179]*kernel.shared_1[((threadIdx.x*72) + 4667)]))
+        conv2d_nchw_1[0] = (conv2d_nchw_1[0] + (pad_temp.shared_1[180]*kernel.shared_1[((threadIdx.x*72) + 60)]))
+        conv2d_nchw_1[1] = (conv2d_nchw_1[1] + (pad_temp.shared_1[181]*kernel.shared_1[((threadIdx.x*72) + 60)]))
+        conv2d_nchw_1[2] = (conv2d_nchw_1[2] + (pad_temp.shared_1[182]*kernel.shared_1[((threadIdx.x*72) + 60)]))
+        conv2d_nchw_1[3] = (conv2d_nchw_1[3] + (pad_temp.shared_1[183]*kernel.shared_1[((threadIdx.x*72) + 60)]))
+        conv2d_nchw_1[4] = (conv2d_nchw_1[4] + (pad_temp.shared_1[184]*kernel.shared_1[((threadIdx.x*72) + 60)]))
+        conv2d_nchw_1[5] = (conv2d_nchw_1[5] + (pad_temp.shared_1[185]*kernel.shared_1[((threadIdx.x*72) + 60)]))
+        conv2d_nchw_1[6] = (conv2d_nchw_1[6] + (pad_temp.shared_1[186]*kernel.shared_1[((threadIdx.x*72) + 60)]))
+        conv2d_nchw_1[7] = (conv2d_nchw_1[7] + (pad_temp.shared_1[180]*kernel.shared_1[((threadIdx.x*72) + 4668)]))
+        conv2d_nchw_1[8] = (conv2d_nchw_1[8] + (pad_temp.shared_1[181]*kernel.shared_1[((threadIdx.x*72) + 4668)]))
+        conv2d_nchw_1[9] = (conv2d_nchw_1[9] + (pad_temp.shared_1[182]*kernel.shared_1[((threadIdx.x*72) + 4668)]))
+        conv2d_nchw_1[10] = (conv2d_nchw_1[10] + (pad_temp.shared_1[183]*kernel.shared_1[((threadIdx.x*72) + 4668)]))
+        conv2d_nchw_1[11] = (conv2d_nchw_1[11] + (pad_temp.shared_1[184]*kernel.shared_1[((threadIdx.x*72) + 4668)]))
+        conv2d_nchw_1[12] = (conv2d_nchw_1[12] + (pad_temp.shared_1[185]*kernel.shared_1[((threadIdx.x*72) + 4668)]))
+        conv2d_nchw_1[13] = (conv2d_nchw_1[13] + (pad_temp.shared_1[186]*kernel.shared_1[((threadIdx.x*72) + 4668)]))
+        conv2d_nchw_1[0] = (conv2d_nchw_1[0] + (pad_temp.shared_1[181]*kernel.shared_1[((threadIdx.x*72) + 61)]))
+        conv2d_nchw_1[1] = (conv2d_nchw_1[1] + (pad_temp.shared_1[182]*kernel.shared_1[((threadIdx.x*72) + 61)]))
+        conv2d_nchw_1[2] = (conv2d_nchw_1[2] + (pad_temp.shared_1[183]*kernel.shared_1[((threadIdx.x*72) + 61)]))
+        conv2d_nchw_1[3] = (conv2d_nchw_1[3] + (pad_temp.shared_1[184]*kernel.shared_1[((threadIdx.x*72) + 61)]))
+        conv2d_nchw_1[4] = (conv2d_nchw_1[4] + (pad_temp.shared_1[185]*kernel.shared_1[((threadIdx.x*72) + 61)]))
+        conv2d_nchw_1[5] = (conv2d_nchw_1[5] + (pad_temp.shared_1[186]*kernel.shared_1[((threadIdx.x*72) + 61)]))
+        conv2d_nchw_1[6] = (conv2d_nchw_1[6] + (pad_temp.shared_1[187]*kernel.shared_1[((threadIdx.x*72) + 61)]))
+        conv2d_nchw_1[7] = (conv2d_nchw_1[7] + (pad_temp.shared_1[181]*kernel.shared_1[((threadIdx.x*72) + 4669)]))
+        conv2d_nchw_1[8] = (conv2d_nchw_1[8] + (pad_temp.shared_1[182]*kernel.shared_1[((threadIdx.x*72) + 4669)]))
+        conv2d_nchw_1[9] = (conv2d_nchw_1[9] + (pad_temp.shared_1[183]*kernel.shared_1[((threadIdx.x*72) + 4669)]))
+        conv2d_nchw_1[10] = (conv2d_nchw_1[10] + (pad_temp.shared_1[184]*kernel.shared_1[((threadIdx.x*72) + 4669)]))
+        conv2d_nchw_1[11] = (conv2d_nchw_1[11] + (pad_temp.shared_1[185]*kernel.shared_1[((threadIdx.x*72) + 4669)]))
+        conv2d_nchw_1[12] = (conv2d_nchw_1[12] + (pad_temp.shared_1[186]*kernel.shared_1[((threadIdx.x*72) + 4669)]))
+        conv2d_nchw_1[13] = (conv2d_nchw_1[13] + (pad_temp.shared_1[187]*kernel.shared_1[((threadIdx.x*72) + 4669)]))
+        conv2d_nchw_1[0] = (conv2d_nchw_1[0] + (pad_temp.shared_1[182]*kernel.shared_1[((threadIdx.x*72) + 62)]))
+        conv2d_nchw_1[1] = (conv2d_nchw_1[1] + (pad_temp.shared_1[183]*kernel.shared_1[((threadIdx.x*72) + 62)]))
+        conv2d_nchw_1[2] = (conv2d_nchw_1[2] + (pad_temp.shared_1[184]*kernel.shared_1[((threadIdx.x*72) + 62)]))
+        conv2d_nchw_1[3] = (conv2d_nchw_1[3] + (pad_temp.shared_1[185]*kernel.shared_1[((threadIdx.x*72) + 62)]))
+        conv2d_nchw_1[4] = (conv2d_nchw_1[4] + (pad_temp.shared_1[186]*kernel.shared_1[((threadIdx.x*72) + 62)]))
+        conv2d_nchw_1[5] = (conv2d_nchw_1[5] + (pad_temp.shared_1[187]*kernel.shared_1[((threadIdx.x*72) + 62)]))
+        conv2d_nchw_1[6] = (conv2d_nchw_1[6] + (pad_temp.shared_1[188]*kernel.shared_1[((threadIdx.x*72) + 62)]))
+        conv2d_nchw_1[7] = (conv2d_nchw_1[7] + (pad_temp.shared_1[182]*kernel.shared_1[((threadIdx.x*72) + 4670)]))
+        conv2d_nchw_1[8] = (conv2d_nchw_1[8] + (pad_temp.shared_1[183]*kernel.shared_1[((threadIdx.x*72) + 4670)]))
+        conv2d_nchw_1[9] = (conv2d_nchw_1[9] + (pad_temp.shared_1[184]*kernel.shared_1[((threadIdx.x*72) + 4670)]))
+        conv2d_nchw_1[10] = (conv2d_nchw_1[10] + (pad_temp.shared_1[185]*kernel.shared_1[((threadIdx.x*72) + 4670)]))
+        conv2d_nchw_1[11] = (conv2d_nchw_1[11] + (pad_temp.shared_1[186]*kernel.shared_1[((threadIdx.x*72) + 4670)]))
+        conv2d_nchw_1[12] = (conv2d_nchw_1[12] + (pad_temp.shared_1[187]*kernel.shared_1[((threadIdx.x*72) + 4670)]))
+        conv2d_nchw_1[13] = (conv2d_nchw_1[13] + (pad_temp.shared_1[188]*kernel.shared_1[((threadIdx.x*72) + 4670)]))
+        conv2d_nchw_1[0] = (conv2d_nchw_1[0] + (pad_temp.shared_1[189]*kernel.shared_1[((threadIdx.x*72) + 63)]))
+        conv2d_nchw_1[1] = (conv2d_nchw_1[1] + (pad_temp.shared_1[190]*kernel.shared_1[((threadIdx.x*72) + 63)]))
+        conv2d_nchw_1[2] = (conv2d_nchw_1[2] + (pad_temp.shared_1[191]*kernel.shared_1[((threadIdx.x*72) + 63)]))
+        conv2d_nchw_1[3] = (conv2d_nchw_1[3] + (pad_temp.shared_1[192]*kernel.shared_1[((threadIdx.x*72) + 63)]))
+        conv2d_nchw_1[4] = (conv2d_nchw_1[4] + (pad_temp.shared_1[193]*kernel.shared_1[((threadIdx.x*72) + 63)]))
+        conv2d_nchw_1[5] = (conv2d_nchw_1[5] + (pad_temp.shared_1[194]*kernel.shared_1[((threadIdx.x*72) + 63)]))
+        conv2d_nchw_1[6] = (conv2d_nchw_1[6] + (pad_temp.shared_1[195]*kernel.shared_1[((threadIdx.x*72) + 63)]))
+        conv2d_nchw_1[7] = (conv2d_nchw_1[7] + (pad_temp.shared_1[189]*kernel.shared_1[((threadIdx.x*72) + 4671)]))
+        conv2d_nchw_1[8] = (conv2d_nchw_1[8] + (pad_temp.shared_1[190]*kernel.shared_1[((threadIdx.x*72) + 4671)]))
+        conv2d_nchw_1[9] = (conv2d_nchw_1[9] + (pad_temp.shared_1[191]*kernel.shared_1[((threadIdx.x*72) + 4671)]))
+        conv2d_nchw_1[10] = (conv2d_nchw_1[10] + (pad_temp.shared_1[192]*kernel.shared_1[((threadIdx.x*72) + 4671)]))
+        conv2d_nchw_1[11] = (conv2d_nchw_1[11] + (pad_temp.shared_1[193]*kernel.shared_1[((threadIdx.x*72) + 4671)]))
+        conv2d_nchw_1[12] = (conv2d_nchw_1[12] + (pad_temp.shared_1[194]*kernel.shared_1[((threadIdx.x*72) + 4671)]))
+        conv2d_nchw_1[13] = (conv2d_nchw_1[13] + (pad_temp.shared_1[195]*kernel.shared_1[((threadIdx.x*72) + 4671)]))
+        conv2d_nchw_1[0] = (conv2d_nchw_1[0] + (pad_temp.shared_1[190]*kernel.shared_1[((threadIdx.x*72) + 64)]))
+        conv2d_nchw_1[1] = (conv2d_nchw_1[1] + (pad_temp.shared_1[191]*kernel.shared_1[((threadIdx.x*72) + 64)]))
+        conv2d_nchw_1[2] = (conv2d_nchw_1[2] + (pad_temp.shared_1[192]*kernel.shared_1[((threadIdx.x*72) + 64)]))
+        conv2d_nchw_1[3] = (conv2d_nchw_1[3] + (pad_temp.shared_1[193]*kernel.shared_1[((threadIdx.x*72) + 64)]))
+        conv2d_nchw_1[4] = (conv2d_nchw_1[4] + (pad_temp.shared_1[194]*kernel.shared_1[((threadIdx.x*72) + 64)]))
+        conv2d_nchw_1[5] = (conv2d_nchw_1[5] + (pad_temp.shared_1[195]*kernel.shared_1[((threadIdx.x*72) + 64)]))
+        conv2d_nchw_1[6] = (conv2d_nchw_1[6] + (pad_temp.shared_1[196]*kernel.shared_1[((threadIdx.x*72) + 64)]))
+        conv2d_nchw_1[7] = (conv2d_nchw_1[7] + (pad_temp.shared_1[190]*kernel.shared_1[((threadIdx.x*72) + 4672)]))
+        conv2d_nchw_1[8] = (conv2d_nchw_1[8] + (pad_temp.shared_1[191]*kernel.shared_1[((threadIdx.x*72) + 4672)]))
+        conv2d_nchw_1[9] = (conv2d_nchw_1[9] + (pad_temp.shared_1[192]*kernel.shared_1[((threadIdx.x*72) + 4672)]))
+        conv2d_nchw_1[10] = (conv2d_nchw_1[10] + (pad_temp.shared_1[193]*kernel.shared_1[((threadIdx.x*72) + 4672)]))
+        conv2d_nchw_1[11] = (conv2d_nchw_1[11] + (pad_temp.shared_1[194]*kernel.shared_1[((threadIdx.x*72) + 4672)]))
+        conv2d_nchw_1[12] = (conv2d_nchw_1[12] + (pad_temp.shared_1[195]*kernel.shared_1[((threadIdx.x*72) + 4672)]))
+        conv2d_nchw_1[13] = (conv2d_nchw_1[13] + (pad_temp.shared_1[196]*kernel.shared_1[((threadIdx.x*72) + 4672)]))
+        conv2d_nchw_1[0] = (conv2d_nchw_1[0] + (pad_temp.shared_1[191]*kernel.shared_1[((threadIdx.x*72) + 65)]))
+        conv2d_nchw_1[1] = (conv2d_nchw_1[1] + (pad_temp.shared_1[192]*kernel.shared_1[((threadIdx.x*72) + 65)]))
+        conv2d_nchw_1[2] = (conv2d_nchw_1[2] + (pad_temp.shared_1[193]*kernel.shared_1[((threadIdx.x*72) + 65)]))
+        conv2d_nchw_1[3] = (conv2d_nchw_1[3] + (pad_temp.shared_1[194]*kernel.shared_1[((threadIdx.x*72) + 65)]))
+        conv2d_nchw_1[4] = (conv2d_nchw_1[4] + (pad_temp.shared_1[195]*kernel.shared_1[((threadIdx.x*72) + 65)]))
+        conv2d_nchw_1[5] = (conv2d_nchw_1[5] + (pad_temp.shared_1[196]*kernel.shared_1[((threadIdx.x*72) + 65)]))
+        conv2d_nchw_1[6] = (conv2d_nchw_1[6] + (pad_temp.shared_1[197]*kernel.shared_1[((threadIdx.x*72) + 65)]))
+        conv2d_nchw_1[7] = (conv2d_nchw_1[7] + (pad_temp.shared_1[191]*kernel.shared_1[((threadIdx.x*72) + 4673)]))
+        conv2d_nchw_1[8] = (conv2d_nchw_1[8] + (pad_temp.shared_1[192]*kernel.shared_1[((threadIdx.x*72) + 4673)]))
+        conv2d_nchw_1[9] = (conv2d_nchw_1[9] + (pad_temp.shared_1[193]*kernel.shared_1[((threadIdx.x*72) + 4673)]))
+        conv2d_nchw_1[10] = (conv2d_nchw_1[10] + (pad_temp.shared_1[194]*kernel.shared_1[((threadIdx.x*72) + 4673)]))
+        conv2d_nchw_1[11] = (conv2d_nchw_1[11] + (pad_temp.shared_1[195]*kernel.shared_1[((threadIdx.x*72) + 4673)]))
+        conv2d_nchw_1[12] = (conv2d_nchw_1[12] + (pad_temp.shared_1[196]*kernel.shared_1[((threadIdx.x*72) + 4673)]))
+        conv2d_nchw_1[13] = (conv2d_nchw_1[13] + (pad_temp.shared_1[197]*kernel.shared_1[((threadIdx.x*72) + 4673)]))
+        conv2d_nchw_1[0] = (conv2d_nchw_1[0] + (pad_temp.shared_1[198]*kernel.shared_1[((threadIdx.x*72) + 66)]))
+        conv2d_nchw_1[1] = (conv2d_nchw_1[1] + (pad_temp.shared_1[199]*kernel.shared_1[((threadIdx.x*72) + 66)]))
+        conv2d_nchw_1[2] = (conv2d_nchw_1[2] + (pad_temp.shared_1[200]*kernel.shared_1[((threadIdx.x*72) + 66)]))
+        conv2d_nchw_1[3] = (conv2d_nchw_1[3] + (pad_temp.shared_1[201]*kernel.shared_1[((threadIdx.x*72) + 66)]))
+        conv2d_nchw_1[4] = (conv2d_nchw_1[4] + (pad_temp.shared_1[202]*kernel.shared_1[((threadIdx.x*72) + 66)]))
+        conv2d_nchw_1[5] = (conv2d_nchw_1[5] + (pad_temp.shared_1[203]*kernel.shared_1[((threadIdx.x*72) + 66)]))
+        conv2d_nchw_1[6] = (conv2d_nchw_1[6] + (pad_temp.shared_1[204]*kernel.shared_1[((threadIdx.x*72) + 66)]))
+        conv2d_nchw_1[7] = (conv2d_nchw_1[7] + (pad_temp.shared_1[198]*kernel.shared_1[((threadIdx.x*72) + 4674)]))
+        conv2d_nchw_1[8] = (conv2d_nchw_1[8] + (pad_temp.shared_1[199]*kernel.shared_1[((threadIdx.x*72) + 4674)]))
+        conv2d_nchw_1[9] = (conv2d_nchw_1[9] + (pad_temp.shared_1[200]*kernel.shared_1[((threadIdx.x*72) + 4674)]))
+        conv2d_nchw_1[10] = (conv2d_nchw_1[10] + (pad_temp.shared_1[201]*kernel.shared_1[((threadIdx.x*72) + 4674)]))
+        conv2d_nchw_1[11] = (conv2d_nchw_1[11] + (pad_temp.shared_1[202]*kernel.shared_1[((threadIdx.x*72) + 4674)]))
+        conv2d_nchw_1[12] = (conv2d_nchw_1[12] + (pad_temp.shared_1[203]*kernel.shared_1[((threadIdx.x*72) + 4674)]))
+        conv2d_nchw_1[13] = (conv2d_nchw_1[13] + (pad_temp.shared_1[204]*kernel.shared_1[((threadIdx.x*72) + 4674)]))
+        conv2d_nchw_1[0] = (conv2d_nchw_1[0] + (pad_temp.shared_1[199]*kernel.shared_1[((threadIdx.x*72) + 67)]))
+        conv2d_nchw_1[1] = (conv2d_nchw_1[1] + (pad_temp.shared_1[200]*kernel.shared_1[((threadIdx.x*72) + 67)]))
+        conv2d_nchw_1[2] = (conv2d_nchw_1[2] + (pad_temp.shared_1[201]*kernel.shared_1[((threadIdx.x*72) + 67)]))
+        conv2d_nchw_1[3] = (conv2d_nchw_1[3] + (pad_temp.shared_1[202]*kernel.shared_1[((threadIdx.x*72) + 67)]))
+        conv2d_nchw_1[4] = (conv2d_nchw_1[4] + (pad_temp.shared_1[203]*kernel.shared_1[((threadIdx.x*72) + 67)]))
+        conv2d_nchw_1[5] = (conv2d_nchw_1[5] + (pad_temp.shared_1[204]*kernel.shared_1[((threadIdx.x*72) + 67)]))
+        conv2d_nchw_1[6] = (conv2d_nchw_1[6] + (pad_temp.shared_1[205]*kernel.shared_1[((threadIdx.x*72) + 67)]))
+        conv2d_nchw_1[7] = (conv2d_nchw_1[7] + (pad_temp.shared_1[199]*kernel.shared_1[((threadIdx.x*72) + 4675)]))
+        conv2d_nchw_1[8] = (conv2d_nchw_1[8] + (pad_temp.shared_1[200]*kernel.shared_1[((threadIdx.x*72) + 4675)]))
+        conv2d_nchw_1[9] = (conv2d_nchw_1[9] + (pad_temp.shared_1[201]*kernel.shared_1[((threadIdx.x*72) + 4675)]))
+        conv2d_nchw_1[10] = (conv2d_nchw_1[10] + (pad_temp.shared_1[202]*kernel.shared_1[((threadIdx.x*72) + 4675)]))
+        conv2d_nchw_1[11] = (conv2d_nchw_1[11] + (pad_temp.shared_1[203]*kernel.shared_1[((threadIdx.x*72) + 4675)]))
+        conv2d_nchw_1[12] = (conv2d_nchw_1[12] + (pad_temp.shared_1[204]*kernel.shared_1[((threadIdx.x*72) + 4675)]))
+        conv2d_nchw_1[13] = (conv2d_nchw_1[13] + (pad_temp.shared_1[205]*kernel.shared_1[((threadIdx.x*72) + 4675)]))
+        conv2d_nchw_1[0] = (conv2d_nchw_1[0] + (pad_temp.shared_1[200]*kernel.shared_1[((threadIdx.x*72) + 68)]))
+        conv2d_nchw_1[1] = (conv2d_nchw_1[1] + (pad_temp.shared_1[201]*kernel.shared_1[((threadIdx.x*72) + 68)]))
+        conv2d_nchw_1[2] = (conv2d_nchw_1[2] + (pad_temp.shared_1[202]*kernel.shared_1[((threadIdx.x*72) + 68)]))
+        conv2d_nchw_1[3] = (conv2d_nchw_1[3] + (pad_temp.shared_1[203]*kernel.shared_1[((threadIdx.x*72) + 68)]))
+        conv2d_nchw_1[4] = (conv2d_nchw_1[4] + (pad_temp.shared_1[204]*kernel.shared_1[((threadIdx.x*72) + 68)]))
+        conv2d_nchw_1[5] = (conv2d_nchw_1[5] + (pad_temp.shared_1[205]*kernel.shared_1[((threadIdx.x*72) + 68)]))
+        conv2d_nchw_1[6] = (conv2d_nchw_1[6] + (pad_temp.shared_1[206]*kernel.shared_1[((threadIdx.x*72) + 68)]))
+        conv2d_nchw_1[7] = (conv2d_nchw_1[7] + (pad_temp.shared_1[200]*kernel.shared_1[((threadIdx.x*72) + 4676)]))
+        conv2d_nchw_1[8] = (conv2d_nchw_1[8] + (pad_temp.shared_1[201]*kernel.shared_1[((threadIdx.x*72) + 4676)]))
+        conv2d_nchw_1[9] = (conv2d_nchw_1[9] + (pad_temp.shared_1[202]*kernel.shared_1[((threadIdx.x*72) + 4676)]))
+        conv2d_nchw_1[10] = (conv2d_nchw_1[10] + (pad_temp.shared_1[203]*kernel.shared_1[((threadIdx.x*72) + 4676)]))
+        conv2d_nchw_1[11] = (conv2d_nchw_1[11] + (pad_temp.shared_1[204]*kernel.shared_1[((threadIdx.x*72) + 4676)]))
+        conv2d_nchw_1[12] = (conv2d_nchw_1[12] + (pad_temp.shared_1[205]*kernel.shared_1[((threadIdx.x*72) + 4676)]))
+        conv2d_nchw_1[13] = (conv2d_nchw_1[13] + (pad_temp.shared_1[206]*kernel.shared_1[((threadIdx.x*72) + 4676)]))
+        conv2d_nchw_1[0] = (conv2d_nchw_1[0] + (pad_temp.shared_1[207]*kernel.shared_1[((threadIdx.x*72) + 69)]))
+        conv2d_nchw_1[1] = (conv2d_nchw_1[1] + (pad_temp.shared_1[208]*kernel.shared_1[((threadIdx.x*72) + 69)]))
+        conv2d_nchw_1[2] = (conv2d_nchw_1[2] + (pad_temp.shared_1[209]*kernel.shared_1[((threadIdx.x*72) + 69)]))
+        conv2d_nchw_1[3] = (conv2d_nchw_1[3] + (pad_temp.shared_1[210]*kernel.shared_1[((threadIdx.x*72) + 69)]))
+        conv2d_nchw_1[4] = (conv2d_nchw_1[4] + (pad_temp.shared_1[211]*kernel.shared_1[((threadIdx.x*72) + 69)]))
+        conv2d_nchw_1[5] = (conv2d_nchw_1[5] + (pad_temp.shared_1[212]*kernel.shared_1[((threadIdx.x*72) + 69)]))
+        conv2d_nchw_1[6] = (conv2d_nchw_1[6] + (pad_temp.shared_1[213]*kernel.shared_1[((threadIdx.x*72) + 69)]))
+        conv2d_nchw_1[7] = (conv2d_nchw_1[7] + (pad_temp.shared_1[207]*kernel.shared_1[((threadIdx.x*72) + 4677)]))
+        conv2d_nchw_1[8] = (conv2d_nchw_1[8] + (pad_temp.shared_1[208]*kernel.shared_1[((threadIdx.x*72) + 4677)]))
+        conv2d_nchw_1[9] = (conv2d_nchw_1[9] + (pad_temp.shared_1[209]*kernel.shared_1[((threadIdx.x*72) + 4677)]))
+        conv2d_nchw_1[10] = (conv2d_nchw_1[10] + (pad_temp.shared_1[210]*kernel.shared_1[((threadIdx.x*72) + 4677)]))
+        conv2d_nchw_1[11] = (conv2d_nchw_1[11] + (pad_temp.shared_1[211]*kernel.shared_1[((threadIdx.x*72) + 4677)]))
+        conv2d_nchw_1[12] = (conv2d_nchw_1[12] + (pad_temp.shared_1[212]*kernel.shared_1[((threadIdx.x*72) + 4677)]))
+        conv2d_nchw_1[13] = (conv2d_nchw_1[13] + (pad_temp.shared_1[213]*kernel.shared_1[((threadIdx.x*72) + 4677)]))
+        conv2d_nchw_1[0] = (conv2d_nchw_1[0] + (pad_temp.shared_1[208]*kernel.shared_1[((threadIdx.x*72) + 70)]))
+        conv2d_nchw_1[1] = (conv2d_nchw_1[1] + (pad_temp.shared_1[209]*kernel.shared_1[((threadIdx.x*72) + 70)]))
+        conv2d_nchw_1[2] = (conv2d_nchw_1[2] + (pad_temp.shared_1[210]*kernel.shared_1[((threadIdx.x*72) + 70)]))
+        conv2d_nchw_1[3] = (conv2d_nchw_1[3] + (pad_temp.shared_1[211]*kernel.shared_1[((threadIdx.x*72) + 70)]))
+        conv2d_nchw_1[4] = (conv2d_nchw_1[4] + (pad_temp.shared_1[212]*kernel.shared_1[((threadIdx.x*72) + 70)]))
+        conv2d_nchw_1[5] = (conv2d_nchw_1[5] + (pad_temp.shared_1[213]*kernel.shared_1[((threadIdx.x*72) + 70)]))
+        conv2d_nchw_1[6] = (conv2d_nchw_1[6] + (pad_temp.shared_1[214]*kernel.shared_1[((threadIdx.x*72) + 70)]))
+        conv2d_nchw_1[7] = (conv2d_nchw_1[7] + (pad_temp.shared_1[208]*kernel.shared_1[((threadIdx.x*72) + 4678)]))
+        conv2d_nchw_1[8] = (conv2d_nchw_1[8] + (pad_temp.shared_1[209]*kernel.shared_1[((threadIdx.x*72) + 4678)]))
+        conv2d_nchw_1[9] = (conv2d_nchw_1[9] + (pad_temp.shared_1[210]*kernel.shared_1[((threadIdx.x*72) + 4678)]))
+        conv2d_nchw_1[10] = (conv2d_nchw_1[10] + (pad_temp.shared_1[211]*kernel.shared_1[((threadIdx.x*72) + 4678)]))
+        conv2d_nchw_1[11] = (conv2d_nchw_1[11] + (pad_temp.shared_1[212]*kernel.shared_1[((threadIdx.x*72) + 4678)]))
+        conv2d_nchw_1[12] = (conv2d_nchw_1[12] + (pad_temp.shared_1[213]*kernel.shared_1[((threadIdx.x*72) + 4678)]))
+        conv2d_nchw_1[13] = (conv2d_nchw_1[13] + (pad_temp.shared_1[214]*kernel.shared_1[((threadIdx.x*72) + 4678)]))
+        conv2d_nchw_1[0] = (conv2d_nchw_1[0] + (pad_temp.shared_1[209]*kernel.shared_1[((threadIdx.x*72) + 71)]))
+        conv2d_nchw_1[1] = (conv2d_nchw_1[1] + (pad_temp.shared_1[210]*kernel.shared_1[((threadIdx.x*72) + 71)]))
+        conv2d_nchw_1[2] = (conv2d_nchw_1[2] + (pad_temp.shared_1[211]*kernel.shared_1[((threadIdx.x*72) + 71)]))
+        conv2d_nchw_1[3] = (conv2d_nchw_1[3] + (pad_temp.shared_1[212]*kernel.shared_1[((threadIdx.x*72) + 71)]))
+        conv2d_nchw_1[4] = (conv2d_nchw_1[4] + (pad_temp.shared_1[213]*kernel.shared_1[((threadIdx.x*72) + 71)]))
+        conv2d_nchw_1[5] = (conv2d_nchw_1[5] + (pad_temp.shared_1[214]*kernel.shared_1[((threadIdx.x*72) + 71)]))
+        conv2d_nchw_1[6] = (conv2d_nchw_1[6] + (pad_temp.shared_1[215]*kernel.shared_1[((threadIdx.x*72) + 71)]))
+        conv2d_nchw_1[7] = (conv2d_nchw_1[7] + (pad_temp.shared_1[209]*kernel.shared_1[((threadIdx.x*72) + 4679)]))
+        conv2d_nchw_1[8] = (conv2d_nchw_1[8] + (pad_temp.shared_1[210]*kernel.shared_1[((threadIdx.x*72) + 4679)]))
+        conv2d_nchw_1[9] = (conv2d_nchw_1[9] + (pad_temp.shared_1[211]*kernel.shared_1[((threadIdx.x*72) + 4679)]))
+        conv2d_nchw_1[10] = (conv2d_nchw_1[10] + (pad_temp.shared_1[212]*kernel.shared_1[((threadIdx.x*72) + 4679)]))
+        conv2d_nchw_1[11] = (conv2d_nchw_1[11] + (pad_temp.shared_1[213]*kernel.shared_1[((threadIdx.x*72) + 4679)]))
+        conv2d_nchw_1[12] = (conv2d_nchw_1[12] + (pad_temp.shared_1[214]*kernel.shared_1[((threadIdx.x*72) + 4679)]))
+        conv2d_nchw_1[13] = (conv2d_nchw_1[13] + (pad_temp.shared_1[215]*kernel.shared_1[((threadIdx.x*72) + 4679)]))
       }
     }
-    for (i1.inner: int32, 0, 2) {
-      for (i3.inner: int32, 0, 7) {
-        compute_3: Buffer(compute_2, float32, [25088], [])[(((((floordiv(blockIdx.x, 7)*6272) + (threadIdx.x*98)) + (i1.inner*49)) + (floormod(blockIdx.x, 7)*7)) + i3.inner)] = max((conv2d_nchw_1[((i1.inner*7) + i3.inner)] + bias_3: Buffer(bias_2, float32, [512], [])[(((floordiv(blockIdx.x, 7)*128) + (threadIdx.x*2)) + i1.inner)]), 0f32)
-      }
-    }
+    compute_3: Buffer(compute_2, float32, [25088], [])[(((floordiv(blockIdx.x, 7)*6272) + (threadIdx.x*49)) + (floormod(blockIdx.x, 7)*7))] = max((conv2d_nchw_1[0] + bias_3: Buffer(bias_2, float32, [512], [])[((floordiv(blockIdx.x, 7)*128) + threadIdx.x)]), 0f32)
+    compute_3[((((floordiv(blockIdx.x, 7)*6272) + (threadIdx.x*49)) + (floormod(blockIdx.x, 7)*7)) + 1)] = max((conv2d_nchw_1[1] + bias_3[((floordiv(blockIdx.x, 7)*128) + threadIdx.x)]), 0f32)
+    compute_3[((((floordiv(blockIdx.x, 7)*6272) + (threadIdx.x*49)) + (floormod(blockIdx.x, 7)*7)) + 2)] = max((conv2d_nchw_1[2] + bias_3[((floordiv(blockIdx.x, 7)*128) + threadIdx.x)]), 0f32)
+    compute_3[((((floordiv(blockIdx.x, 7)*6272) + (threadIdx.x*49)) + (floormod(blockIdx.x, 7)*7)) + 3)] = max((conv2d_nchw_1[3] + bias_3[((floordiv(blockIdx.x, 7)*128) + threadIdx.x)]), 0f32)
+    compute_3[((((floordiv(blockIdx.x, 7)*6272) + (threadIdx.x*49)) + (floormod(blockIdx.x, 7)*7)) + 4)] = max((conv2d_nchw_1[4] + bias_3[((floordiv(blockIdx.x, 7)*128) + threadIdx.x)]), 0f32)
+    compute_3[((((floordiv(blockIdx.x, 7)*6272) + (threadIdx.x*49)) + (floormod(blockIdx.x, 7)*7)) + 5)] = max((conv2d_nchw_1[5] + bias_3[((floordiv(blockIdx.x, 7)*128) + threadIdx.x)]), 0f32)
+    compute_3[((((floordiv(blockIdx.x, 7)*6272) + (threadIdx.x*49)) + (floormod(blockIdx.x, 7)*7)) + 6)] = max((conv2d_nchw_1[6] + bias_3[((floordiv(blockIdx.x, 7)*128) + threadIdx.x)]), 0f32)
+    compute_3[((((floordiv(blockIdx.x, 7)*6272) + (threadIdx.x*49)) + (floormod(blockIdx.x, 7)*7)) + 3136)] = max((conv2d_nchw_1[7] + bias_3[(((floordiv(blockIdx.x, 7)*128) + threadIdx.x) + 64)]), 0f32)
+    compute_3[((((floordiv(blockIdx.x, 7)*6272) + (threadIdx.x*49)) + (floormod(blockIdx.x, 7)*7)) + 3137)] = max((conv2d_nchw_1[8] + bias_3[(((floordiv(blockIdx.x, 7)*128) + threadIdx.x) + 64)]), 0f32)
+    compute_3[((((floordiv(blockIdx.x, 7)*6272) + (threadIdx.x*49)) + (floormod(blockIdx.x, 7)*7)) + 3138)] = max((conv2d_nchw_1[9] + bias_3[(((floordiv(blockIdx.x, 7)*128) + threadIdx.x) + 64)]), 0f32)
+    compute_3[((((floordiv(blockIdx.x, 7)*6272) + (threadIdx.x*49)) + (floormod(blockIdx.x, 7)*7)) + 3139)] = max((conv2d_nchw_1[10] + bias_3[(((floordiv(blockIdx.x, 7)*128) + threadIdx.x) + 64)]), 0f32)
+    compute_3[((((floordiv(blockIdx.x, 7)*6272) + (threadIdx.x*49)) + (floormod(blockIdx.x, 7)*7)) + 3140)] = max((conv2d_nchw_1[11] + bias_3[(((floordiv(blockIdx.x, 7)*128) + threadIdx.x) + 64)]), 0f32)
+    compute_3[((((floordiv(blockIdx.x, 7)*6272) + (threadIdx.x*49)) + (floormod(blockIdx.x, 7)*7)) + 3141)] = max((conv2d_nchw_1[12] + bias_3[(((floordiv(blockIdx.x, 7)*128) + threadIdx.x) + 64)]), 0f32)
+    compute_3[((((floordiv(blockIdx.x, 7)*6272) + (threadIdx.x*49)) + (floormod(blockIdx.x, 7)*7)) + 3142)] = max((conv2d_nchw_1[13] + bias_3[(((floordiv(blockIdx.x, 7)*128) + threadIdx.x) + 64)]), 0f32)
   }
 }
 </pre></div>
@@ -1020,7 +1884,7 @@ cooperative fetching, unrolling and operator fusion.</p>
 <span class="p">)</span>
 </pre></div>
 </div>
-<div class="sphx-glr-script-out highlight-none notranslate"><div class="highlight"><pre><span></span>Execution time of this operator: 0.360 ms
+<div class="sphx-glr-script-out highlight-none notranslate"><div class="highlight"><pre><span></span>Execution time of this operator: 0.271 ms
 </pre></div>
 </div>
 </div>
@@ -1050,36 +1914,36 @@ conv2d_nchw_nn_o_o_i, conv2d_nchw_nn_o_i = s[conv2d_nchw].split(conv2d_nchw_nn_o
 conv2d_nchw_nn_o_o_o_i, conv2d_nchw_nn_o_o_i = s[conv2d_nchw].split(conv2d_nchw_nn_o_o_i, factor=1)
 conv2d_nchw_nn_o_o_o_o, conv2d_nchw_nn_o_o_o_i = s[conv2d_nchw].split(conv2d_nchw_nn_o_o_o_i, factor=1)
 conv2d_nchw_ff_o_i, conv2d_nchw_ff_i = s[conv2d_nchw].split(conv2d_nchw_ff, factor=1)
-conv2d_nchw_ff_o_o_i, conv2d_nchw_ff_o_i = s[conv2d_nchw].split(conv2d_nchw_ff_o_i, factor=2)
+conv2d_nchw_ff_o_o_i, conv2d_nchw_ff_o_i = s[conv2d_nchw].split(conv2d_nchw_ff_o_i, factor=1)
 conv2d_nchw_ff_o_o_o_i, conv2d_nchw_ff_o_o_i = s[conv2d_nchw].split(conv2d_nchw_ff_o_o_i, factor=64)
-conv2d_nchw_ff_o_o_o_o, conv2d_nchw_ff_o_o_o_i = s[conv2d_nchw].split(conv2d_nchw_ff_o_o_o_i, factor=1)
+conv2d_nchw_ff_o_o_o_o, conv2d_nchw_ff_o_o_o_i = s[conv2d_nchw].split(conv2d_nchw_ff_o_o_o_i, factor=2)
 conv2d_nchw_yy_o_i, conv2d_nchw_yy_i = s[conv2d_nchw].split(conv2d_nchw_yy, factor=1)
 conv2d_nchw_yy_o_o_i, conv2d_nchw_yy_o_i = s[conv2d_nchw].split(conv2d_nchw_yy_o_i, factor=1)
 conv2d_nchw_yy_o_o_o_i, conv2d_nchw_yy_o_o_i = s[conv2d_nchw].split(conv2d_nchw_yy_o_o_i, factor=1)
 conv2d_nchw_yy_o_o_o_o, conv2d_nchw_yy_o_o_o_i = s[conv2d_nchw].split(conv2d_nchw_yy_o_o_o_i, factor=1)
 conv2d_nchw_xx_o_i, conv2d_nchw_xx_i = s[conv2d_nchw].split(conv2d_nchw_xx, factor=1)
-conv2d_nchw_xx_o_o_i, conv2d_nchw_xx_o_i = s[conv2d_nchw].split(conv2d_nchw_xx_o_i, factor=7)
+conv2d_nchw_xx_o_o_i, conv2d_nchw_xx_o_i = s[conv2d_nchw].split(conv2d_nchw_xx_o_i, factor=1)
 conv2d_nchw_xx_o_o_o_i, conv2d_nchw_xx_o_o_i = s[conv2d_nchw].split(conv2d_nchw_xx_o_o_i, factor=1)
-conv2d_nchw_xx_o_o_o_o, conv2d_nchw_xx_o_o_o_i = s[conv2d_nchw].split(conv2d_nchw_xx_o_o_o_i, factor=1)
-conv2d_nchw_rc_o_i, conv2d_nchw_rc_i = s[conv2d_nchw].split(conv2d_nchw_rc, factor=2)
-conv2d_nchw_rc_o_o, conv2d_nchw_rc_o_i = s[conv2d_nchw].split(conv2d_nchw_rc_o_i, factor=4)
-conv2d_nchw_ry_o_i, conv2d_nchw_ry_i = s[conv2d_nchw].split(conv2d_nchw_ry, factor=1)
+conv2d_nchw_xx_o_o_o_o, conv2d_nchw_xx_o_o_o_i = s[conv2d_nchw].split(conv2d_nchw_xx_o_o_o_i, factor=7)
+conv2d_nchw_rc_o_i, conv2d_nchw_rc_i = s[conv2d_nchw].split(conv2d_nchw_rc, factor=1)
+conv2d_nchw_rc_o_o, conv2d_nchw_rc_o_i = s[conv2d_nchw].split(conv2d_nchw_rc_o_i, factor=8)
+conv2d_nchw_ry_o_i, conv2d_nchw_ry_i = s[conv2d_nchw].split(conv2d_nchw_ry, factor=3)
 conv2d_nchw_ry_o_o, conv2d_nchw_ry_o_i = s[conv2d_nchw].split(conv2d_nchw_ry_o_i, factor=1)
-conv2d_nchw_rx_o_i, conv2d_nchw_rx_i = s[conv2d_nchw].split(conv2d_nchw_rx, factor=1)
-conv2d_nchw_rx_o_o, conv2d_nchw_rx_o_i = s[conv2d_nchw].split(conv2d_nchw_rx_o_i, factor=3)
+conv2d_nchw_rx_o_i, conv2d_nchw_rx_i = s[conv2d_nchw].split(conv2d_nchw_rx, factor=3)
+conv2d_nchw_rx_o_o, conv2d_nchw_rx_o_i = s[conv2d_nchw].split(conv2d_nchw_rx_o_i, factor=1)
 s[conv2d_nchw].reorder(conv2d_nchw_nn_o_o_o_o, conv2d_nchw_ff_o_o_o_o, conv2d_nchw_yy_o_o_o_o, conv2d_nchw_xx_o_o_o_o, conv2d_nchw_nn_o_o_o_i, conv2d_nchw_ff_o_o_o_i, conv2d_nchw_yy_o_o_o_i, conv2d_nchw_xx_o_o_o_i, conv2d_nchw_nn_o_o_i, conv2d_nchw_ff_o_o_i, conv2d_nchw_yy_o_o_i, conv2d_nchw_xx_o_o_i, conv2d_nchw_rc_o_o, conv2d_nchw_ry_o_o, conv2d_nchw_rx_o_o, conv2d_nchw_rc_o_i, conv2d_nchw_ry_o_i, conv2d_nchw_rx_o_i, conv2d_nchw_nn_o_i, conv2d_nchw_ff_o_i, conv2d_nchw_yy_o_i, conv2d_nc [...]
 compute_i0_o_i, compute_i0_i = s[compute].split(compute_i0, factor=1)
 compute_i0_o_o_i, compute_i0_o_i = s[compute].split(compute_i0_o_i, factor=1)
 compute_i0_o_o_o, compute_i0_o_o_i = s[compute].split(compute_i0_o_o_i, factor=1)
-compute_i1_o_i, compute_i1_i = s[compute].split(compute_i1, factor=2)
+compute_i1_o_i, compute_i1_i = s[compute].split(compute_i1, factor=1)
 compute_i1_o_o_i, compute_i1_o_i = s[compute].split(compute_i1_o_i, factor=64)
-compute_i1_o_o_o, compute_i1_o_o_i = s[compute].split(compute_i1_o_o_i, factor=1)
+compute_i1_o_o_o, compute_i1_o_o_i = s[compute].split(compute_i1_o_o_i, factor=2)
 compute_i2_o_i, compute_i2_i = s[compute].split(compute_i2, factor=1)
 compute_i2_o_o_i, compute_i2_o_i = s[compute].split(compute_i2_o_i, factor=1)
 compute_i2_o_o_o, compute_i2_o_o_i = s[compute].split(compute_i2_o_o_i, factor=1)
-compute_i3_o_i, compute_i3_i = s[compute].split(compute_i3, factor=7)
+compute_i3_o_i, compute_i3_i = s[compute].split(compute_i3, factor=1)
 compute_i3_o_o_i, compute_i3_o_i = s[compute].split(compute_i3_o_i, factor=1)
-compute_i3_o_o_o, compute_i3_o_o_i = s[compute].split(compute_i3_o_o_i, factor=1)
+compute_i3_o_o_o, compute_i3_o_o_i = s[compute].split(compute_i3_o_o_i, factor=7)
 s[compute].reorder(compute_i0_o_o_o, compute_i1_o_o_o, compute_i2_o_o_o, compute_i3_o_o_o, compute_i0_o_o_i, compute_i1_o_o_i, compute_i2_o_o_i, compute_i3_o_o_i, compute_i0_o_i, compute_i1_o_i, compute_i2_o_i, compute_i3_o_i, compute_i0_i, compute_i1_i, compute_i2_i, compute_i3_i)
 s[conv2d_nchw].compute_at(s[compute], compute_i3_o_i)
 kernel_shared = s.cache_read(kernel, &quot;shared&quot;, [conv2d_nchw])
@@ -1101,11 +1965,11 @@ s[kernel_shared].vectorize(kernel_shared_ax0_ax1_fused_ax2_fused_ax3_fused_i)
 kernel_shared_ax0_ax1_fused_ax2_fused_ax3_fused_o_o, kernel_shared_ax0_ax1_fused_ax2_fused_ax3_fused_o_i = s[kernel_shared].split(kernel_shared_ax0_ax1_fused_ax2_fused_ax3_fused_o, factor=64)
 s[kernel_shared].bind(kernel_shared_ax0_ax1_fused_ax2_fused_ax3_fused_o_i, te.thread_axis(&quot;threadIdx.x&quot;))
 pad_temp_shared_ax0_ax1_fused_ax2_fused_ax3_fused = s[pad_temp_shared].fuse(pad_temp_shared_ax0, pad_temp_shared_ax1, pad_temp_shared_ax2, pad_temp_shared_ax3)
-pad_temp_shared_ax0_ax1_fused_ax2_fused_ax3_fused_o, pad_temp_shared_ax0_ax1_fused_ax2_fused_ax3_fused_i = s[pad_temp_shared].split(pad_temp_shared_ax0_ax1_fused_ax2_fused_ax3_fused, factor=4)
+pad_temp_shared_ax0_ax1_fused_ax2_fused_ax3_fused_o, pad_temp_shared_ax0_ax1_fused_ax2_fused_ax3_fused_i = s[pad_temp_shared].split(pad_temp_shared_ax0_ax1_fused_ax2_fused_ax3_fused, factor=1)
 s[pad_temp_shared].vectorize(pad_temp_shared_ax0_ax1_fused_ax2_fused_ax3_fused_i)
 pad_temp_shared_ax0_ax1_fused_ax2_fused_ax3_fused_o_o, pad_temp_shared_ax0_ax1_fused_ax2_fused_ax3_fused_o_i = s[pad_temp_shared].split(pad_temp_shared_ax0_ax1_fused_ax2_fused_ax3_fused_o, factor=64)
 s[pad_temp_shared].bind(pad_temp_shared_ax0_ax1_fused_ax2_fused_ax3_fused_o_i, te.thread_axis(&quot;threadIdx.x&quot;))
-s[conv2d_nchw].pragma(conv2d_nchw_nn_o_o_o_o, &quot;auto_unroll_max_step&quot;, 512)
+s[conv2d_nchw].pragma(conv2d_nchw_nn_o_o_o_o, &quot;auto_unroll_max_step&quot;, 1024)
 s[conv2d_nchw].pragma(conv2d_nchw_nn_o_o_o_o, &quot;unroll_explicit&quot;, True)
 
 CUDA source code:
@@ -1125,8 +1989,8 @@ CUDA source code:
 #endif
 extern &quot;C&quot; __global__ void __launch_bounds__(64) default_function_kernel0(float* __restrict__ data, float* __restrict__ kernel, float* __restrict__ compute, float* __restrict__ bias) {
   float conv2d_nchw[14];
-  __shared__ float pad_temp_shared[72];
-  __shared__ float kernel_shared[3072];
+  __shared__ float pad_temp_shared[216];
+  __shared__ float kernel_shared[9216];
   conv2d_nchw[0] = 0.000000e+00f;
   conv2d_nchw[1] = 0.000000e+00f;
   conv2d_nchw[2] = 0.000000e+00f;
@@ -1142,412 +2006,1181 @@ extern &quot;C&quot; __global__ void __launch_bounds__(64) default_function_kern
   conv2d_nchw[12] = 0.000000e+00f;
   conv2d_nchw[13] = 0.000000e+00f;
   for (int rc_outer_outer = 0; rc_outer_outer &lt; 64; ++rc_outer_outer) {
-    for (int ry_outer_outer = 0; ry_outer_outer &lt; 3; ++ry_outer_outer) {
-      __syncthreads();
-      if (((int)threadIdx.x) &lt; 18) {
-        pad_temp_shared[(((int)threadIdx.x) * 4)] = (((((1 &lt;= (ry_outer_outer + (((int)blockIdx.x) % 7))) &amp;&amp; ((ry_outer_outer + (((int)blockIdx.x) % 7)) &lt; 8)) &amp;&amp; (1 &lt;= ((((int)threadIdx.x) * 4) % 9))) &amp;&amp; (((((int)threadIdx.x) * 4) % 9) &lt; 8)) ? data[((((((rc_outer_outer * 392) + (((((int)threadIdx.x) * 4) / 9) * 49)) + (ry_outer_outer * 7)) + ((((int)blockIdx.x) % 7) * 7)) + ((((int)threadIdx.x) * 4) % 9)) - 8)] : 0.000000e+00f);
-      }
-      if (((int)threadIdx.x) &lt; 18) {
-        pad_temp_shared[((((int)threadIdx.x) * 4) + 1)] = (((((1 &lt;= (ry_outer_outer + (((int)blockIdx.x) % 7))) &amp;&amp; ((ry_outer_outer + (((int)blockIdx.x) % 7)) &lt; 8)) &amp;&amp; (1 &lt;= (((((int)threadIdx.x) * 4) + 1) % 9))) &amp;&amp; ((((((int)threadIdx.x) * 4) + 1) % 9) &lt; 8)) ? data[((((((rc_outer_outer * 392) + ((((((int)threadIdx.x) * 4) + 1) / 9) * 49)) + (ry_outer_outer * 7)) + ((((int)blockIdx.x) % 7) * 7)) + (((((int)threadIdx.x) * 4) + 1) % 9)) - 8)] : 0.000000e+00f);
-      }
-      if (((int)threadIdx.x) &lt; 18) {
-        pad_temp_shared[((((int)threadIdx.x) * 4) + 2)] = (((((1 &lt;= (ry_outer_outer + (((int)blockIdx.x) % 7))) &amp;&amp; ((ry_outer_outer + (((int)blockIdx.x) % 7)) &lt; 8)) &amp;&amp; (1 &lt;= (((((int)threadIdx.x) * 4) + 2) % 9))) &amp;&amp; ((((((int)threadIdx.x) * 4) + 2) % 9) &lt; 8)) ? data[((((((rc_outer_outer * 392) + ((((((int)threadIdx.x) * 4) + 2) / 9) * 49)) + (ry_outer_outer * 7)) + ((((int)blockIdx.x) % 7) * 7)) + (((((int)threadIdx.x) * 4) + 2) % 9)) - 8)] : 0.000000e+00f);
-      }
-      if (((int)threadIdx.x) &lt; 18) {
-        pad_temp_shared[((((int)threadIdx.x) * 4) + 3)] = (((((1 &lt;= (ry_outer_outer + (((int)blockIdx.x) % 7))) &amp;&amp; ((ry_outer_outer + (((int)blockIdx.x) % 7)) &lt; 8)) &amp;&amp; (1 &lt;= (((((int)threadIdx.x) * 4) + 3) % 9))) &amp;&amp; ((((((int)threadIdx.x) * 4) + 3) % 9) &lt; 8)) ? data[((((((rc_outer_outer * 392) + ((((((int)threadIdx.x) * 4) + 3) / 9) * 49)) + (ry_outer_outer * 7)) + ((((int)blockIdx.x) % 7) * 7)) + (((((int)threadIdx.x) * 4) + 3) % 9)) - 8)] : 0.000000e+00f);
-      }
-      kernel_shared[((int)threadIdx.x)] = kernel[(((((((((int)blockIdx.x) / 7) * 589824) + ((((int)threadIdx.x) / 24) * 4608)) + (rc_outer_outer * 72)) + (((((int)threadIdx.x) % 24) / 3) * 9)) + (ry_outer_outer * 3)) + (((int)threadIdx.x) % 3))];
-      kernel_shared[(((int)threadIdx.x) + 64)] = kernel[(((((((((int)blockIdx.x) / 7) * 589824) + (((((int)threadIdx.x) + 64) / 24) * 4608)) + (rc_outer_outer * 72)) + ((((((int)threadIdx.x) + 16) % 24) / 3) * 9)) + (ry_outer_outer * 3)) + ((((int)threadIdx.x) + 1) % 3))];
-      kernel_shared[(((int)threadIdx.x) + 128)] = kernel[(((((((((int)blockIdx.x) / 7) * 589824) + (((((int)threadIdx.x) + 128) / 24) * 4608)) + (rc_outer_outer * 72)) + ((((((int)threadIdx.x) + 8) % 24) / 3) * 9)) + (ry_outer_outer * 3)) + ((((int)threadIdx.x) + 2) % 3))];
-      kernel_shared[(((int)threadIdx.x) + 192)] = kernel[((((((((((int)blockIdx.x) / 7) * 589824) + ((((int)threadIdx.x) / 24) * 4608)) + (rc_outer_outer * 72)) + (((((int)threadIdx.x) % 24) / 3) * 9)) + (ry_outer_outer * 3)) + (((int)threadIdx.x) % 3)) + 36864)];
-      kernel_shared[(((int)threadIdx.x) + 256)] = kernel[(((((((((int)blockIdx.x) / 7) * 589824) + (((((int)threadIdx.x) + 256) / 24) * 4608)) + (rc_outer_outer * 72)) + ((((((int)threadIdx.x) + 16) % 24) / 3) * 9)) + (ry_outer_outer * 3)) + ((((int)threadIdx.x) + 1) % 3))];
-      kernel_shared[(((int)threadIdx.x) + 320)] = kernel[(((((((((int)blockIdx.x) / 7) * 589824) + (((((int)threadIdx.x) + 320) / 24) * 4608)) + (rc_outer_outer * 72)) + ((((((int)threadIdx.x) + 8) % 24) / 3) * 9)) + (ry_outer_outer * 3)) + ((((int)threadIdx.x) + 2) % 3))];
-      kernel_shared[(((int)threadIdx.x) + 384)] = kernel[((((((((((int)blockIdx.x) / 7) * 589824) + ((((int)threadIdx.x) / 24) * 4608)) + (rc_outer_outer * 72)) + (((((int)threadIdx.x) % 24) / 3) * 9)) + (ry_outer_outer * 3)) + (((int)threadIdx.x) % 3)) + 73728)];
-      kernel_shared[(((int)threadIdx.x) + 448)] = kernel[(((((((((int)blockIdx.x) / 7) * 589824) + (((((int)threadIdx.x) + 448) / 24) * 4608)) + (rc_outer_outer * 72)) + ((((((int)threadIdx.x) + 16) % 24) / 3) * 9)) + (ry_outer_outer * 3)) + ((((int)threadIdx.x) + 1) % 3))];
-      kernel_shared[(((int)threadIdx.x) + 512)] = kernel[(((((((((int)blockIdx.x) / 7) * 589824) + (((((int)threadIdx.x) + 512) / 24) * 4608)) + (rc_outer_outer * 72)) + ((((((int)threadIdx.x) + 8) % 24) / 3) * 9)) + (ry_outer_outer * 3)) + ((((int)threadIdx.x) + 2) % 3))];
-      kernel_shared[(((int)threadIdx.x) + 576)] = kernel[((((((((((int)blockIdx.x) / 7) * 589824) + ((((int)threadIdx.x) / 24) * 4608)) + (rc_outer_outer * 72)) + (((((int)threadIdx.x) % 24) / 3) * 9)) + (ry_outer_outer * 3)) + (((int)threadIdx.x) % 3)) + 110592)];
-      kernel_shared[(((int)threadIdx.x) + 640)] = kernel[(((((((((int)blockIdx.x) / 7) * 589824) + (((((int)threadIdx.x) + 640) / 24) * 4608)) + (rc_outer_outer * 72)) + ((((((int)threadIdx.x) + 16) % 24) / 3) * 9)) + (ry_outer_outer * 3)) + ((((int)threadIdx.x) + 1) % 3))];
-      kernel_shared[(((int)threadIdx.x) + 704)] = kernel[(((((((((int)blockIdx.x) / 7) * 589824) + (((((int)threadIdx.x) + 704) / 24) * 4608)) + (rc_outer_outer * 72)) + ((((((int)threadIdx.x) + 8) % 24) / 3) * 9)) + (ry_outer_outer * 3)) + ((((int)threadIdx.x) + 2) % 3))];
-      kernel_shared[(((int)threadIdx.x) + 768)] = kernel[((((((((((int)blockIdx.x) / 7) * 589824) + ((((int)threadIdx.x) / 24) * 4608)) + (rc_outer_outer * 72)) + (((((int)threadIdx.x) % 24) / 3) * 9)) + (ry_outer_outer * 3)) + (((int)threadIdx.x) % 3)) + 147456)];
-      kernel_shared[(((int)threadIdx.x) + 832)] = kernel[(((((((((int)blockIdx.x) / 7) * 589824) + (((((int)threadIdx.x) + 832) / 24) * 4608)) + (rc_outer_outer * 72)) + ((((((int)threadIdx.x) + 16) % 24) / 3) * 9)) + (ry_outer_outer * 3)) + ((((int)threadIdx.x) + 1) % 3))];
-      kernel_shared[(((int)threadIdx.x) + 896)] = kernel[(((((((((int)blockIdx.x) / 7) * 589824) + (((((int)threadIdx.x) + 896) / 24) * 4608)) + (rc_outer_outer * 72)) + ((((((int)threadIdx.x) + 8) % 24) / 3) * 9)) + (ry_outer_outer * 3)) + ((((int)threadIdx.x) + 2) % 3))];
-      kernel_shared[(((int)threadIdx.x) + 960)] = kernel[((((((((((int)blockIdx.x) / 7) * 589824) + ((((int)threadIdx.x) / 24) * 4608)) + (rc_outer_outer * 72)) + (((((int)threadIdx.x) % 24) / 3) * 9)) + (ry_outer_outer * 3)) + (((int)threadIdx.x) % 3)) + 184320)];
-      kernel_shared[(((int)threadIdx.x) + 1024)] = kernel[(((((((((int)blockIdx.x) / 7) * 589824) + (((((int)threadIdx.x) + 1024) / 24) * 4608)) + (rc_outer_outer * 72)) + ((((((int)threadIdx.x) + 16) % 24) / 3) * 9)) + (ry_outer_outer * 3)) + ((((int)threadIdx.x) + 1) % 3))];
-      kernel_shared[(((int)threadIdx.x) + 1088)] = kernel[(((((((((int)blockIdx.x) / 7) * 589824) + (((((int)threadIdx.x) + 1088) / 24) * 4608)) + (rc_outer_outer * 72)) + ((((((int)threadIdx.x) + 8) % 24) / 3) * 9)) + (ry_outer_outer * 3)) + ((((int)threadIdx.x) + 2) % 3))];
-      kernel_shared[(((int)threadIdx.x) + 1152)] = kernel[((((((((((int)blockIdx.x) / 7) * 589824) + ((((int)threadIdx.x) / 24) * 4608)) + (rc_outer_outer * 72)) + (((((int)threadIdx.x) % 24) / 3) * 9)) + (ry_outer_outer * 3)) + (((int)threadIdx.x) % 3)) + 221184)];
-      kernel_shared[(((int)threadIdx.x) + 1216)] = kernel[(((((((((int)blockIdx.x) / 7) * 589824) + (((((int)threadIdx.x) + 1216) / 24) * 4608)) + (rc_outer_outer * 72)) + ((((((int)threadIdx.x) + 16) % 24) / 3) * 9)) + (ry_outer_outer * 3)) + ((((int)threadIdx.x) + 1) % 3))];
-      kernel_shared[(((int)threadIdx.x) + 1280)] = kernel[(((((((((int)blockIdx.x) / 7) * 589824) + (((((int)threadIdx.x) + 1280) / 24) * 4608)) + (rc_outer_outer * 72)) + ((((((int)threadIdx.x) + 8) % 24) / 3) * 9)) + (ry_outer_outer * 3)) + ((((int)threadIdx.x) + 2) % 3))];
-      kernel_shared[(((int)threadIdx.x) + 1344)] = kernel[((((((((((int)blockIdx.x) / 7) * 589824) + ((((int)threadIdx.x) / 24) * 4608)) + (rc_outer_outer * 72)) + (((((int)threadIdx.x) % 24) / 3) * 9)) + (ry_outer_outer * 3)) + (((int)threadIdx.x) % 3)) + 258048)];
-      kernel_shared[(((int)threadIdx.x) + 1408)] = kernel[(((((((((int)blockIdx.x) / 7) * 589824) + (((((int)threadIdx.x) + 1408) / 24) * 4608)) + (rc_outer_outer * 72)) + ((((((int)threadIdx.x) + 16) % 24) / 3) * 9)) + (ry_outer_outer * 3)) + ((((int)threadIdx.x) + 1) % 3))];
-      kernel_shared[(((int)threadIdx.x) + 1472)] = kernel[(((((((((int)blockIdx.x) / 7) * 589824) + (((((int)threadIdx.x) + 1472) / 24) * 4608)) + (rc_outer_outer * 72)) + ((((((int)threadIdx.x) + 8) % 24) / 3) * 9)) + (ry_outer_outer * 3)) + ((((int)threadIdx.x) + 2) % 3))];
-      kernel_shared[(((int)threadIdx.x) + 1536)] = kernel[((((((((((int)blockIdx.x) / 7) * 589824) + ((((int)threadIdx.x) / 24) * 4608)) + (rc_outer_outer * 72)) + (((((int)threadIdx.x) % 24) / 3) * 9)) + (ry_outer_outer * 3)) + (((int)threadIdx.x) % 3)) + 294912)];
-      kernel_shared[(((int)threadIdx.x) + 1600)] = kernel[(((((((((int)blockIdx.x) / 7) * 589824) + (((((int)threadIdx.x) + 1600) / 24) * 4608)) + (rc_outer_outer * 72)) + ((((((int)threadIdx.x) + 16) % 24) / 3) * 9)) + (ry_outer_outer * 3)) + ((((int)threadIdx.x) + 1) % 3))];
-      kernel_shared[(((int)threadIdx.x) + 1664)] = kernel[(((((((((int)blockIdx.x) / 7) * 589824) + (((((int)threadIdx.x) + 1664) / 24) * 4608)) + (rc_outer_outer * 72)) + ((((((int)threadIdx.x) + 8) % 24) / 3) * 9)) + (ry_outer_outer * 3)) + ((((int)threadIdx.x) + 2) % 3))];
-      kernel_shared[(((int)threadIdx.x) + 1728)] = kernel[((((((((((int)blockIdx.x) / 7) * 589824) + ((((int)threadIdx.x) / 24) * 4608)) + (rc_outer_outer * 72)) + (((((int)threadIdx.x) % 24) / 3) * 9)) + (ry_outer_outer * 3)) + (((int)threadIdx.x) % 3)) + 331776)];
-      kernel_shared[(((int)threadIdx.x) + 1792)] = kernel[(((((((((int)blockIdx.x) / 7) * 589824) + (((((int)threadIdx.x) + 1792) / 24) * 4608)) + (rc_outer_outer * 72)) + ((((((int)threadIdx.x) + 16) % 24) / 3) * 9)) + (ry_outer_outer * 3)) + ((((int)threadIdx.x) + 1) % 3))];
-      kernel_shared[(((int)threadIdx.x) + 1856)] = kernel[(((((((((int)blockIdx.x) / 7) * 589824) + (((((int)threadIdx.x) + 1856) / 24) * 4608)) + (rc_outer_outer * 72)) + ((((((int)threadIdx.x) + 8) % 24) / 3) * 9)) + (ry_outer_outer * 3)) + ((((int)threadIdx.x) + 2) % 3))];
-      kernel_shared[(((int)threadIdx.x) + 1920)] = kernel[((((((((((int)blockIdx.x) / 7) * 589824) + ((((int)threadIdx.x) / 24) * 4608)) + (rc_outer_outer * 72)) + (((((int)threadIdx.x) % 24) / 3) * 9)) + (ry_outer_outer * 3)) + (((int)threadIdx.x) % 3)) + 368640)];
-      kernel_shared[(((int)threadIdx.x) + 1984)] = kernel[(((((((((int)blockIdx.x) / 7) * 589824) + (((((int)threadIdx.x) + 1984) / 24) * 4608)) + (rc_outer_outer * 72)) + ((((((int)threadIdx.x) + 16) % 24) / 3) * 9)) + (ry_outer_outer * 3)) + ((((int)threadIdx.x) + 1) % 3))];
-      kernel_shared[(((int)threadIdx.x) + 2048)] = kernel[(((((((((int)blockIdx.x) / 7) * 589824) + (((((int)threadIdx.x) + 2048) / 24) * 4608)) + (rc_outer_outer * 72)) + ((((((int)threadIdx.x) + 8) % 24) / 3) * 9)) + (ry_outer_outer * 3)) + ((((int)threadIdx.x) + 2) % 3))];
-      kernel_shared[(((int)threadIdx.x) + 2112)] = kernel[((((((((((int)blockIdx.x) / 7) * 589824) + ((((int)threadIdx.x) / 24) * 4608)) + (rc_outer_outer * 72)) + (((((int)threadIdx.x) % 24) / 3) * 9)) + (ry_outer_outer * 3)) + (((int)threadIdx.x) % 3)) + 405504)];
-      kernel_shared[(((int)threadIdx.x) + 2176)] = kernel[(((((((((int)blockIdx.x) / 7) * 589824) + (((((int)threadIdx.x) + 2176) / 24) * 4608)) + (rc_outer_outer * 72)) + ((((((int)threadIdx.x) + 16) % 24) / 3) * 9)) + (ry_outer_outer * 3)) + ((((int)threadIdx.x) + 1) % 3))];
-      kernel_shared[(((int)threadIdx.x) + 2240)] = kernel[(((((((((int)blockIdx.x) / 7) * 589824) + (((((int)threadIdx.x) + 2240) / 24) * 4608)) + (rc_outer_outer * 72)) + ((((((int)threadIdx.x) + 8) % 24) / 3) * 9)) + (ry_outer_outer * 3)) + ((((int)threadIdx.x) + 2) % 3))];
-      kernel_shared[(((int)threadIdx.x) + 2304)] = kernel[((((((((((int)blockIdx.x) / 7) * 589824) + ((((int)threadIdx.x) / 24) * 4608)) + (rc_outer_outer * 72)) + (((((int)threadIdx.x) % 24) / 3) * 9)) + (ry_outer_outer * 3)) + (((int)threadIdx.x) % 3)) + 442368)];
-      kernel_shared[(((int)threadIdx.x) + 2368)] = kernel[(((((((((int)blockIdx.x) / 7) * 589824) + (((((int)threadIdx.x) + 2368) / 24) * 4608)) + (rc_outer_outer * 72)) + ((((((int)threadIdx.x) + 16) % 24) / 3) * 9)) + (ry_outer_outer * 3)) + ((((int)threadIdx.x) + 1) % 3))];
-      kernel_shared[(((int)threadIdx.x) + 2432)] = kernel[(((((((((int)blockIdx.x) / 7) * 589824) + (((((int)threadIdx.x) + 2432) / 24) * 4608)) + (rc_outer_outer * 72)) + ((((((int)threadIdx.x) + 8) % 24) / 3) * 9)) + (ry_outer_outer * 3)) + ((((int)threadIdx.x) + 2) % 3))];
-      kernel_shared[(((int)threadIdx.x) + 2496)] = kernel[((((((((((int)blockIdx.x) / 7) * 589824) + ((((int)threadIdx.x) / 24) * 4608)) + (rc_outer_outer * 72)) + (((((int)threadIdx.x) % 24) / 3) * 9)) + (ry_outer_outer * 3)) + (((int)threadIdx.x) % 3)) + 479232)];
-      kernel_shared[(((int)threadIdx.x) + 2560)] = kernel[(((((((((int)blockIdx.x) / 7) * 589824) + (((((int)threadIdx.x) + 2560) / 24) * 4608)) + (rc_outer_outer * 72)) + ((((((int)threadIdx.x) + 16) % 24) / 3) * 9)) + (ry_outer_outer * 3)) + ((((int)threadIdx.x) + 1) % 3))];
-      kernel_shared[(((int)threadIdx.x) + 2624)] = kernel[(((((((((int)blockIdx.x) / 7) * 589824) + (((((int)threadIdx.x) + 2624) / 24) * 4608)) + (rc_outer_outer * 72)) + ((((((int)threadIdx.x) + 8) % 24) / 3) * 9)) + (ry_outer_outer * 3)) + ((((int)threadIdx.x) + 2) % 3))];
-      kernel_shared[(((int)threadIdx.x) + 2688)] = kernel[((((((((((int)blockIdx.x) / 7) * 589824) + ((((int)threadIdx.x) / 24) * 4608)) + (rc_outer_outer * 72)) + (((((int)threadIdx.x) % 24) / 3) * 9)) + (ry_outer_outer * 3)) + (((int)threadIdx.x) % 3)) + 516096)];
-      kernel_shared[(((int)threadIdx.x) + 2752)] = kernel[(((((((((int)blockIdx.x) / 7) * 589824) + (((((int)threadIdx.x) + 2752) / 24) * 4608)) + (rc_outer_outer * 72)) + ((((((int)threadIdx.x) + 16) % 24) / 3) * 9)) + (ry_outer_outer * 3)) + ((((int)threadIdx.x) + 1) % 3))];
-      kernel_shared[(((int)threadIdx.x) + 2816)] = kernel[(((((((((int)blockIdx.x) / 7) * 589824) + (((((int)threadIdx.x) + 2816) / 24) * 4608)) + (rc_outer_outer * 72)) + ((((((int)threadIdx.x) + 8) % 24) / 3) * 9)) + (ry_outer_outer * 3)) + ((((int)threadIdx.x) + 2) % 3))];
-      kernel_shared[(((int)threadIdx.x) + 2880)] = kernel[((((((((((int)blockIdx.x) / 7) * 589824) + ((((int)threadIdx.x) / 24) * 4608)) + (rc_outer_outer * 72)) + (((((int)threadIdx.x) % 24) / 3) * 9)) + (ry_outer_outer * 3)) + (((int)threadIdx.x) % 3)) + 552960)];
-      kernel_shared[(((int)threadIdx.x) + 2944)] = kernel[(((((((((int)blockIdx.x) / 7) * 589824) + (((((int)threadIdx.x) + 2944) / 24) * 4608)) + (rc_outer_outer * 72)) + ((((((int)threadIdx.x) + 16) % 24) / 3) * 9)) + (ry_outer_outer * 3)) + ((((int)threadIdx.x) + 1) % 3))];
-      kernel_shared[(((int)threadIdx.x) + 3008)] = kernel[(((((((((int)blockIdx.x) / 7) * 589824) + (((((int)threadIdx.x) + 3008) / 24) * 4608)) + (rc_outer_outer * 72)) + ((((((int)threadIdx.x) + 8) % 24) / 3) * 9)) + (ry_outer_outer * 3)) + ((((int)threadIdx.x) + 2) % 3))];
-      __syncthreads();
-      conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[0] * kernel_shared[(((int)threadIdx.x) * 48)]));
-      conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[9] * kernel_shared[((((int)threadIdx.x) * 48) + 3)]));
-      conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[1] * kernel_shared[(((int)threadIdx.x) * 48)]));
-      conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[10] * kernel_shared[((((int)threadIdx.x) * 48) + 3)]));
-      conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[2] * kernel_shared[(((int)threadIdx.x) * 48)]));
-      conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[11] * kernel_shared[((((int)threadIdx.x) * 48) + 3)]));
-      conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[3] * kernel_shared[(((int)threadIdx.x) * 48)]));
-      conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[12] * kernel_shared[((((int)threadIdx.x) * 48) + 3)]));
-      conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[4] * kernel_shared[(((int)threadIdx.x) * 48)]));
-      conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[13] * kernel_shared[((((int)threadIdx.x) * 48) + 3)]));
-      conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[5] * kernel_shared[(((int)threadIdx.x) * 48)]));
-      conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[14] * kernel_shared[((((int)threadIdx.x) * 48) + 3)]));
-      conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[6] * kernel_shared[(((int)threadIdx.x) * 48)]));
-      conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[15] * kernel_shared[((((int)threadIdx.x) * 48) + 3)]));
-      conv2d_nchw[7] = (conv2d_nchw[7] + (pad_temp_shared[0] * kernel_shared[((((int)threadIdx.x) * 48) + 24)]));
-      conv2d_nchw[7] = (conv2d_nchw[7] + (pad_temp_shared[9] * kernel_shared[((((int)threadIdx.x) * 48) + 27)]));
-      conv2d_nchw[8] = (conv2d_nchw[8] + (pad_temp_shared[1] * kernel_shared[((((int)threadIdx.x) * 48) + 24)]));
-      conv2d_nchw[8] = (conv2d_nchw[8] + (pad_temp_shared[10] * kernel_shared[((((int)threadIdx.x) * 48) + 27)]));
-      conv2d_nchw[9] = (conv2d_nchw[9] + (pad_temp_shared[2] * kernel_shared[((((int)threadIdx.x) * 48) + 24)]));
-      conv2d_nchw[9] = (conv2d_nchw[9] + (pad_temp_shared[11] * kernel_shared[((((int)threadIdx.x) * 48) + 27)]));
-      conv2d_nchw[10] = (conv2d_nchw[10] + (pad_temp_shared[3] * kernel_shared[((((int)threadIdx.x) * 48) + 24)]));
-      conv2d_nchw[10] = (conv2d_nchw[10] + (pad_temp_shared[12] * kernel_shared[((((int)threadIdx.x) * 48) + 27)]));
-      conv2d_nchw[11] = (conv2d_nchw[11] + (pad_temp_shared[4] * kernel_shared[((((int)threadIdx.x) * 48) + 24)]));
-      conv2d_nchw[11] = (conv2d_nchw[11] + (pad_temp_shared[13] * kernel_shared[((((int)threadIdx.x) * 48) + 27)]));
-      conv2d_nchw[12] = (conv2d_nchw[12] + (pad_temp_shared[5] * kernel_shared[((((int)threadIdx.x) * 48) + 24)]));
-      conv2d_nchw[12] = (conv2d_nchw[12] + (pad_temp_shared[14] * kernel_shared[((((int)threadIdx.x) * 48) + 27)]));
-      conv2d_nchw[13] = (conv2d_nchw[13] + (pad_temp_shared[6] * kernel_shared[((((int)threadIdx.x) * 48) + 24)]));
-      conv2d_nchw[13] = (conv2d_nchw[13] + (pad_temp_shared[15] * kernel_shared[((((int)threadIdx.x) * 48) + 27)]));
-      conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[1] * kernel_shared[((((int)threadIdx.x) * 48) + 1)]));
-      conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[10] * kernel_shared[((((int)threadIdx.x) * 48) + 4)]));
-      conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[2] * kernel_shared[((((int)threadIdx.x) * 48) + 1)]));
-      conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[11] * kernel_shared[((((int)threadIdx.x) * 48) + 4)]));
-      conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[3] * kernel_shared[((((int)threadIdx.x) * 48) + 1)]));
-      conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[12] * kernel_shared[((((int)threadIdx.x) * 48) + 4)]));
-      conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[4] * kernel_shared[((((int)threadIdx.x) * 48) + 1)]));
-      conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[13] * kernel_shared[((((int)threadIdx.x) * 48) + 4)]));
-      conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[5] * kernel_shared[((((int)threadIdx.x) * 48) + 1)]));
-      conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[14] * kernel_shared[((((int)threadIdx.x) * 48) + 4)]));
-      conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[6] * kernel_shared[((((int)threadIdx.x) * 48) + 1)]));
-      conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[15] * kernel_shared[((((int)threadIdx.x) * 48) + 4)]));
-      conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[7] * kernel_shared[((((int)threadIdx.x) * 48) + 1)]));
-      conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[16] * kernel_shared[((((int)threadIdx.x) * 48) + 4)]));
-      conv2d_nchw[7] = (conv2d_nchw[7] + (pad_temp_shared[1] * kernel_shared[((((int)threadIdx.x) * 48) + 25)]));
-      conv2d_nchw[7] = (conv2d_nchw[7] + (pad_temp_shared[10] * kernel_shared[((((int)threadIdx.x) * 48) + 28)]));
-      conv2d_nchw[8] = (conv2d_nchw[8] + (pad_temp_shared[2] * kernel_shared[((((int)threadIdx.x) * 48) + 25)]));
-      conv2d_nchw[8] = (conv2d_nchw[8] + (pad_temp_shared[11] * kernel_shared[((((int)threadIdx.x) * 48) + 28)]));
-      conv2d_nchw[9] = (conv2d_nchw[9] + (pad_temp_shared[3] * kernel_shared[((((int)threadIdx.x) * 48) + 25)]));
-      conv2d_nchw[9] = (conv2d_nchw[9] + (pad_temp_shared[12] * kernel_shared[((((int)threadIdx.x) * 48) + 28)]));
-      conv2d_nchw[10] = (conv2d_nchw[10] + (pad_temp_shared[4] * kernel_shared[((((int)threadIdx.x) * 48) + 25)]));
-      conv2d_nchw[10] = (conv2d_nchw[10] + (pad_temp_shared[13] * kernel_shared[((((int)threadIdx.x) * 48) + 28)]));
-      conv2d_nchw[11] = (conv2d_nchw[11] + (pad_temp_shared[5] * kernel_shared[((((int)threadIdx.x) * 48) + 25)]));
-      conv2d_nchw[11] = (conv2d_nchw[11] + (pad_temp_shared[14] * kernel_shared[((((int)threadIdx.x) * 48) + 28)]));
-      conv2d_nchw[12] = (conv2d_nchw[12] + (pad_temp_shared[6] * kernel_shared[((((int)threadIdx.x) * 48) + 25)]));
-      conv2d_nchw[12] = (conv2d_nchw[12] + (pad_temp_shared[15] * kernel_shared[((((int)threadIdx.x) * 48) + 28)]));
-      conv2d_nchw[13] = (conv2d_nchw[13] + (pad_temp_shared[7] * kernel_shared[((((int)threadIdx.x) * 48) + 25)]));
-      conv2d_nchw[13] = (conv2d_nchw[13] + (pad_temp_shared[16] * kernel_shared[((((int)threadIdx.x) * 48) + 28)]));
-      conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[2] * kernel_shared[((((int)threadIdx.x) * 48) + 2)]));
-      conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[11] * kernel_shared[((((int)threadIdx.x) * 48) + 5)]));
-      conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[3] * kernel_shared[((((int)threadIdx.x) * 48) + 2)]));
-      conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[12] * kernel_shared[((((int)threadIdx.x) * 48) + 5)]));
-      conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[4] * kernel_shared[((((int)threadIdx.x) * 48) + 2)]));
-      conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[13] * kernel_shared[((((int)threadIdx.x) * 48) + 5)]));
-      conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[5] * kernel_shared[((((int)threadIdx.x) * 48) + 2)]));
-      conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[14] * kernel_shared[((((int)threadIdx.x) * 48) + 5)]));
-      conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[6] * kernel_shared[((((int)threadIdx.x) * 48) + 2)]));
-      conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[15] * kernel_shared[((((int)threadIdx.x) * 48) + 5)]));
-      conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[7] * kernel_shared[((((int)threadIdx.x) * 48) + 2)]));
-      conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[16] * kernel_shared[((((int)threadIdx.x) * 48) + 5)]));
-      conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[8] * kernel_shared[((((int)threadIdx.x) * 48) + 2)]));
-      conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[17] * kernel_shared[((((int)threadIdx.x) * 48) + 5)]));
-      conv2d_nchw[7] = (conv2d_nchw[7] + (pad_temp_shared[2] * kernel_shared[((((int)threadIdx.x) * 48) + 26)]));
-      conv2d_nchw[7] = (conv2d_nchw[7] + (pad_temp_shared[11] * kernel_shared[((((int)threadIdx.x) * 48) + 29)]));
-      conv2d_nchw[8] = (conv2d_nchw[8] + (pad_temp_shared[3] * kernel_shared[((((int)threadIdx.x) * 48) + 26)]));
-      conv2d_nchw[8] = (conv2d_nchw[8] + (pad_temp_shared[12] * kernel_shared[((((int)threadIdx.x) * 48) + 29)]));
-      conv2d_nchw[9] = (conv2d_nchw[9] + (pad_temp_shared[4] * kernel_shared[((((int)threadIdx.x) * 48) + 26)]));
-      conv2d_nchw[9] = (conv2d_nchw[9] + (pad_temp_shared[13] * kernel_shared[((((int)threadIdx.x) * 48) + 29)]));
-      conv2d_nchw[10] = (conv2d_nchw[10] + (pad_temp_shared[5] * kernel_shared[((((int)threadIdx.x) * 48) + 26)]));
-      conv2d_nchw[10] = (conv2d_nchw[10] + (pad_temp_shared[14] * kernel_shared[((((int)threadIdx.x) * 48) + 29)]));
-      conv2d_nchw[11] = (conv2d_nchw[11] + (pad_temp_shared[6] * kernel_shared[((((int)threadIdx.x) * 48) + 26)]));
-      conv2d_nchw[11] = (conv2d_nchw[11] + (pad_temp_shared[15] * kernel_shared[((((int)threadIdx.x) * 48) + 29)]));
-      conv2d_nchw[12] = (conv2d_nchw[12] + (pad_temp_shared[7] * kernel_shared[((((int)threadIdx.x) * 48) + 26)]));
-      conv2d_nchw[12] = (conv2d_nchw[12] + (pad_temp_shared[16] * kernel_shared[((((int)threadIdx.x) * 48) + 29)]));
-      conv2d_nchw[13] = (conv2d_nchw[13] + (pad_temp_shared[8] * kernel_shared[((((int)threadIdx.x) * 48) + 26)]));
-      conv2d_nchw[13] = (conv2d_nchw[13] + (pad_temp_shared[17] * kernel_shared[((((int)threadIdx.x) * 48) + 29)]));
-      conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[18] * kernel_shared[((((int)threadIdx.x) * 48) + 6)]));
-      conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[27] * kernel_shared[((((int)threadIdx.x) * 48) + 9)]));
-      conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[19] * kernel_shared[((((int)threadIdx.x) * 48) + 6)]));
-      conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[28] * kernel_shared[((((int)threadIdx.x) * 48) + 9)]));
-      conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[20] * kernel_shared[((((int)threadIdx.x) * 48) + 6)]));
-      conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[29] * kernel_shared[((((int)threadIdx.x) * 48) + 9)]));
-      conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[21] * kernel_shared[((((int)threadIdx.x) * 48) + 6)]));
-      conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[30] * kernel_shared[((((int)threadIdx.x) * 48) + 9)]));
-      conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[22] * kernel_shared[((((int)threadIdx.x) * 48) + 6)]));
-      conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[31] * kernel_shared[((((int)threadIdx.x) * 48) + 9)]));
-      conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[23] * kernel_shared[((((int)threadIdx.x) * 48) + 6)]));
-      conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[32] * kernel_shared[((((int)threadIdx.x) * 48) + 9)]));
-      conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[24] * kernel_shared[((((int)threadIdx.x) * 48) + 6)]));
-      conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[33] * kernel_shared[((((int)threadIdx.x) * 48) + 9)]));
-      conv2d_nchw[7] = (conv2d_nchw[7] + (pad_temp_shared[18] * kernel_shared[((((int)threadIdx.x) * 48) + 30)]));
-      conv2d_nchw[7] = (conv2d_nchw[7] + (pad_temp_shared[27] * kernel_shared[((((int)threadIdx.x) * 48) + 33)]));
-      conv2d_nchw[8] = (conv2d_nchw[8] + (pad_temp_shared[19] * kernel_shared[((((int)threadIdx.x) * 48) + 30)]));
-      conv2d_nchw[8] = (conv2d_nchw[8] + (pad_temp_shared[28] * kernel_shared[((((int)threadIdx.x) * 48) + 33)]));
-      conv2d_nchw[9] = (conv2d_nchw[9] + (pad_temp_shared[20] * kernel_shared[((((int)threadIdx.x) * 48) + 30)]));
-      conv2d_nchw[9] = (conv2d_nchw[9] + (pad_temp_shared[29] * kernel_shared[((((int)threadIdx.x) * 48) + 33)]));
-      conv2d_nchw[10] = (conv2d_nchw[10] + (pad_temp_shared[21] * kernel_shared[((((int)threadIdx.x) * 48) + 30)]));
-      conv2d_nchw[10] = (conv2d_nchw[10] + (pad_temp_shared[30] * kernel_shared[((((int)threadIdx.x) * 48) + 33)]));
-      conv2d_nchw[11] = (conv2d_nchw[11] + (pad_temp_shared[22] * kernel_shared[((((int)threadIdx.x) * 48) + 30)]));
-      conv2d_nchw[11] = (conv2d_nchw[11] + (pad_temp_shared[31] * kernel_shared[((((int)threadIdx.x) * 48) + 33)]));
-      conv2d_nchw[12] = (conv2d_nchw[12] + (pad_temp_shared[23] * kernel_shared[((((int)threadIdx.x) * 48) + 30)]));
-      conv2d_nchw[12] = (conv2d_nchw[12] + (pad_temp_shared[32] * kernel_shared[((((int)threadIdx.x) * 48) + 33)]));
-      conv2d_nchw[13] = (conv2d_nchw[13] + (pad_temp_shared[24] * kernel_shared[((((int)threadIdx.x) * 48) + 30)]));
-      conv2d_nchw[13] = (conv2d_nchw[13] + (pad_temp_shared[33] * kernel_shared[((((int)threadIdx.x) * 48) + 33)]));
-      conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[19] * kernel_shared[((((int)threadIdx.x) * 48) + 7)]));
-      conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[28] * kernel_shared[((((int)threadIdx.x) * 48) + 10)]));
-      conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[20] * kernel_shared[((((int)threadIdx.x) * 48) + 7)]));
-      conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[29] * kernel_shared[((((int)threadIdx.x) * 48) + 10)]));
-      conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[21] * kernel_shared[((((int)threadIdx.x) * 48) + 7)]));
-      conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[30] * kernel_shared[((((int)threadIdx.x) * 48) + 10)]));
-      conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[22] * kernel_shared[((((int)threadIdx.x) * 48) + 7)]));
-      conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[31] * kernel_shared[((((int)threadIdx.x) * 48) + 10)]));
-      conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[23] * kernel_shared[((((int)threadIdx.x) * 48) + 7)]));
-      conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[32] * kernel_shared[((((int)threadIdx.x) * 48) + 10)]));
-      conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[24] * kernel_shared[((((int)threadIdx.x) * 48) + 7)]));
-      conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[33] * kernel_shared[((((int)threadIdx.x) * 48) + 10)]));
-      conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[25] * kernel_shared[((((int)threadIdx.x) * 48) + 7)]));
-      conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[34] * kernel_shared[((((int)threadIdx.x) * 48) + 10)]));
-      conv2d_nchw[7] = (conv2d_nchw[7] + (pad_temp_shared[19] * kernel_shared[((((int)threadIdx.x) * 48) + 31)]));
-      conv2d_nchw[7] = (conv2d_nchw[7] + (pad_temp_shared[28] * kernel_shared[((((int)threadIdx.x) * 48) + 34)]));
-      conv2d_nchw[8] = (conv2d_nchw[8] + (pad_temp_shared[20] * kernel_shared[((((int)threadIdx.x) * 48) + 31)]));
-      conv2d_nchw[8] = (conv2d_nchw[8] + (pad_temp_shared[29] * kernel_shared[((((int)threadIdx.x) * 48) + 34)]));
-      conv2d_nchw[9] = (conv2d_nchw[9] + (pad_temp_shared[21] * kernel_shared[((((int)threadIdx.x) * 48) + 31)]));
-      conv2d_nchw[9] = (conv2d_nchw[9] + (pad_temp_shared[30] * kernel_shared[((((int)threadIdx.x) * 48) + 34)]));
-      conv2d_nchw[10] = (conv2d_nchw[10] + (pad_temp_shared[22] * kernel_shared[((((int)threadIdx.x) * 48) + 31)]));
-      conv2d_nchw[10] = (conv2d_nchw[10] + (pad_temp_shared[31] * kernel_shared[((((int)threadIdx.x) * 48) + 34)]));
-      conv2d_nchw[11] = (conv2d_nchw[11] + (pad_temp_shared[23] * kernel_shared[((((int)threadIdx.x) * 48) + 31)]));
-      conv2d_nchw[11] = (conv2d_nchw[11] + (pad_temp_shared[32] * kernel_shared[((((int)threadIdx.x) * 48) + 34)]));
-      conv2d_nchw[12] = (conv2d_nchw[12] + (pad_temp_shared[24] * kernel_shared[((((int)threadIdx.x) * 48) + 31)]));
-      conv2d_nchw[12] = (conv2d_nchw[12] + (pad_temp_shared[33] * kernel_shared[((((int)threadIdx.x) * 48) + 34)]));
-      conv2d_nchw[13] = (conv2d_nchw[13] + (pad_temp_shared[25] * kernel_shared[((((int)threadIdx.x) * 48) + 31)]));
-      conv2d_nchw[13] = (conv2d_nchw[13] + (pad_temp_shared[34] * kernel_shared[((((int)threadIdx.x) * 48) + 34)]));
-      conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[20] * kernel_shared[((((int)threadIdx.x) * 48) + 8)]));
-      conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[29] * kernel_shared[((((int)threadIdx.x) * 48) + 11)]));
-      conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[21] * kernel_shared[((((int)threadIdx.x) * 48) + 8)]));
-      conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[30] * kernel_shared[((((int)threadIdx.x) * 48) + 11)]));
-      conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[22] * kernel_shared[((((int)threadIdx.x) * 48) + 8)]));
-      conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[31] * kernel_shared[((((int)threadIdx.x) * 48) + 11)]));
-      conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[23] * kernel_shared[((((int)threadIdx.x) * 48) + 8)]));
-      conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[32] * kernel_shared[((((int)threadIdx.x) * 48) + 11)]));
-      conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[24] * kernel_shared[((((int)threadIdx.x) * 48) + 8)]));
-      conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[33] * kernel_shared[((((int)threadIdx.x) * 48) + 11)]));
-      conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[25] * kernel_shared[((((int)threadIdx.x) * 48) + 8)]));
-      conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[34] * kernel_shared[((((int)threadIdx.x) * 48) + 11)]));
-      conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[26] * kernel_shared[((((int)threadIdx.x) * 48) + 8)]));
-      conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[35] * kernel_shared[((((int)threadIdx.x) * 48) + 11)]));
-      conv2d_nchw[7] = (conv2d_nchw[7] + (pad_temp_shared[20] * kernel_shared[((((int)threadIdx.x) * 48) + 32)]));
-      conv2d_nchw[7] = (conv2d_nchw[7] + (pad_temp_shared[29] * kernel_shared[((((int)threadIdx.x) * 48) + 35)]));
-      conv2d_nchw[8] = (conv2d_nchw[8] + (pad_temp_shared[21] * kernel_shared[((((int)threadIdx.x) * 48) + 32)]));
-      conv2d_nchw[8] = (conv2d_nchw[8] + (pad_temp_shared[30] * kernel_shared[((((int)threadIdx.x) * 48) + 35)]));
-      conv2d_nchw[9] = (conv2d_nchw[9] + (pad_temp_shared[22] * kernel_shared[((((int)threadIdx.x) * 48) + 32)]));
-      conv2d_nchw[9] = (conv2d_nchw[9] + (pad_temp_shared[31] * kernel_shared[((((int)threadIdx.x) * 48) + 35)]));
-      conv2d_nchw[10] = (conv2d_nchw[10] + (pad_temp_shared[23] * kernel_shared[((((int)threadIdx.x) * 48) + 32)]));
-      conv2d_nchw[10] = (conv2d_nchw[10] + (pad_temp_shared[32] * kernel_shared[((((int)threadIdx.x) * 48) + 35)]));
-      conv2d_nchw[11] = (conv2d_nchw[11] + (pad_temp_shared[24] * kernel_shared[((((int)threadIdx.x) * 48) + 32)]));
-      conv2d_nchw[11] = (conv2d_nchw[11] + (pad_temp_shared[33] * kernel_shared[((((int)threadIdx.x) * 48) + 35)]));
-      conv2d_nchw[12] = (conv2d_nchw[12] + (pad_temp_shared[25] * kernel_shared[((((int)threadIdx.x) * 48) + 32)]));
-      conv2d_nchw[12] = (conv2d_nchw[12] + (pad_temp_shared[34] * kernel_shared[((((int)threadIdx.x) * 48) + 35)]));
-      conv2d_nchw[13] = (conv2d_nchw[13] + (pad_temp_shared[26] * kernel_shared[((((int)threadIdx.x) * 48) + 32)]));
-      conv2d_nchw[13] = (conv2d_nchw[13] + (pad_temp_shared[35] * kernel_shared[((((int)threadIdx.x) * 48) + 35)]));
-      conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[36] * kernel_shared[((((int)threadIdx.x) * 48) + 12)]));
-      conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[45] * kernel_shared[((((int)threadIdx.x) * 48) + 15)]));
-      conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[37] * kernel_shared[((((int)threadIdx.x) * 48) + 12)]));
-      conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[46] * kernel_shared[((((int)threadIdx.x) * 48) + 15)]));
-      conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[38] * kernel_shared[((((int)threadIdx.x) * 48) + 12)]));
-      conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[47] * kernel_shared[((((int)threadIdx.x) * 48) + 15)]));
-      conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[39] * kernel_shared[((((int)threadIdx.x) * 48) + 12)]));
-      conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[48] * kernel_shared[((((int)threadIdx.x) * 48) + 15)]));
-      conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[40] * kernel_shared[((((int)threadIdx.x) * 48) + 12)]));
-      conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[49] * kernel_shared[((((int)threadIdx.x) * 48) + 15)]));
-      conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[41] * kernel_shared[((((int)threadIdx.x) * 48) + 12)]));
-      conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[50] * kernel_shared[((((int)threadIdx.x) * 48) + 15)]));
-      conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[42] * kernel_shared[((((int)threadIdx.x) * 48) + 12)]));
-      conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[51] * kernel_shared[((((int)threadIdx.x) * 48) + 15)]));
-      conv2d_nchw[7] = (conv2d_nchw[7] + (pad_temp_shared[36] * kernel_shared[((((int)threadIdx.x) * 48) + 36)]));
-      conv2d_nchw[7] = (conv2d_nchw[7] + (pad_temp_shared[45] * kernel_shared[((((int)threadIdx.x) * 48) + 39)]));
-      conv2d_nchw[8] = (conv2d_nchw[8] + (pad_temp_shared[37] * kernel_shared[((((int)threadIdx.x) * 48) + 36)]));
-      conv2d_nchw[8] = (conv2d_nchw[8] + (pad_temp_shared[46] * kernel_shared[((((int)threadIdx.x) * 48) + 39)]));
-      conv2d_nchw[9] = (conv2d_nchw[9] + (pad_temp_shared[38] * kernel_shared[((((int)threadIdx.x) * 48) + 36)]));
-      conv2d_nchw[9] = (conv2d_nchw[9] + (pad_temp_shared[47] * kernel_shared[((((int)threadIdx.x) * 48) + 39)]));
-      conv2d_nchw[10] = (conv2d_nchw[10] + (pad_temp_shared[39] * kernel_shared[((((int)threadIdx.x) * 48) + 36)]));
-      conv2d_nchw[10] = (conv2d_nchw[10] + (pad_temp_shared[48] * kernel_shared[((((int)threadIdx.x) * 48) + 39)]));
-      conv2d_nchw[11] = (conv2d_nchw[11] + (pad_temp_shared[40] * kernel_shared[((((int)threadIdx.x) * 48) + 36)]));
-      conv2d_nchw[11] = (conv2d_nchw[11] + (pad_temp_shared[49] * kernel_shared[((((int)threadIdx.x) * 48) + 39)]));
-      conv2d_nchw[12] = (conv2d_nchw[12] + (pad_temp_shared[41] * kernel_shared[((((int)threadIdx.x) * 48) + 36)]));
-      conv2d_nchw[12] = (conv2d_nchw[12] + (pad_temp_shared[50] * kernel_shared[((((int)threadIdx.x) * 48) + 39)]));
-      conv2d_nchw[13] = (conv2d_nchw[13] + (pad_temp_shared[42] * kernel_shared[((((int)threadIdx.x) * 48) + 36)]));
-      conv2d_nchw[13] = (conv2d_nchw[13] + (pad_temp_shared[51] * kernel_shared[((((int)threadIdx.x) * 48) + 39)]));
-      conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[37] * kernel_shared[((((int)threadIdx.x) * 48) + 13)]));
-      conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[46] * kernel_shared[((((int)threadIdx.x) * 48) + 16)]));
-      conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[38] * kernel_shared[((((int)threadIdx.x) * 48) + 13)]));
-      conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[47] * kernel_shared[((((int)threadIdx.x) * 48) + 16)]));
-      conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[39] * kernel_shared[((((int)threadIdx.x) * 48) + 13)]));
-      conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[48] * kernel_shared[((((int)threadIdx.x) * 48) + 16)]));
-      conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[40] * kernel_shared[((((int)threadIdx.x) * 48) + 13)]));
-      conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[49] * kernel_shared[((((int)threadIdx.x) * 48) + 16)]));
-      conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[41] * kernel_shared[((((int)threadIdx.x) * 48) + 13)]));
-      conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[50] * kernel_shared[((((int)threadIdx.x) * 48) + 16)]));
-      conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[42] * kernel_shared[((((int)threadIdx.x) * 48) + 13)]));
-      conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[51] * kernel_shared[((((int)threadIdx.x) * 48) + 16)]));
-      conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[43] * kernel_shared[((((int)threadIdx.x) * 48) + 13)]));
-      conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[52] * kernel_shared[((((int)threadIdx.x) * 48) + 16)]));
-      conv2d_nchw[7] = (conv2d_nchw[7] + (pad_temp_shared[37] * kernel_shared[((((int)threadIdx.x) * 48) + 37)]));
-      conv2d_nchw[7] = (conv2d_nchw[7] + (pad_temp_shared[46] * kernel_shared[((((int)threadIdx.x) * 48) + 40)]));
-      conv2d_nchw[8] = (conv2d_nchw[8] + (pad_temp_shared[38] * kernel_shared[((((int)threadIdx.x) * 48) + 37)]));
-      conv2d_nchw[8] = (conv2d_nchw[8] + (pad_temp_shared[47] * kernel_shared[((((int)threadIdx.x) * 48) + 40)]));
-      conv2d_nchw[9] = (conv2d_nchw[9] + (pad_temp_shared[39] * kernel_shared[((((int)threadIdx.x) * 48) + 37)]));
-      conv2d_nchw[9] = (conv2d_nchw[9] + (pad_temp_shared[48] * kernel_shared[((((int)threadIdx.x) * 48) + 40)]));
-      conv2d_nchw[10] = (conv2d_nchw[10] + (pad_temp_shared[40] * kernel_shared[((((int)threadIdx.x) * 48) + 37)]));
-      conv2d_nchw[10] = (conv2d_nchw[10] + (pad_temp_shared[49] * kernel_shared[((((int)threadIdx.x) * 48) + 40)]));
-      conv2d_nchw[11] = (conv2d_nchw[11] + (pad_temp_shared[41] * kernel_shared[((((int)threadIdx.x) * 48) + 37)]));
-      conv2d_nchw[11] = (conv2d_nchw[11] + (pad_temp_shared[50] * kernel_shared[((((int)threadIdx.x) * 48) + 40)]));
-      conv2d_nchw[12] = (conv2d_nchw[12] + (pad_temp_shared[42] * kernel_shared[((((int)threadIdx.x) * 48) + 37)]));
-      conv2d_nchw[12] = (conv2d_nchw[12] + (pad_temp_shared[51] * kernel_shared[((((int)threadIdx.x) * 48) + 40)]));
-      conv2d_nchw[13] = (conv2d_nchw[13] + (pad_temp_shared[43] * kernel_shared[((((int)threadIdx.x) * 48) + 37)]));
-      conv2d_nchw[13] = (conv2d_nchw[13] + (pad_temp_shared[52] * kernel_shared[((((int)threadIdx.x) * 48) + 40)]));
-      conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[38] * kernel_shared[((((int)threadIdx.x) * 48) + 14)]));
-      conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[47] * kernel_shared[((((int)threadIdx.x) * 48) + 17)]));
-      conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[39] * kernel_shared[((((int)threadIdx.x) * 48) + 14)]));
-      conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[48] * kernel_shared[((((int)threadIdx.x) * 48) + 17)]));
-      conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[40] * kernel_shared[((((int)threadIdx.x) * 48) + 14)]));
-      conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[49] * kernel_shared[((((int)threadIdx.x) * 48) + 17)]));
-      conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[41] * kernel_shared[((((int)threadIdx.x) * 48) + 14)]));
-      conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[50] * kernel_shared[((((int)threadIdx.x) * 48) + 17)]));
-      conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[42] * kernel_shared[((((int)threadIdx.x) * 48) + 14)]));
-      conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[51] * kernel_shared[((((int)threadIdx.x) * 48) + 17)]));
-      conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[43] * kernel_shared[((((int)threadIdx.x) * 48) + 14)]));
-      conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[52] * kernel_shared[((((int)threadIdx.x) * 48) + 17)]));
-      conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[44] * kernel_shared[((((int)threadIdx.x) * 48) + 14)]));
-      conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[53] * kernel_shared[((((int)threadIdx.x) * 48) + 17)]));
-      conv2d_nchw[7] = (conv2d_nchw[7] + (pad_temp_shared[38] * kernel_shared[((((int)threadIdx.x) * 48) + 38)]));
-      conv2d_nchw[7] = (conv2d_nchw[7] + (pad_temp_shared[47] * kernel_shared[((((int)threadIdx.x) * 48) + 41)]));
-      conv2d_nchw[8] = (conv2d_nchw[8] + (pad_temp_shared[39] * kernel_shared[((((int)threadIdx.x) * 48) + 38)]));
-      conv2d_nchw[8] = (conv2d_nchw[8] + (pad_temp_shared[48] * kernel_shared[((((int)threadIdx.x) * 48) + 41)]));
-      conv2d_nchw[9] = (conv2d_nchw[9] + (pad_temp_shared[40] * kernel_shared[((((int)threadIdx.x) * 48) + 38)]));
-      conv2d_nchw[9] = (conv2d_nchw[9] + (pad_temp_shared[49] * kernel_shared[((((int)threadIdx.x) * 48) + 41)]));
-      conv2d_nchw[10] = (conv2d_nchw[10] + (pad_temp_shared[41] * kernel_shared[((((int)threadIdx.x) * 48) + 38)]));
-      conv2d_nchw[10] = (conv2d_nchw[10] + (pad_temp_shared[50] * kernel_shared[((((int)threadIdx.x) * 48) + 41)]));
-      conv2d_nchw[11] = (conv2d_nchw[11] + (pad_temp_shared[42] * kernel_shared[((((int)threadIdx.x) * 48) + 38)]));
-      conv2d_nchw[11] = (conv2d_nchw[11] + (pad_temp_shared[51] * kernel_shared[((((int)threadIdx.x) * 48) + 41)]));
-      conv2d_nchw[12] = (conv2d_nchw[12] + (pad_temp_shared[43] * kernel_shared[((((int)threadIdx.x) * 48) + 38)]));
-      conv2d_nchw[12] = (conv2d_nchw[12] + (pad_temp_shared[52] * kernel_shared[((((int)threadIdx.x) * 48) + 41)]));
-      conv2d_nchw[13] = (conv2d_nchw[13] + (pad_temp_shared[44] * kernel_shared[((((int)threadIdx.x) * 48) + 38)]));
-      conv2d_nchw[13] = (conv2d_nchw[13] + (pad_temp_shared[53] * kernel_shared[((((int)threadIdx.x) * 48) + 41)]));
-      conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[54] * kernel_shared[((((int)threadIdx.x) * 48) + 18)]));
-      conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[63] * kernel_shared[((((int)threadIdx.x) * 48) + 21)]));
-      conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[55] * kernel_shared[((((int)threadIdx.x) * 48) + 18)]));
-      conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[64] * kernel_shared[((((int)threadIdx.x) * 48) + 21)]));
-      conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[56] * kernel_shared[((((int)threadIdx.x) * 48) + 18)]));
-      conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[65] * kernel_shared[((((int)threadIdx.x) * 48) + 21)]));
-      conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[57] * kernel_shared[((((int)threadIdx.x) * 48) + 18)]));
-      conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[66] * kernel_shared[((((int)threadIdx.x) * 48) + 21)]));
-      conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[58] * kernel_shared[((((int)threadIdx.x) * 48) + 18)]));
-      conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[67] * kernel_shared[((((int)threadIdx.x) * 48) + 21)]));
-      conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[59] * kernel_shared[((((int)threadIdx.x) * 48) + 18)]));
-      conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[68] * kernel_shared[((((int)threadIdx.x) * 48) + 21)]));
-      conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[60] * kernel_shared[((((int)threadIdx.x) * 48) + 18)]));
-      conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[69] * kernel_shared[((((int)threadIdx.x) * 48) + 21)]));
-      conv2d_nchw[7] = (conv2d_nchw[7] + (pad_temp_shared[54] * kernel_shared[((((int)threadIdx.x) * 48) + 42)]));
-      conv2d_nchw[7] = (conv2d_nchw[7] + (pad_temp_shared[63] * kernel_shared[((((int)threadIdx.x) * 48) + 45)]));
-      conv2d_nchw[8] = (conv2d_nchw[8] + (pad_temp_shared[55] * kernel_shared[((((int)threadIdx.x) * 48) + 42)]));
-      conv2d_nchw[8] = (conv2d_nchw[8] + (pad_temp_shared[64] * kernel_shared[((((int)threadIdx.x) * 48) + 45)]));
-      conv2d_nchw[9] = (conv2d_nchw[9] + (pad_temp_shared[56] * kernel_shared[((((int)threadIdx.x) * 48) + 42)]));
-      conv2d_nchw[9] = (conv2d_nchw[9] + (pad_temp_shared[65] * kernel_shared[((((int)threadIdx.x) * 48) + 45)]));
-      conv2d_nchw[10] = (conv2d_nchw[10] + (pad_temp_shared[57] * kernel_shared[((((int)threadIdx.x) * 48) + 42)]));
-      conv2d_nchw[10] = (conv2d_nchw[10] + (pad_temp_shared[66] * kernel_shared[((((int)threadIdx.x) * 48) + 45)]));
-      conv2d_nchw[11] = (conv2d_nchw[11] + (pad_temp_shared[58] * kernel_shared[((((int)threadIdx.x) * 48) + 42)]));
-      conv2d_nchw[11] = (conv2d_nchw[11] + (pad_temp_shared[67] * kernel_shared[((((int)threadIdx.x) * 48) + 45)]));
-      conv2d_nchw[12] = (conv2d_nchw[12] + (pad_temp_shared[59] * kernel_shared[((((int)threadIdx.x) * 48) + 42)]));
-      conv2d_nchw[12] = (conv2d_nchw[12] + (pad_temp_shared[68] * kernel_shared[((((int)threadIdx.x) * 48) + 45)]));
-      conv2d_nchw[13] = (conv2d_nchw[13] + (pad_temp_shared[60] * kernel_shared[((((int)threadIdx.x) * 48) + 42)]));
-      conv2d_nchw[13] = (conv2d_nchw[13] + (pad_temp_shared[69] * kernel_shared[((((int)threadIdx.x) * 48) + 45)]));
-      conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[55] * kernel_shared[((((int)threadIdx.x) * 48) + 19)]));
-      conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[64] * kernel_shared[((((int)threadIdx.x) * 48) + 22)]));
-      conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[56] * kernel_shared[((((int)threadIdx.x) * 48) + 19)]));
-      conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[65] * kernel_shared[((((int)threadIdx.x) * 48) + 22)]));
-      conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[57] * kernel_shared[((((int)threadIdx.x) * 48) + 19)]));
-      conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[66] * kernel_shared[((((int)threadIdx.x) * 48) + 22)]));
-      conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[58] * kernel_shared[((((int)threadIdx.x) * 48) + 19)]));
-      conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[67] * kernel_shared[((((int)threadIdx.x) * 48) + 22)]));
-      conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[59] * kernel_shared[((((int)threadIdx.x) * 48) + 19)]));
-      conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[68] * kernel_shared[((((int)threadIdx.x) * 48) + 22)]));
-      conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[60] * kernel_shared[((((int)threadIdx.x) * 48) + 19)]));
-      conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[69] * kernel_shared[((((int)threadIdx.x) * 48) + 22)]));
-      conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[61] * kernel_shared[((((int)threadIdx.x) * 48) + 19)]));
-      conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[70] * kernel_shared[((((int)threadIdx.x) * 48) + 22)]));
-      conv2d_nchw[7] = (conv2d_nchw[7] + (pad_temp_shared[55] * kernel_shared[((((int)threadIdx.x) * 48) + 43)]));
-      conv2d_nchw[7] = (conv2d_nchw[7] + (pad_temp_shared[64] * kernel_shared[((((int)threadIdx.x) * 48) + 46)]));
-      conv2d_nchw[8] = (conv2d_nchw[8] + (pad_temp_shared[56] * kernel_shared[((((int)threadIdx.x) * 48) + 43)]));
-      conv2d_nchw[8] = (conv2d_nchw[8] + (pad_temp_shared[65] * kernel_shared[((((int)threadIdx.x) * 48) + 46)]));
-      conv2d_nchw[9] = (conv2d_nchw[9] + (pad_temp_shared[57] * kernel_shared[((((int)threadIdx.x) * 48) + 43)]));
-      conv2d_nchw[9] = (conv2d_nchw[9] + (pad_temp_shared[66] * kernel_shared[((((int)threadIdx.x) * 48) + 46)]));
-      conv2d_nchw[10] = (conv2d_nchw[10] + (pad_temp_shared[58] * kernel_shared[((((int)threadIdx.x) * 48) + 43)]));
-      conv2d_nchw[10] = (conv2d_nchw[10] + (pad_temp_shared[67] * kernel_shared[((((int)threadIdx.x) * 48) + 46)]));
-      conv2d_nchw[11] = (conv2d_nchw[11] + (pad_temp_shared[59] * kernel_shared[((((int)threadIdx.x) * 48) + 43)]));
-      conv2d_nchw[11] = (conv2d_nchw[11] + (pad_temp_shared[68] * kernel_shared[((((int)threadIdx.x) * 48) + 46)]));
-      conv2d_nchw[12] = (conv2d_nchw[12] + (pad_temp_shared[60] * kernel_shared[((((int)threadIdx.x) * 48) + 43)]));
-      conv2d_nchw[12] = (conv2d_nchw[12] + (pad_temp_shared[69] * kernel_shared[((((int)threadIdx.x) * 48) + 46)]));
-      conv2d_nchw[13] = (conv2d_nchw[13] + (pad_temp_shared[61] * kernel_shared[((((int)threadIdx.x) * 48) + 43)]));
-      conv2d_nchw[13] = (conv2d_nchw[13] + (pad_temp_shared[70] * kernel_shared[((((int)threadIdx.x) * 48) + 46)]));
-      conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[56] * kernel_shared[((((int)threadIdx.x) * 48) + 20)]));
-      conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[65] * kernel_shared[((((int)threadIdx.x) * 48) + 23)]));
-      conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[57] * kernel_shared[((((int)threadIdx.x) * 48) + 20)]));
-      conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[66] * kernel_shared[((((int)threadIdx.x) * 48) + 23)]));
-      conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[58] * kernel_shared[((((int)threadIdx.x) * 48) + 20)]));
-      conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[67] * kernel_shared[((((int)threadIdx.x) * 48) + 23)]));
-      conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[59] * kernel_shared[((((int)threadIdx.x) * 48) + 20)]));
-      conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[68] * kernel_shared[((((int)threadIdx.x) * 48) + 23)]));
-      conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[60] * kernel_shared[((((int)threadIdx.x) * 48) + 20)]));
-      conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[69] * kernel_shared[((((int)threadIdx.x) * 48) + 23)]));
-      conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[61] * kernel_shared[((((int)threadIdx.x) * 48) + 20)]));
-      conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[70] * kernel_shared[((((int)threadIdx.x) * 48) + 23)]));
-      conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[62] * kernel_shared[((((int)threadIdx.x) * 48) + 20)]));
-      conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[71] * kernel_shared[((((int)threadIdx.x) * 48) + 23)]));
-      conv2d_nchw[7] = (conv2d_nchw[7] + (pad_temp_shared[56] * kernel_shared[((((int)threadIdx.x) * 48) + 44)]));
-      conv2d_nchw[7] = (conv2d_nchw[7] + (pad_temp_shared[65] * kernel_shared[((((int)threadIdx.x) * 48) + 47)]));
-      conv2d_nchw[8] = (conv2d_nchw[8] + (pad_temp_shared[57] * kernel_shared[((((int)threadIdx.x) * 48) + 44)]));
-      conv2d_nchw[8] = (conv2d_nchw[8] + (pad_temp_shared[66] * kernel_shared[((((int)threadIdx.x) * 48) + 47)]));
-      conv2d_nchw[9] = (conv2d_nchw[9] + (pad_temp_shared[58] * kernel_shared[((((int)threadIdx.x) * 48) + 44)]));
-      conv2d_nchw[9] = (conv2d_nchw[9] + (pad_temp_shared[67] * kernel_shared[((((int)threadIdx.x) * 48) + 47)]));
-      conv2d_nchw[10] = (conv2d_nchw[10] + (pad_temp_shared[59] * kernel_shared[((((int)threadIdx.x) * 48) + 44)]));
-      conv2d_nchw[10] = (conv2d_nchw[10] + (pad_temp_shared[68] * kernel_shared[((((int)threadIdx.x) * 48) + 47)]));
-      conv2d_nchw[11] = (conv2d_nchw[11] + (pad_temp_shared[60] * kernel_shared[((((int)threadIdx.x) * 48) + 44)]));
-      conv2d_nchw[11] = (conv2d_nchw[11] + (pad_temp_shared[69] * kernel_shared[((((int)threadIdx.x) * 48) + 47)]));
-      conv2d_nchw[12] = (conv2d_nchw[12] + (pad_temp_shared[61] * kernel_shared[((((int)threadIdx.x) * 48) + 44)]));
-      conv2d_nchw[12] = (conv2d_nchw[12] + (pad_temp_shared[70] * kernel_shared[((((int)threadIdx.x) * 48) + 47)]));
-      conv2d_nchw[13] = (conv2d_nchw[13] + (pad_temp_shared[62] * kernel_shared[((((int)threadIdx.x) * 48) + 44)]));
-      conv2d_nchw[13] = (conv2d_nchw[13] + (pad_temp_shared[71] * kernel_shared[((((int)threadIdx.x) * 48) + 47)]));
-    }
-  }
-  for (int i1_inner = 0; i1_inner &lt; 2; ++i1_inner) {
-    for (int i3_inner = 0; i3_inner &lt; 7; ++i3_inner) {
-      compute[((((((((int)blockIdx.x) / 7) * 6272) + (((int)threadIdx.x) * 98)) + (i1_inner * 49)) + ((((int)blockIdx.x) % 7) * 7)) + i3_inner)] = max((conv2d_nchw[((i1_inner * 7) + i3_inner)] + bias[((((((int)blockIdx.x) / 7) * 128) + (((int)threadIdx.x) * 2)) + i1_inner)]), 0.000000e+00f);
+    __syncthreads();
+    pad_temp_shared[((int)threadIdx.x)] = (((((1 &lt;= (((((int)threadIdx.x) % 27) / 9) + (((int)blockIdx.x) % 7))) &amp;&amp; ((((((int)threadIdx.x) % 27) / 9) + (((int)blockIdx.x) % 7)) &lt; 8)) &amp;&amp; (1 &lt;= (((int)threadIdx.x) % 9))) &amp;&amp; ((((int)threadIdx.x) % 9) &lt; 8)) ? data[((((((rc_outer_outer * 392) + ((((int)threadIdx.x) / 27) * 49)) + (((((int)threadIdx.x) % 27) / 9) * 7)) + ((((int)blockIdx.x) % 7) * 7)) + (((int)threadIdx.x) % 9)) - 8)] : 0.000000e+00f);
+    pad_temp_shared[(((int)threadIdx.x) + 64)] = (((((1 &lt;= ((((((int)threadIdx.x) + 10) % 27) / 9) + (((int)blockIdx.x) % 7))) &amp;&amp; (((((((int)threadIdx.x) + 10) % 27) / 9) + (((int)blockIdx.x) % 7)) &lt; 8)) &amp;&amp; (1 &lt;= ((((int)threadIdx.x) + 1) % 9))) &amp;&amp; (((((int)threadIdx.x) + 1) % 9) &lt; 8)) ? data[((((((rc_outer_outer * 392) + (((((int)threadIdx.x) + 64) / 27) * 49)) + ((((((int)threadIdx.x) + 10) % 27) / 9) * 7)) + ((((int)blockIdx.x) % 7) * 7)) + ((((int) [...]
+    pad_temp_shared[(((int)threadIdx.x) + 128)] = (((((1 &lt;= ((((((int)threadIdx.x) + 20) % 27) / 9) + (((int)blockIdx.x) % 7))) &amp;&amp; (((((((int)threadIdx.x) + 20) % 27) / 9) + (((int)blockIdx.x) % 7)) &lt; 8)) &amp;&amp; (1 &lt;= ((((int)threadIdx.x) + 2) % 9))) &amp;&amp; (((((int)threadIdx.x) + 2) % 9) &lt; 8)) ? data[((((((rc_outer_outer * 392) + (((((int)threadIdx.x) + 128) / 27) * 49)) + ((((((int)threadIdx.x) + 20) % 27) / 9) * 7)) + ((((int)blockIdx.x) % 7) * 7)) + ((((in [...]
+    if (((int)threadIdx.x) &lt; 24) {
+      pad_temp_shared[(((int)threadIdx.x) + 192)] = (((((1 &lt;= (((((int)threadIdx.x) + 3) / 9) + (((int)blockIdx.x) % 7))) &amp;&amp; ((((((int)threadIdx.x) + 3) / 9) + (((int)blockIdx.x) % 7)) &lt; 8)) &amp;&amp; (1 &lt;= ((((int)threadIdx.x) + 3) % 9))) &amp;&amp; (((((int)threadIdx.x) + 3) % 9) &lt; 8)) ? data[((((((rc_outer_outer * 392) + (((((int)threadIdx.x) + 192) / 27) * 49)) + (((((int)threadIdx.x) + 3) / 9) * 7)) + ((((int)blockIdx.x) % 7) * 7)) + ((((int)threadIdx.x) + 3) %  [...]
     }
+    kernel_shared[((int)threadIdx.x)] = kernel[((((((int)blockIdx.x) / 7) * 589824) + (rc_outer_outer * 72)) + ((int)threadIdx.x))];
+    kernel_shared[(((int)threadIdx.x) + 64)] = kernel[((((((((int)blockIdx.x) / 7) * 589824) + (((((int)threadIdx.x) + 64) / 72) * 4608)) + (rc_outer_outer * 72)) + ((((((int)threadIdx.x) + 64) % 72) / 3) * 3)) + ((((int)threadIdx.x) + 1) % 3))];
+    kernel_shared[(((int)threadIdx.x) + 128)] = kernel[((((((((int)blockIdx.x) / 7) * 589824) + (((((int)threadIdx.x) + 128) / 72) * 4608)) + (rc_outer_outer * 72)) + ((((((int)threadIdx.x) + 56) % 72) / 3) * 3)) + ((((int)threadIdx.x) + 2) % 3))];
+    kernel_shared[(((int)threadIdx.x) + 192)] = kernel[((((((((int)blockIdx.x) / 7) * 589824) + (((((int)threadIdx.x) + 192) / 72) * 4608)) + (rc_outer_outer * 72)) + ((((((int)threadIdx.x) / 3) + 16) % 24) * 3)) + (((int)threadIdx.x) % 3))];
+    kernel_shared[(((int)threadIdx.x) + 256)] = kernel[((((((((int)blockIdx.x) / 7) * 589824) + (((((int)threadIdx.x) + 256) / 72) * 4608)) + (rc_outer_outer * 72)) + ((((((int)threadIdx.x) + 40) % 72) / 3) * 3)) + ((((int)threadIdx.x) + 1) % 3))];
+    kernel_shared[(((int)threadIdx.x) + 320)] = kernel[((((((((int)blockIdx.x) / 7) * 589824) + (((((int)threadIdx.x) + 320) / 72) * 4608)) + (rc_outer_outer * 72)) + ((((((int)threadIdx.x) + 32) % 72) / 3) * 3)) + ((((int)threadIdx.x) + 2) % 3))];
+    kernel_shared[(((int)threadIdx.x) + 384)] = kernel[((((((((int)blockIdx.x) / 7) * 589824) + (((((int)threadIdx.x) + 384) / 72) * 4608)) + (rc_outer_outer * 72)) + ((((((int)threadIdx.x) / 3) + 8) % 24) * 3)) + (((int)threadIdx.x) % 3))];
+    kernel_shared[(((int)threadIdx.x) + 448)] = kernel[((((((((int)blockIdx.x) / 7) * 589824) + (((((int)threadIdx.x) + 448) / 72) * 4608)) + (rc_outer_outer * 72)) + ((((((int)threadIdx.x) + 16) % 72) / 3) * 3)) + ((((int)threadIdx.x) + 1) % 3))];
+    kernel_shared[(((int)threadIdx.x) + 512)] = kernel[((((((((int)blockIdx.x) / 7) * 589824) + (((((int)threadIdx.x) + 512) / 72) * 4608)) + (rc_outer_outer * 72)) + (((((int)threadIdx.x) + 8) / 3) * 3)) + ((((int)threadIdx.x) + 2) % 3))];
+    kernel_shared[(((int)threadIdx.x) + 576)] = kernel[(((((((int)blockIdx.x) / 7) * 589824) + (rc_outer_outer * 72)) + ((int)threadIdx.x)) + 36864)];
+    kernel_shared[(((int)threadIdx.x) + 640)] = kernel[((((((((int)blockIdx.x) / 7) * 589824) + (((((int)threadIdx.x) + 640) / 72) * 4608)) + (rc_outer_outer * 72)) + ((((((int)threadIdx.x) + 64) % 72) / 3) * 3)) + ((((int)threadIdx.x) + 1) % 3))];
+    kernel_shared[(((int)threadIdx.x) + 704)] = kernel[((((((((int)blockIdx.x) / 7) * 589824) + (((((int)threadIdx.x) + 704) / 72) * 4608)) + (rc_outer_outer * 72)) + ((((((int)threadIdx.x) + 56) % 72) / 3) * 3)) + ((((int)threadIdx.x) + 2) % 3))];
+    kernel_shared[(((int)threadIdx.x) + 768)] = kernel[((((((((int)blockIdx.x) / 7) * 589824) + (((((int)threadIdx.x) + 768) / 72) * 4608)) + (rc_outer_outer * 72)) + ((((((int)threadIdx.x) / 3) + 16) % 24) * 3)) + (((int)threadIdx.x) % 3))];
+    kernel_shared[(((int)threadIdx.x) + 832)] = kernel[((((((((int)blockIdx.x) / 7) * 589824) + (((((int)threadIdx.x) + 832) / 72) * 4608)) + (rc_outer_outer * 72)) + ((((((int)threadIdx.x) + 40) % 72) / 3) * 3)) + ((((int)threadIdx.x) + 1) % 3))];
+    kernel_shared[(((int)threadIdx.x) + 896)] = kernel[((((((((int)blockIdx.x) / 7) * 589824) + (((((int)threadIdx.x) + 896) / 72) * 4608)) + (rc_outer_outer * 72)) + ((((((int)threadIdx.x) + 32) % 72) / 3) * 3)) + ((((int)threadIdx.x) + 2) % 3))];
+    kernel_shared[(((int)threadIdx.x) + 960)] = kernel[((((((((int)blockIdx.x) / 7) * 589824) + (((((int)threadIdx.x) + 960) / 72) * 4608)) + (rc_outer_outer * 72)) + ((((((int)threadIdx.x) / 3) + 8) % 24) * 3)) + (((int)threadIdx.x) % 3))];
+    kernel_shared[(((int)threadIdx.x) + 1024)] = kernel[((((((((int)blockIdx.x) / 7) * 589824) + (((((int)threadIdx.x) + 1024) / 72) * 4608)) + (rc_outer_outer * 72)) + ((((((int)threadIdx.x) + 16) % 72) / 3) * 3)) + ((((int)threadIdx.x) + 1) % 3))];
+    kernel_shared[(((int)threadIdx.x) + 1088)] = kernel[((((((((int)blockIdx.x) / 7) * 589824) + (((((int)threadIdx.x) + 1088) / 72) * 4608)) + (rc_outer_outer * 72)) + (((((int)threadIdx.x) + 8) / 3) * 3)) + ((((int)threadIdx.x) + 2) % 3))];
+    kernel_shared[(((int)threadIdx.x) + 1152)] = kernel[(((((((int)blockIdx.x) / 7) * 589824) + (rc_outer_outer * 72)) + ((int)threadIdx.x)) + 73728)];
+    kernel_shared[(((int)threadIdx.x) + 1216)] = kernel[((((((((int)blockIdx.x) / 7) * 589824) + (((((int)threadIdx.x) + 1216) / 72) * 4608)) + (rc_outer_outer * 72)) + ((((((int)threadIdx.x) + 64) % 72) / 3) * 3)) + ((((int)threadIdx.x) + 1) % 3))];
+    kernel_shared[(((int)threadIdx.x) + 1280)] = kernel[((((((((int)blockIdx.x) / 7) * 589824) + (((((int)threadIdx.x) + 1280) / 72) * 4608)) + (rc_outer_outer * 72)) + ((((((int)threadIdx.x) + 56) % 72) / 3) * 3)) + ((((int)threadIdx.x) + 2) % 3))];
+    kernel_shared[(((int)threadIdx.x) + 1344)] = kernel[((((((((int)blockIdx.x) / 7) * 589824) + (((((int)threadIdx.x) + 1344) / 72) * 4608)) + (rc_outer_outer * 72)) + ((((((int)threadIdx.x) / 3) + 16) % 24) * 3)) + (((int)threadIdx.x) % 3))];
+    kernel_shared[(((int)threadIdx.x) + 1408)] = kernel[((((((((int)blockIdx.x) / 7) * 589824) + (((((int)threadIdx.x) + 1408) / 72) * 4608)) + (rc_outer_outer * 72)) + ((((((int)threadIdx.x) + 40) % 72) / 3) * 3)) + ((((int)threadIdx.x) + 1) % 3))];
+    kernel_shared[(((int)threadIdx.x) + 1472)] = kernel[((((((((int)blockIdx.x) / 7) * 589824) + (((((int)threadIdx.x) + 1472) / 72) * 4608)) + (rc_outer_outer * 72)) + ((((((int)threadIdx.x) + 32) % 72) / 3) * 3)) + ((((int)threadIdx.x) + 2) % 3))];
+    kernel_shared[(((int)threadIdx.x) + 1536)] = kernel[((((((((int)blockIdx.x) / 7) * 589824) + (((((int)threadIdx.x) + 1536) / 72) * 4608)) + (rc_outer_outer * 72)) + ((((((int)threadIdx.x) / 3) + 8) % 24) * 3)) + (((int)threadIdx.x) % 3))];
+    kernel_shared[(((int)threadIdx.x) + 1600)] = kernel[((((((((int)blockIdx.x) / 7) * 589824) + (((((int)threadIdx.x) + 1600) / 72) * 4608)) + (rc_outer_outer * 72)) + ((((((int)threadIdx.x) + 16) % 72) / 3) * 3)) + ((((int)threadIdx.x) + 1) % 3))];
+    kernel_shared[(((int)threadIdx.x) + 1664)] = kernel[((((((((int)blockIdx.x) / 7) * 589824) + (((((int)threadIdx.x) + 1664) / 72) * 4608)) + (rc_outer_outer * 72)) + (((((int)threadIdx.x) + 8) / 3) * 3)) + ((((int)threadIdx.x) + 2) % 3))];
+    kernel_shared[(((int)threadIdx.x) + 1728)] = kernel[(((((((int)blockIdx.x) / 7) * 589824) + (rc_outer_outer * 72)) + ((int)threadIdx.x)) + 110592)];
+    kernel_shared[(((int)threadIdx.x) + 1792)] = kernel[((((((((int)blockIdx.x) / 7) * 589824) + (((((int)threadIdx.x) + 1792) / 72) * 4608)) + (rc_outer_outer * 72)) + ((((((int)threadIdx.x) + 64) % 72) / 3) * 3)) + ((((int)threadIdx.x) + 1) % 3))];
+    kernel_shared[(((int)threadIdx.x) + 1856)] = kernel[((((((((int)blockIdx.x) / 7) * 589824) + (((((int)threadIdx.x) + 1856) / 72) * 4608)) + (rc_outer_outer * 72)) + ((((((int)threadIdx.x) + 56) % 72) / 3) * 3)) + ((((int)threadIdx.x) + 2) % 3))];
+    kernel_shared[(((int)threadIdx.x) + 1920)] = kernel[((((((((int)blockIdx.x) / 7) * 589824) + (((((int)threadIdx.x) + 1920) / 72) * 4608)) + (rc_outer_outer * 72)) + ((((((int)threadIdx.x) / 3) + 16) % 24) * 3)) + (((int)threadIdx.x) % 3))];
+    kernel_shared[(((int)threadIdx.x) + 1984)] = kernel[((((((((int)blockIdx.x) / 7) * 589824) + (((((int)threadIdx.x) + 1984) / 72) * 4608)) + (rc_outer_outer * 72)) + ((((((int)threadIdx.x) + 40) % 72) / 3) * 3)) + ((((int)threadIdx.x) + 1) % 3))];
+    kernel_shared[(((int)threadIdx.x) + 2048)] = kernel[((((((((int)blockIdx.x) / 7) * 589824) + (((((int)threadIdx.x) + 2048) / 72) * 4608)) + (rc_outer_outer * 72)) + ((((((int)threadIdx.x) + 32) % 72) / 3) * 3)) + ((((int)threadIdx.x) + 2) % 3))];
+    kernel_shared[(((int)threadIdx.x) + 2112)] = kernel[((((((((int)blockIdx.x) / 7) * 589824) + (((((int)threadIdx.x) + 2112) / 72) * 4608)) + (rc_outer_outer * 72)) + ((((((int)threadIdx.x) / 3) + 8) % 24) * 3)) + (((int)threadIdx.x) % 3))];
+    kernel_shared[(((int)threadIdx.x) + 2176)] = kernel[((((((((int)blockIdx.x) / 7) * 589824) + (((((int)threadIdx.x) + 2176) / 72) * 4608)) + (rc_outer_outer * 72)) + ((((((int)threadIdx.x) + 16) % 72) / 3) * 3)) + ((((int)threadIdx.x) + 1) % 3))];
+    kernel_shared[(((int)threadIdx.x) + 2240)] = kernel[((((((((int)blockIdx.x) / 7) * 589824) + (((((int)threadIdx.x) + 2240) / 72) * 4608)) + (rc_outer_outer * 72)) + (((((int)threadIdx.x) + 8) / 3) * 3)) + ((((int)threadIdx.x) + 2) % 3))];
+    kernel_shared[(((int)threadIdx.x) + 2304)] = kernel[(((((((int)blockIdx.x) / 7) * 589824) + (rc_outer_outer * 72)) + ((int)threadIdx.x)) + 147456)];
+    kernel_shared[(((int)threadIdx.x) + 2368)] = kernel[((((((((int)blockIdx.x) / 7) * 589824) + (((((int)threadIdx.x) + 2368) / 72) * 4608)) + (rc_outer_outer * 72)) + ((((((int)threadIdx.x) + 64) % 72) / 3) * 3)) + ((((int)threadIdx.x) + 1) % 3))];
+    kernel_shared[(((int)threadIdx.x) + 2432)] = kernel[((((((((int)blockIdx.x) / 7) * 589824) + (((((int)threadIdx.x) + 2432) / 72) * 4608)) + (rc_outer_outer * 72)) + ((((((int)threadIdx.x) + 56) % 72) / 3) * 3)) + ((((int)threadIdx.x) + 2) % 3))];
+    kernel_shared[(((int)threadIdx.x) + 2496)] = kernel[((((((((int)blockIdx.x) / 7) * 589824) + (((((int)threadIdx.x) + 2496) / 72) * 4608)) + (rc_outer_outer * 72)) + ((((((int)threadIdx.x) / 3) + 16) % 24) * 3)) + (((int)threadIdx.x) % 3))];
+    kernel_shared[(((int)threadIdx.x) + 2560)] = kernel[((((((((int)blockIdx.x) / 7) * 589824) + (((((int)threadIdx.x) + 2560) / 72) * 4608)) + (rc_outer_outer * 72)) + ((((((int)threadIdx.x) + 40) % 72) / 3) * 3)) + ((((int)threadIdx.x) + 1) % 3))];
+    kernel_shared[(((int)threadIdx.x) + 2624)] = kernel[((((((((int)blockIdx.x) / 7) * 589824) + (((((int)threadIdx.x) + 2624) / 72) * 4608)) + (rc_outer_outer * 72)) + ((((((int)threadIdx.x) + 32) % 72) / 3) * 3)) + ((((int)threadIdx.x) + 2) % 3))];
+    kernel_shared[(((int)threadIdx.x) + 2688)] = kernel[((((((((int)blockIdx.x) / 7) * 589824) + (((((int)threadIdx.x) + 2688) / 72) * 4608)) + (rc_outer_outer * 72)) + ((((((int)threadIdx.x) / 3) + 8) % 24) * 3)) + (((int)threadIdx.x) % 3))];
+    kernel_shared[(((int)threadIdx.x) + 2752)] = kernel[((((((((int)blockIdx.x) / 7) * 589824) + (((((int)threadIdx.x) + 2752) / 72) * 4608)) + (rc_outer_outer * 72)) + ((((((int)threadIdx.x) + 16) % 72) / 3) * 3)) + ((((int)threadIdx.x) + 1) % 3))];
+    kernel_shared[(((int)threadIdx.x) + 2816)] = kernel[((((((((int)blockIdx.x) / 7) * 589824) + (((((int)threadIdx.x) + 2816) / 72) * 4608)) + (rc_outer_outer * 72)) + (((((int)threadIdx.x) + 8) / 3) * 3)) + ((((int)threadIdx.x) + 2) % 3))];
+    kernel_shared[(((int)threadIdx.x) + 2880)] = kernel[(((((((int)blockIdx.x) / 7) * 589824) + (rc_outer_outer * 72)) + ((int)threadIdx.x)) + 184320)];
+    kernel_shared[(((int)threadIdx.x) + 2944)] = kernel[((((((((int)blockIdx.x) / 7) * 589824) + (((((int)threadIdx.x) + 2944) / 72) * 4608)) + (rc_outer_outer * 72)) + ((((((int)threadIdx.x) + 64) % 72) / 3) * 3)) + ((((int)threadIdx.x) + 1) % 3))];
+    kernel_shared[(((int)threadIdx.x) + 3008)] = kernel[((((((((int)blockIdx.x) / 7) * 589824) + (((((int)threadIdx.x) + 3008) / 72) * 4608)) + (rc_outer_outer * 72)) + ((((((int)threadIdx.x) + 56) % 72) / 3) * 3)) + ((((int)threadIdx.x) + 2) % 3))];
+    kernel_shared[(((int)threadIdx.x) + 3072)] = kernel[((((((((int)blockIdx.x) / 7) * 589824) + (((((int)threadIdx.x) + 3072) / 72) * 4608)) + (rc_outer_outer * 72)) + ((((((int)threadIdx.x) / 3) + 16) % 24) * 3)) + (((int)threadIdx.x) % 3))];
+    kernel_shared[(((int)threadIdx.x) + 3136)] = kernel[((((((((int)blockIdx.x) / 7) * 589824) + (((((int)threadIdx.x) + 3136) / 72) * 4608)) + (rc_outer_outer * 72)) + ((((((int)threadIdx.x) + 40) % 72) / 3) * 3)) + ((((int)threadIdx.x) + 1) % 3))];
+    kernel_shared[(((int)threadIdx.x) + 3200)] = kernel[((((((((int)blockIdx.x) / 7) * 589824) + (((((int)threadIdx.x) + 3200) / 72) * 4608)) + (rc_outer_outer * 72)) + ((((((int)threadIdx.x) + 32) % 72) / 3) * 3)) + ((((int)threadIdx.x) + 2) % 3))];
+    kernel_shared[(((int)threadIdx.x) + 3264)] = kernel[((((((((int)blockIdx.x) / 7) * 589824) + (((((int)threadIdx.x) + 3264) / 72) * 4608)) + (rc_outer_outer * 72)) + ((((((int)threadIdx.x) / 3) + 8) % 24) * 3)) + (((int)threadIdx.x) % 3))];
+    kernel_shared[(((int)threadIdx.x) + 3328)] = kernel[((((((((int)blockIdx.x) / 7) * 589824) + (((((int)threadIdx.x) + 3328) / 72) * 4608)) + (rc_outer_outer * 72)) + ((((((int)threadIdx.x) + 16) % 72) / 3) * 3)) + ((((int)threadIdx.x) + 1) % 3))];
+    kernel_shared[(((int)threadIdx.x) + 3392)] = kernel[((((((((int)blockIdx.x) / 7) * 589824) + (((((int)threadIdx.x) + 3392) / 72) * 4608)) + (rc_outer_outer * 72)) + (((((int)threadIdx.x) + 8) / 3) * 3)) + ((((int)threadIdx.x) + 2) % 3))];
+    kernel_shared[(((int)threadIdx.x) + 3456)] = kernel[(((((((int)blockIdx.x) / 7) * 589824) + (rc_outer_outer * 72)) + ((int)threadIdx.x)) + 221184)];
+    kernel_shared[(((int)threadIdx.x) + 3520)] = kernel[((((((((int)blockIdx.x) / 7) * 589824) + (((((int)threadIdx.x) + 3520) / 72) * 4608)) + (rc_outer_outer * 72)) + ((((((int)threadIdx.x) + 64) % 72) / 3) * 3)) + ((((int)threadIdx.x) + 1) % 3))];
+    kernel_shared[(((int)threadIdx.x) + 3584)] = kernel[((((((((int)blockIdx.x) / 7) * 589824) + (((((int)threadIdx.x) + 3584) / 72) * 4608)) + (rc_outer_outer * 72)) + ((((((int)threadIdx.x) + 56) % 72) / 3) * 3)) + ((((int)threadIdx.x) + 2) % 3))];
+    kernel_shared[(((int)threadIdx.x) + 3648)] = kernel[((((((((int)blockIdx.x) / 7) * 589824) + (((((int)threadIdx.x) + 3648) / 72) * 4608)) + (rc_outer_outer * 72)) + ((((((int)threadIdx.x) / 3) + 16) % 24) * 3)) + (((int)threadIdx.x) % 3))];
+    kernel_shared[(((int)threadIdx.x) + 3712)] = kernel[((((((((int)blockIdx.x) / 7) * 589824) + (((((int)threadIdx.x) + 3712) / 72) * 4608)) + (rc_outer_outer * 72)) + ((((((int)threadIdx.x) + 40) % 72) / 3) * 3)) + ((((int)threadIdx.x) + 1) % 3))];
+    kernel_shared[(((int)threadIdx.x) + 3776)] = kernel[((((((((int)blockIdx.x) / 7) * 589824) + (((((int)threadIdx.x) + 3776) / 72) * 4608)) + (rc_outer_outer * 72)) + ((((((int)threadIdx.x) + 32) % 72) / 3) * 3)) + ((((int)threadIdx.x) + 2) % 3))];
+    kernel_shared[(((int)threadIdx.x) + 3840)] = kernel[((((((((int)blockIdx.x) / 7) * 589824) + (((((int)threadIdx.x) + 3840) / 72) * 4608)) + (rc_outer_outer * 72)) + ((((((int)threadIdx.x) / 3) + 8) % 24) * 3)) + (((int)threadIdx.x) % 3))];
+    kernel_shared[(((int)threadIdx.x) + 3904)] = kernel[((((((((int)blockIdx.x) / 7) * 589824) + (((((int)threadIdx.x) + 3904) / 72) * 4608)) + (rc_outer_outer * 72)) + ((((((int)threadIdx.x) + 16) % 72) / 3) * 3)) + ((((int)threadIdx.x) + 1) % 3))];
+    kernel_shared[(((int)threadIdx.x) + 3968)] = kernel[((((((((int)blockIdx.x) / 7) * 589824) + (((((int)threadIdx.x) + 3968) / 72) * 4608)) + (rc_outer_outer * 72)) + (((((int)threadIdx.x) + 8) / 3) * 3)) + ((((int)threadIdx.x) + 2) % 3))];
+    kernel_shared[(((int)threadIdx.x) + 4032)] = kernel[(((((((int)blockIdx.x) / 7) * 589824) + (rc_outer_outer * 72)) + ((int)threadIdx.x)) + 258048)];
+    kernel_shared[(((int)threadIdx.x) + 4096)] = kernel[((((((((int)blockIdx.x) / 7) * 589824) + (((((int)threadIdx.x) + 4096) / 72) * 4608)) + (rc_outer_outer * 72)) + ((((((int)threadIdx.x) + 64) % 72) / 3) * 3)) + ((((int)threadIdx.x) + 1) % 3))];
+    kernel_shared[(((int)threadIdx.x) + 4160)] = kernel[((((((((int)blockIdx.x) / 7) * 589824) + (((((int)threadIdx.x) + 4160) / 72) * 4608)) + (rc_outer_outer * 72)) + ((((((int)threadIdx.x) + 56) % 72) / 3) * 3)) + ((((int)threadIdx.x) + 2) % 3))];
+    kernel_shared[(((int)threadIdx.x) + 4224)] = kernel[((((((((int)blockIdx.x) / 7) * 589824) + (((((int)threadIdx.x) + 4224) / 72) * 4608)) + (rc_outer_outer * 72)) + ((((((int)threadIdx.x) / 3) + 16) % 24) * 3)) + (((int)threadIdx.x) % 3))];
+    kernel_shared[(((int)threadIdx.x) + 4288)] = kernel[((((((((int)blockIdx.x) / 7) * 589824) + (((((int)threadIdx.x) + 4288) / 72) * 4608)) + (rc_outer_outer * 72)) + ((((((int)threadIdx.x) + 40) % 72) / 3) * 3)) + ((((int)threadIdx.x) + 1) % 3))];
+    kernel_shared[(((int)threadIdx.x) + 4352)] = kernel[((((((((int)blockIdx.x) / 7) * 589824) + (((((int)threadIdx.x) + 4352) / 72) * 4608)) + (rc_outer_outer * 72)) + ((((((int)threadIdx.x) + 32) % 72) / 3) * 3)) + ((((int)threadIdx.x) + 2) % 3))];
+    kernel_shared[(((int)threadIdx.x) + 4416)] = kernel[((((((((int)blockIdx.x) / 7) * 589824) + (((((int)threadIdx.x) + 4416) / 72) * 4608)) + (rc_outer_outer * 72)) + ((((((int)threadIdx.x) / 3) + 8) % 24) * 3)) + (((int)threadIdx.x) % 3))];
+    kernel_shared[(((int)threadIdx.x) + 4480)] = kernel[((((((((int)blockIdx.x) / 7) * 589824) + (((((int)threadIdx.x) + 4480) / 72) * 4608)) + (rc_outer_outer * 72)) + ((((((int)threadIdx.x) + 16) % 72) / 3) * 3)) + ((((int)threadIdx.x) + 1) % 3))];
+    kernel_shared[(((int)threadIdx.x) + 4544)] = kernel[((((((((int)blockIdx.x) / 7) * 589824) + (((((int)threadIdx.x) + 4544) / 72) * 4608)) + (rc_outer_outer * 72)) + (((((int)threadIdx.x) + 8) / 3) * 3)) + ((((int)threadIdx.x) + 2) % 3))];
+    kernel_shared[(((int)threadIdx.x) + 4608)] = kernel[(((((((int)blockIdx.x) / 7) * 589824) + (rc_outer_outer * 72)) + ((int)threadIdx.x)) + 294912)];
+    kernel_shared[(((int)threadIdx.x) + 4672)] = kernel[((((((((int)blockIdx.x) / 7) * 589824) + (((((int)threadIdx.x) + 4672) / 72) * 4608)) + (rc_outer_outer * 72)) + ((((((int)threadIdx.x) + 64) % 72) / 3) * 3)) + ((((int)threadIdx.x) + 1) % 3))];
+    kernel_shared[(((int)threadIdx.x) + 4736)] = kernel[((((((((int)blockIdx.x) / 7) * 589824) + (((((int)threadIdx.x) + 4736) / 72) * 4608)) + (rc_outer_outer * 72)) + ((((((int)threadIdx.x) + 56) % 72) / 3) * 3)) + ((((int)threadIdx.x) + 2) % 3))];
+    kernel_shared[(((int)threadIdx.x) + 4800)] = kernel[((((((((int)blockIdx.x) / 7) * 589824) + (((((int)threadIdx.x) + 4800) / 72) * 4608)) + (rc_outer_outer * 72)) + ((((((int)threadIdx.x) / 3) + 16) % 24) * 3)) + (((int)threadIdx.x) % 3))];
+    kernel_shared[(((int)threadIdx.x) + 4864)] = kernel[((((((((int)blockIdx.x) / 7) * 589824) + (((((int)threadIdx.x) + 4864) / 72) * 4608)) + (rc_outer_outer * 72)) + ((((((int)threadIdx.x) + 40) % 72) / 3) * 3)) + ((((int)threadIdx.x) + 1) % 3))];
+    kernel_shared[(((int)threadIdx.x) + 4928)] = kernel[((((((((int)blockIdx.x) / 7) * 589824) + (((((int)threadIdx.x) + 4928) / 72) * 4608)) + (rc_outer_outer * 72)) + ((((((int)threadIdx.x) + 32) % 72) / 3) * 3)) + ((((int)threadIdx.x) + 2) % 3))];
+    kernel_shared[(((int)threadIdx.x) + 4992)] = kernel[((((((((int)blockIdx.x) / 7) * 589824) + (((((int)threadIdx.x) + 4992) / 72) * 4608)) + (rc_outer_outer * 72)) + ((((((int)threadIdx.x) / 3) + 8) % 24) * 3)) + (((int)threadIdx.x) % 3))];
+    kernel_shared[(((int)threadIdx.x) + 5056)] = kernel[((((((((int)blockIdx.x) / 7) * 589824) + (((((int)threadIdx.x) + 5056) / 72) * 4608)) + (rc_outer_outer * 72)) + ((((((int)threadIdx.x) + 16) % 72) / 3) * 3)) + ((((int)threadIdx.x) + 1) % 3))];
+    kernel_shared[(((int)threadIdx.x) + 5120)] = kernel[((((((((int)blockIdx.x) / 7) * 589824) + (((((int)threadIdx.x) + 5120) / 72) * 4608)) + (rc_outer_outer * 72)) + (((((int)threadIdx.x) + 8) / 3) * 3)) + ((((int)threadIdx.x) + 2) % 3))];
+    kernel_shared[(((int)threadIdx.x) + 5184)] = kernel[(((((((int)blockIdx.x) / 7) * 589824) + (rc_outer_outer * 72)) + ((int)threadIdx.x)) + 331776)];
+    kernel_shared[(((int)threadIdx.x) + 5248)] = kernel[((((((((int)blockIdx.x) / 7) * 589824) + (((((int)threadIdx.x) + 5248) / 72) * 4608)) + (rc_outer_outer * 72)) + ((((((int)threadIdx.x) + 64) % 72) / 3) * 3)) + ((((int)threadIdx.x) + 1) % 3))];
+    kernel_shared[(((int)threadIdx.x) + 5312)] = kernel[((((((((int)blockIdx.x) / 7) * 589824) + (((((int)threadIdx.x) + 5312) / 72) * 4608)) + (rc_outer_outer * 72)) + ((((((int)threadIdx.x) + 56) % 72) / 3) * 3)) + ((((int)threadIdx.x) + 2) % 3))];
+    kernel_shared[(((int)threadIdx.x) + 5376)] = kernel[((((((((int)blockIdx.x) / 7) * 589824) + (((((int)threadIdx.x) + 5376) / 72) * 4608)) + (rc_outer_outer * 72)) + ((((((int)threadIdx.x) / 3) + 16) % 24) * 3)) + (((int)threadIdx.x) % 3))];
+    kernel_shared[(((int)threadIdx.x) + 5440)] = kernel[((((((((int)blockIdx.x) / 7) * 589824) + (((((int)threadIdx.x) + 5440) / 72) * 4608)) + (rc_outer_outer * 72)) + ((((((int)threadIdx.x) + 40) % 72) / 3) * 3)) + ((((int)threadIdx.x) + 1) % 3))];
+    kernel_shared[(((int)threadIdx.x) + 5504)] = kernel[((((((((int)blockIdx.x) / 7) * 589824) + (((((int)threadIdx.x) + 5504) / 72) * 4608)) + (rc_outer_outer * 72)) + ((((((int)threadIdx.x) + 32) % 72) / 3) * 3)) + ((((int)threadIdx.x) + 2) % 3))];
+    kernel_shared[(((int)threadIdx.x) + 5568)] = kernel[((((((((int)blockIdx.x) / 7) * 589824) + (((((int)threadIdx.x) + 5568) / 72) * 4608)) + (rc_outer_outer * 72)) + ((((((int)threadIdx.x) / 3) + 8) % 24) * 3)) + (((int)threadIdx.x) % 3))];
+    kernel_shared[(((int)threadIdx.x) + 5632)] = kernel[((((((((int)blockIdx.x) / 7) * 589824) + (((((int)threadIdx.x) + 5632) / 72) * 4608)) + (rc_outer_outer * 72)) + ((((((int)threadIdx.x) + 16) % 72) / 3) * 3)) + ((((int)threadIdx.x) + 1) % 3))];
+    kernel_shared[(((int)threadIdx.x) + 5696)] = kernel[((((((((int)blockIdx.x) / 7) * 589824) + (((((int)threadIdx.x) + 5696) / 72) * 4608)) + (rc_outer_outer * 72)) + (((((int)threadIdx.x) + 8) / 3) * 3)) + ((((int)threadIdx.x) + 2) % 3))];
+    kernel_shared[(((int)threadIdx.x) + 5760)] = kernel[(((((((int)blockIdx.x) / 7) * 589824) + (rc_outer_outer * 72)) + ((int)threadIdx.x)) + 368640)];
+    kernel_shared[(((int)threadIdx.x) + 5824)] = kernel[((((((((int)blockIdx.x) / 7) * 589824) + (((((int)threadIdx.x) + 5824) / 72) * 4608)) + (rc_outer_outer * 72)) + ((((((int)threadIdx.x) + 64) % 72) / 3) * 3)) + ((((int)threadIdx.x) + 1) % 3))];
+    kernel_shared[(((int)threadIdx.x) + 5888)] = kernel[((((((((int)blockIdx.x) / 7) * 589824) + (((((int)threadIdx.x) + 5888) / 72) * 4608)) + (rc_outer_outer * 72)) + ((((((int)threadIdx.x) + 56) % 72) / 3) * 3)) + ((((int)threadIdx.x) + 2) % 3))];
+    kernel_shared[(((int)threadIdx.x) + 5952)] = kernel[((((((((int)blockIdx.x) / 7) * 589824) + (((((int)threadIdx.x) + 5952) / 72) * 4608)) + (rc_outer_outer * 72)) + ((((((int)threadIdx.x) / 3) + 16) % 24) * 3)) + (((int)threadIdx.x) % 3))];
+    kernel_shared[(((int)threadIdx.x) + 6016)] = kernel[((((((((int)blockIdx.x) / 7) * 589824) + (((((int)threadIdx.x) + 6016) / 72) * 4608)) + (rc_outer_outer * 72)) + ((((((int)threadIdx.x) + 40) % 72) / 3) * 3)) + ((((int)threadIdx.x) + 1) % 3))];
+    kernel_shared[(((int)threadIdx.x) + 6080)] = kernel[((((((((int)blockIdx.x) / 7) * 589824) + (((((int)threadIdx.x) + 6080) / 72) * 4608)) + (rc_outer_outer * 72)) + ((((((int)threadIdx.x) + 32) % 72) / 3) * 3)) + ((((int)threadIdx.x) + 2) % 3))];
+    kernel_shared[(((int)threadIdx.x) + 6144)] = kernel[((((((((int)blockIdx.x) / 7) * 589824) + (((((int)threadIdx.x) + 6144) / 72) * 4608)) + (rc_outer_outer * 72)) + ((((((int)threadIdx.x) / 3) + 8) % 24) * 3)) + (((int)threadIdx.x) % 3))];
+    kernel_shared[(((int)threadIdx.x) + 6208)] = kernel[((((((((int)blockIdx.x) / 7) * 589824) + (((((int)threadIdx.x) + 6208) / 72) * 4608)) + (rc_outer_outer * 72)) + ((((((int)threadIdx.x) + 16) % 72) / 3) * 3)) + ((((int)threadIdx.x) + 1) % 3))];
+    kernel_shared[(((int)threadIdx.x) + 6272)] = kernel[((((((((int)blockIdx.x) / 7) * 589824) + (((((int)threadIdx.x) + 6272) / 72) * 4608)) + (rc_outer_outer * 72)) + (((((int)threadIdx.x) + 8) / 3) * 3)) + ((((int)threadIdx.x) + 2) % 3))];
+    kernel_shared[(((int)threadIdx.x) + 6336)] = kernel[(((((((int)blockIdx.x) / 7) * 589824) + (rc_outer_outer * 72)) + ((int)threadIdx.x)) + 405504)];
+    kernel_shared[(((int)threadIdx.x) + 6400)] = kernel[((((((((int)blockIdx.x) / 7) * 589824) + (((((int)threadIdx.x) + 6400) / 72) * 4608)) + (rc_outer_outer * 72)) + ((((((int)threadIdx.x) + 64) % 72) / 3) * 3)) + ((((int)threadIdx.x) + 1) % 3))];
+    kernel_shared[(((int)threadIdx.x) + 6464)] = kernel[((((((((int)blockIdx.x) / 7) * 589824) + (((((int)threadIdx.x) + 6464) / 72) * 4608)) + (rc_outer_outer * 72)) + ((((((int)threadIdx.x) + 56) % 72) / 3) * 3)) + ((((int)threadIdx.x) + 2) % 3))];
+    kernel_shared[(((int)threadIdx.x) + 6528)] = kernel[((((((((int)blockIdx.x) / 7) * 589824) + (((((int)threadIdx.x) + 6528) / 72) * 4608)) + (rc_outer_outer * 72)) + ((((((int)threadIdx.x) / 3) + 16) % 24) * 3)) + (((int)threadIdx.x) % 3))];
+    kernel_shared[(((int)threadIdx.x) + 6592)] = kernel[((((((((int)blockIdx.x) / 7) * 589824) + (((((int)threadIdx.x) + 6592) / 72) * 4608)) + (rc_outer_outer * 72)) + ((((((int)threadIdx.x) + 40) % 72) / 3) * 3)) + ((((int)threadIdx.x) + 1) % 3))];
+    kernel_shared[(((int)threadIdx.x) + 6656)] = kernel[((((((((int)blockIdx.x) / 7) * 589824) + (((((int)threadIdx.x) + 6656) / 72) * 4608)) + (rc_outer_outer * 72)) + ((((((int)threadIdx.x) + 32) % 72) / 3) * 3)) + ((((int)threadIdx.x) + 2) % 3))];
+    kernel_shared[(((int)threadIdx.x) + 6720)] = kernel[((((((((int)blockIdx.x) / 7) * 589824) + (((((int)threadIdx.x) + 6720) / 72) * 4608)) + (rc_outer_outer * 72)) + ((((((int)threadIdx.x) / 3) + 8) % 24) * 3)) + (((int)threadIdx.x) % 3))];
+    kernel_shared[(((int)threadIdx.x) + 6784)] = kernel[((((((((int)blockIdx.x) / 7) * 589824) + (((((int)threadIdx.x) + 6784) / 72) * 4608)) + (rc_outer_outer * 72)) + ((((((int)threadIdx.x) + 16) % 72) / 3) * 3)) + ((((int)threadIdx.x) + 1) % 3))];
+    kernel_shared[(((int)threadIdx.x) + 6848)] = kernel[((((((((int)blockIdx.x) / 7) * 589824) + (((((int)threadIdx.x) + 6848) / 72) * 4608)) + (rc_outer_outer * 72)) + (((((int)threadIdx.x) + 8) / 3) * 3)) + ((((int)threadIdx.x) + 2) % 3))];
+    kernel_shared[(((int)threadIdx.x) + 6912)] = kernel[(((((((int)blockIdx.x) / 7) * 589824) + (rc_outer_outer * 72)) + ((int)threadIdx.x)) + 442368)];
+    kernel_shared[(((int)threadIdx.x) + 6976)] = kernel[((((((((int)blockIdx.x) / 7) * 589824) + (((((int)threadIdx.x) + 6976) / 72) * 4608)) + (rc_outer_outer * 72)) + ((((((int)threadIdx.x) + 64) % 72) / 3) * 3)) + ((((int)threadIdx.x) + 1) % 3))];
+    kernel_shared[(((int)threadIdx.x) + 7040)] = kernel[((((((((int)blockIdx.x) / 7) * 589824) + (((((int)threadIdx.x) + 7040) / 72) * 4608)) + (rc_outer_outer * 72)) + ((((((int)threadIdx.x) + 56) % 72) / 3) * 3)) + ((((int)threadIdx.x) + 2) % 3))];
+    kernel_shared[(((int)threadIdx.x) + 7104)] = kernel[((((((((int)blockIdx.x) / 7) * 589824) + (((((int)threadIdx.x) + 7104) / 72) * 4608)) + (rc_outer_outer * 72)) + ((((((int)threadIdx.x) / 3) + 16) % 24) * 3)) + (((int)threadIdx.x) % 3))];
+    kernel_shared[(((int)threadIdx.x) + 7168)] = kernel[((((((((int)blockIdx.x) / 7) * 589824) + (((((int)threadIdx.x) + 7168) / 72) * 4608)) + (rc_outer_outer * 72)) + ((((((int)threadIdx.x) + 40) % 72) / 3) * 3)) + ((((int)threadIdx.x) + 1) % 3))];
+    kernel_shared[(((int)threadIdx.x) + 7232)] = kernel[((((((((int)blockIdx.x) / 7) * 589824) + (((((int)threadIdx.x) + 7232) / 72) * 4608)) + (rc_outer_outer * 72)) + ((((((int)threadIdx.x) + 32) % 72) / 3) * 3)) + ((((int)threadIdx.x) + 2) % 3))];
+    kernel_shared[(((int)threadIdx.x) + 7296)] = kernel[((((((((int)blockIdx.x) / 7) * 589824) + (((((int)threadIdx.x) + 7296) / 72) * 4608)) + (rc_outer_outer * 72)) + ((((((int)threadIdx.x) / 3) + 8) % 24) * 3)) + (((int)threadIdx.x) % 3))];
+    kernel_shared[(((int)threadIdx.x) + 7360)] = kernel[((((((((int)blockIdx.x) / 7) * 589824) + (((((int)threadIdx.x) + 7360) / 72) * 4608)) + (rc_outer_outer * 72)) + ((((((int)threadIdx.x) + 16) % 72) / 3) * 3)) + ((((int)threadIdx.x) + 1) % 3))];
+    kernel_shared[(((int)threadIdx.x) + 7424)] = kernel[((((((((int)blockIdx.x) / 7) * 589824) + (((((int)threadIdx.x) + 7424) / 72) * 4608)) + (rc_outer_outer * 72)) + (((((int)threadIdx.x) + 8) / 3) * 3)) + ((((int)threadIdx.x) + 2) % 3))];
+    kernel_shared[(((int)threadIdx.x) + 7488)] = kernel[(((((((int)blockIdx.x) / 7) * 589824) + (rc_outer_outer * 72)) + ((int)threadIdx.x)) + 479232)];
+    kernel_shared[(((int)threadIdx.x) + 7552)] = kernel[((((((((int)blockIdx.x) / 7) * 589824) + (((((int)threadIdx.x) + 7552) / 72) * 4608)) + (rc_outer_outer * 72)) + ((((((int)threadIdx.x) + 64) % 72) / 3) * 3)) + ((((int)threadIdx.x) + 1) % 3))];
+    kernel_shared[(((int)threadIdx.x) + 7616)] = kernel[((((((((int)blockIdx.x) / 7) * 589824) + (((((int)threadIdx.x) + 7616) / 72) * 4608)) + (rc_outer_outer * 72)) + ((((((int)threadIdx.x) + 56) % 72) / 3) * 3)) + ((((int)threadIdx.x) + 2) % 3))];
+    kernel_shared[(((int)threadIdx.x) + 7680)] = kernel[((((((((int)blockIdx.x) / 7) * 589824) + (((((int)threadIdx.x) + 7680) / 72) * 4608)) + (rc_outer_outer * 72)) + ((((((int)threadIdx.x) / 3) + 16) % 24) * 3)) + (((int)threadIdx.x) % 3))];
+    kernel_shared[(((int)threadIdx.x) + 7744)] = kernel[((((((((int)blockIdx.x) / 7) * 589824) + (((((int)threadIdx.x) + 7744) / 72) * 4608)) + (rc_outer_outer * 72)) + ((((((int)threadIdx.x) + 40) % 72) / 3) * 3)) + ((((int)threadIdx.x) + 1) % 3))];
+    kernel_shared[(((int)threadIdx.x) + 7808)] = kernel[((((((((int)blockIdx.x) / 7) * 589824) + (((((int)threadIdx.x) + 7808) / 72) * 4608)) + (rc_outer_outer * 72)) + ((((((int)threadIdx.x) + 32) % 72) / 3) * 3)) + ((((int)threadIdx.x) + 2) % 3))];
+    kernel_shared[(((int)threadIdx.x) + 7872)] = kernel[((((((((int)blockIdx.x) / 7) * 589824) + (((((int)threadIdx.x) + 7872) / 72) * 4608)) + (rc_outer_outer * 72)) + ((((((int)threadIdx.x) / 3) + 8) % 24) * 3)) + (((int)threadIdx.x) % 3))];
+    kernel_shared[(((int)threadIdx.x) + 7936)] = kernel[((((((((int)blockIdx.x) / 7) * 589824) + (((((int)threadIdx.x) + 7936) / 72) * 4608)) + (rc_outer_outer * 72)) + ((((((int)threadIdx.x) + 16) % 72) / 3) * 3)) + ((((int)threadIdx.x) + 1) % 3))];
+    kernel_shared[(((int)threadIdx.x) + 8000)] = kernel[((((((((int)blockIdx.x) / 7) * 589824) + (((((int)threadIdx.x) + 8000) / 72) * 4608)) + (rc_outer_outer * 72)) + (((((int)threadIdx.x) + 8) / 3) * 3)) + ((((int)threadIdx.x) + 2) % 3))];
+    kernel_shared[(((int)threadIdx.x) + 8064)] = kernel[(((((((int)blockIdx.x) / 7) * 589824) + (rc_outer_outer * 72)) + ((int)threadIdx.x)) + 516096)];
+    kernel_shared[(((int)threadIdx.x) + 8128)] = kernel[((((((((int)blockIdx.x) / 7) * 589824) + (((((int)threadIdx.x) + 8128) / 72) * 4608)) + (rc_outer_outer * 72)) + ((((((int)threadIdx.x) + 64) % 72) / 3) * 3)) + ((((int)threadIdx.x) + 1) % 3))];
+    kernel_shared[(((int)threadIdx.x) + 8192)] = kernel[((((((((int)blockIdx.x) / 7) * 589824) + (((((int)threadIdx.x) + 8192) / 72) * 4608)) + (rc_outer_outer * 72)) + ((((((int)threadIdx.x) + 56) % 72) / 3) * 3)) + ((((int)threadIdx.x) + 2) % 3))];
+    kernel_shared[(((int)threadIdx.x) + 8256)] = kernel[((((((((int)blockIdx.x) / 7) * 589824) + (((((int)threadIdx.x) + 8256) / 72) * 4608)) + (rc_outer_outer * 72)) + ((((((int)threadIdx.x) / 3) + 16) % 24) * 3)) + (((int)threadIdx.x) % 3))];
+    kernel_shared[(((int)threadIdx.x) + 8320)] = kernel[((((((((int)blockIdx.x) / 7) * 589824) + (((((int)threadIdx.x) + 8320) / 72) * 4608)) + (rc_outer_outer * 72)) + ((((((int)threadIdx.x) + 40) % 72) / 3) * 3)) + ((((int)threadIdx.x) + 1) % 3))];
+    kernel_shared[(((int)threadIdx.x) + 8384)] = kernel[((((((((int)blockIdx.x) / 7) * 589824) + (((((int)threadIdx.x) + 8384) / 72) * 4608)) + (rc_outer_outer * 72)) + ((((((int)threadIdx.x) + 32) % 72) / 3) * 3)) + ((((int)threadIdx.x) + 2) % 3))];
+    kernel_shared[(((int)threadIdx.x) + 8448)] = kernel[((((((((int)blockIdx.x) / 7) * 589824) + (((((int)threadIdx.x) + 8448) / 72) * 4608)) + (rc_outer_outer * 72)) + ((((((int)threadIdx.x) / 3) + 8) % 24) * 3)) + (((int)threadIdx.x) % 3))];
+    kernel_shared[(((int)threadIdx.x) + 8512)] = kernel[((((((((int)blockIdx.x) / 7) * 589824) + (((((int)threadIdx.x) + 8512) / 72) * 4608)) + (rc_outer_outer * 72)) + ((((((int)threadIdx.x) + 16) % 72) / 3) * 3)) + ((((int)threadIdx.x) + 1) % 3))];
+    kernel_shared[(((int)threadIdx.x) + 8576)] = kernel[((((((((int)blockIdx.x) / 7) * 589824) + (((((int)threadIdx.x) + 8576) / 72) * 4608)) + (rc_outer_outer * 72)) + (((((int)threadIdx.x) + 8) / 3) * 3)) + ((((int)threadIdx.x) + 2) % 3))];
+    kernel_shared[(((int)threadIdx.x) + 8640)] = kernel[(((((((int)blockIdx.x) / 7) * 589824) + (rc_outer_outer * 72)) + ((int)threadIdx.x)) + 552960)];
+    kernel_shared[(((int)threadIdx.x) + 8704)] = kernel[((((((((int)blockIdx.x) / 7) * 589824) + (((((int)threadIdx.x) + 8704) / 72) * 4608)) + (rc_outer_outer * 72)) + ((((((int)threadIdx.x) + 64) % 72) / 3) * 3)) + ((((int)threadIdx.x) + 1) % 3))];
+    kernel_shared[(((int)threadIdx.x) + 8768)] = kernel[((((((((int)blockIdx.x) / 7) * 589824) + (((((int)threadIdx.x) + 8768) / 72) * 4608)) + (rc_outer_outer * 72)) + ((((((int)threadIdx.x) + 56) % 72) / 3) * 3)) + ((((int)threadIdx.x) + 2) % 3))];
+    kernel_shared[(((int)threadIdx.x) + 8832)] = kernel[((((((((int)blockIdx.x) / 7) * 589824) + (((((int)threadIdx.x) + 8832) / 72) * 4608)) + (rc_outer_outer * 72)) + ((((((int)threadIdx.x) / 3) + 16) % 24) * 3)) + (((int)threadIdx.x) % 3))];
+    kernel_shared[(((int)threadIdx.x) + 8896)] = kernel[((((((((int)blockIdx.x) / 7) * 589824) + (((((int)threadIdx.x) + 8896) / 72) * 4608)) + (rc_outer_outer * 72)) + ((((((int)threadIdx.x) + 40) % 72) / 3) * 3)) + ((((int)threadIdx.x) + 1) % 3))];
+    kernel_shared[(((int)threadIdx.x) + 8960)] = kernel[((((((((int)blockIdx.x) / 7) * 589824) + (((((int)threadIdx.x) + 8960) / 72) * 4608)) + (rc_outer_outer * 72)) + ((((((int)threadIdx.x) + 32) % 72) / 3) * 3)) + ((((int)threadIdx.x) + 2) % 3))];
+    kernel_shared[(((int)threadIdx.x) + 9024)] = kernel[((((((((int)blockIdx.x) / 7) * 589824) + (((((int)threadIdx.x) + 9024) / 72) * 4608)) + (rc_outer_outer * 72)) + ((((((int)threadIdx.x) / 3) + 8) % 24) * 3)) + (((int)threadIdx.x) % 3))];
+    kernel_shared[(((int)threadIdx.x) + 9088)] = kernel[((((((((int)blockIdx.x) / 7) * 589824) + (((((int)threadIdx.x) + 9088) / 72) * 4608)) + (rc_outer_outer * 72)) + ((((((int)threadIdx.x) + 16) % 72) / 3) * 3)) + ((((int)threadIdx.x) + 1) % 3))];
+    kernel_shared[(((int)threadIdx.x) + 9152)] = kernel[((((((((int)blockIdx.x) / 7) * 589824) + (((((int)threadIdx.x) + 9152) / 72) * 4608)) + (rc_outer_outer * 72)) + (((((int)threadIdx.x) + 8) / 3) * 3)) + ((((int)threadIdx.x) + 2) % 3))];
+    __syncthreads();
+    conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[0] * kernel_shared[(((int)threadIdx.x) * 72)]));
+    conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[1] * kernel_shared[(((int)threadIdx.x) * 72)]));
+    conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[2] * kernel_shared[(((int)threadIdx.x) * 72)]));
+    conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[3] * kernel_shared[(((int)threadIdx.x) * 72)]));
+    conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[4] * kernel_shared[(((int)threadIdx.x) * 72)]));
+    conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[5] * kernel_shared[(((int)threadIdx.x) * 72)]));
+    conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[6] * kernel_shared[(((int)threadIdx.x) * 72)]));
+    conv2d_nchw[7] = (conv2d_nchw[7] + (pad_temp_shared[0] * kernel_shared[((((int)threadIdx.x) * 72) + 4608)]));
+    conv2d_nchw[8] = (conv2d_nchw[8] + (pad_temp_shared[1] * kernel_shared[((((int)threadIdx.x) * 72) + 4608)]));
+    conv2d_nchw[9] = (conv2d_nchw[9] + (pad_temp_shared[2] * kernel_shared[((((int)threadIdx.x) * 72) + 4608)]));
+    conv2d_nchw[10] = (conv2d_nchw[10] + (pad_temp_shared[3] * kernel_shared[((((int)threadIdx.x) * 72) + 4608)]));
+    conv2d_nchw[11] = (conv2d_nchw[11] + (pad_temp_shared[4] * kernel_shared[((((int)threadIdx.x) * 72) + 4608)]));
+    conv2d_nchw[12] = (conv2d_nchw[12] + (pad_temp_shared[5] * kernel_shared[((((int)threadIdx.x) * 72) + 4608)]));
+    conv2d_nchw[13] = (conv2d_nchw[13] + (pad_temp_shared[6] * kernel_shared[((((int)threadIdx.x) * 72) + 4608)]));
+    conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[1] * kernel_shared[((((int)threadIdx.x) * 72) + 1)]));
+    conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[2] * kernel_shared[((((int)threadIdx.x) * 72) + 1)]));
+    conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[3] * kernel_shared[((((int)threadIdx.x) * 72) + 1)]));
+    conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[4] * kernel_shared[((((int)threadIdx.x) * 72) + 1)]));
+    conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[5] * kernel_shared[((((int)threadIdx.x) * 72) + 1)]));
+    conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[6] * kernel_shared[((((int)threadIdx.x) * 72) + 1)]));
+    conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[7] * kernel_shared[((((int)threadIdx.x) * 72) + 1)]));
+    conv2d_nchw[7] = (conv2d_nchw[7] + (pad_temp_shared[1] * kernel_shared[((((int)threadIdx.x) * 72) + 4609)]));
+    conv2d_nchw[8] = (conv2d_nchw[8] + (pad_temp_shared[2] * kernel_shared[((((int)threadIdx.x) * 72) + 4609)]));
+    conv2d_nchw[9] = (conv2d_nchw[9] + (pad_temp_shared[3] * kernel_shared[((((int)threadIdx.x) * 72) + 4609)]));
+    conv2d_nchw[10] = (conv2d_nchw[10] + (pad_temp_shared[4] * kernel_shared[((((int)threadIdx.x) * 72) + 4609)]));
+    conv2d_nchw[11] = (conv2d_nchw[11] + (pad_temp_shared[5] * kernel_shared[((((int)threadIdx.x) * 72) + 4609)]));
+    conv2d_nchw[12] = (conv2d_nchw[12] + (pad_temp_shared[6] * kernel_shared[((((int)threadIdx.x) * 72) + 4609)]));
+    conv2d_nchw[13] = (conv2d_nchw[13] + (pad_temp_shared[7] * kernel_shared[((((int)threadIdx.x) * 72) + 4609)]));
+    conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[2] * kernel_shared[((((int)threadIdx.x) * 72) + 2)]));
+    conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[3] * kernel_shared[((((int)threadIdx.x) * 72) + 2)]));
+    conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[4] * kernel_shared[((((int)threadIdx.x) * 72) + 2)]));
+    conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[5] * kernel_shared[((((int)threadIdx.x) * 72) + 2)]));
+    conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[6] * kernel_shared[((((int)threadIdx.x) * 72) + 2)]));
+    conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[7] * kernel_shared[((((int)threadIdx.x) * 72) + 2)]));
+    conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[8] * kernel_shared[((((int)threadIdx.x) * 72) + 2)]));
+    conv2d_nchw[7] = (conv2d_nchw[7] + (pad_temp_shared[2] * kernel_shared[((((int)threadIdx.x) * 72) + 4610)]));
+    conv2d_nchw[8] = (conv2d_nchw[8] + (pad_temp_shared[3] * kernel_shared[((((int)threadIdx.x) * 72) + 4610)]));
+    conv2d_nchw[9] = (conv2d_nchw[9] + (pad_temp_shared[4] * kernel_shared[((((int)threadIdx.x) * 72) + 4610)]));
+    conv2d_nchw[10] = (conv2d_nchw[10] + (pad_temp_shared[5] * kernel_shared[((((int)threadIdx.x) * 72) + 4610)]));
+    conv2d_nchw[11] = (conv2d_nchw[11] + (pad_temp_shared[6] * kernel_shared[((((int)threadIdx.x) * 72) + 4610)]));
+    conv2d_nchw[12] = (conv2d_nchw[12] + (pad_temp_shared[7] * kernel_shared[((((int)threadIdx.x) * 72) + 4610)]));
+    conv2d_nchw[13] = (conv2d_nchw[13] + (pad_temp_shared[8] * kernel_shared[((((int)threadIdx.x) * 72) + 4610)]));
+    conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[9] * kernel_shared[((((int)threadIdx.x) * 72) + 3)]));
+    conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[10] * kernel_shared[((((int)threadIdx.x) * 72) + 3)]));
+    conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[11] * kernel_shared[((((int)threadIdx.x) * 72) + 3)]));
+    conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[12] * kernel_shared[((((int)threadIdx.x) * 72) + 3)]));
+    conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[13] * kernel_shared[((((int)threadIdx.x) * 72) + 3)]));
+    conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[14] * kernel_shared[((((int)threadIdx.x) * 72) + 3)]));
+    conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[15] * kernel_shared[((((int)threadIdx.x) * 72) + 3)]));
+    conv2d_nchw[7] = (conv2d_nchw[7] + (pad_temp_shared[9] * kernel_shared[((((int)threadIdx.x) * 72) + 4611)]));
+    conv2d_nchw[8] = (conv2d_nchw[8] + (pad_temp_shared[10] * kernel_shared[((((int)threadIdx.x) * 72) + 4611)]));
+    conv2d_nchw[9] = (conv2d_nchw[9] + (pad_temp_shared[11] * kernel_shared[((((int)threadIdx.x) * 72) + 4611)]));
+    conv2d_nchw[10] = (conv2d_nchw[10] + (pad_temp_shared[12] * kernel_shared[((((int)threadIdx.x) * 72) + 4611)]));
+    conv2d_nchw[11] = (conv2d_nchw[11] + (pad_temp_shared[13] * kernel_shared[((((int)threadIdx.x) * 72) + 4611)]));
+    conv2d_nchw[12] = (conv2d_nchw[12] + (pad_temp_shared[14] * kernel_shared[((((int)threadIdx.x) * 72) + 4611)]));
+    conv2d_nchw[13] = (conv2d_nchw[13] + (pad_temp_shared[15] * kernel_shared[((((int)threadIdx.x) * 72) + 4611)]));
+    conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[10] * kernel_shared[((((int)threadIdx.x) * 72) + 4)]));
+    conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[11] * kernel_shared[((((int)threadIdx.x) * 72) + 4)]));
+    conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[12] * kernel_shared[((((int)threadIdx.x) * 72) + 4)]));
+    conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[13] * kernel_shared[((((int)threadIdx.x) * 72) + 4)]));
+    conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[14] * kernel_shared[((((int)threadIdx.x) * 72) + 4)]));
+    conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[15] * kernel_shared[((((int)threadIdx.x) * 72) + 4)]));
+    conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[16] * kernel_shared[((((int)threadIdx.x) * 72) + 4)]));
+    conv2d_nchw[7] = (conv2d_nchw[7] + (pad_temp_shared[10] * kernel_shared[((((int)threadIdx.x) * 72) + 4612)]));
+    conv2d_nchw[8] = (conv2d_nchw[8] + (pad_temp_shared[11] * kernel_shared[((((int)threadIdx.x) * 72) + 4612)]));
+    conv2d_nchw[9] = (conv2d_nchw[9] + (pad_temp_shared[12] * kernel_shared[((((int)threadIdx.x) * 72) + 4612)]));
+    conv2d_nchw[10] = (conv2d_nchw[10] + (pad_temp_shared[13] * kernel_shared[((((int)threadIdx.x) * 72) + 4612)]));
+    conv2d_nchw[11] = (conv2d_nchw[11] + (pad_temp_shared[14] * kernel_shared[((((int)threadIdx.x) * 72) + 4612)]));
+    conv2d_nchw[12] = (conv2d_nchw[12] + (pad_temp_shared[15] * kernel_shared[((((int)threadIdx.x) * 72) + 4612)]));
+    conv2d_nchw[13] = (conv2d_nchw[13] + (pad_temp_shared[16] * kernel_shared[((((int)threadIdx.x) * 72) + 4612)]));
+    conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[11] * kernel_shared[((((int)threadIdx.x) * 72) + 5)]));
+    conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[12] * kernel_shared[((((int)threadIdx.x) * 72) + 5)]));
+    conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[13] * kernel_shared[((((int)threadIdx.x) * 72) + 5)]));
+    conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[14] * kernel_shared[((((int)threadIdx.x) * 72) + 5)]));
+    conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[15] * kernel_shared[((((int)threadIdx.x) * 72) + 5)]));
+    conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[16] * kernel_shared[((((int)threadIdx.x) * 72) + 5)]));
+    conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[17] * kernel_shared[((((int)threadIdx.x) * 72) + 5)]));
+    conv2d_nchw[7] = (conv2d_nchw[7] + (pad_temp_shared[11] * kernel_shared[((((int)threadIdx.x) * 72) + 4613)]));
+    conv2d_nchw[8] = (conv2d_nchw[8] + (pad_temp_shared[12] * kernel_shared[((((int)threadIdx.x) * 72) + 4613)]));
+    conv2d_nchw[9] = (conv2d_nchw[9] + (pad_temp_shared[13] * kernel_shared[((((int)threadIdx.x) * 72) + 4613)]));
+    conv2d_nchw[10] = (conv2d_nchw[10] + (pad_temp_shared[14] * kernel_shared[((((int)threadIdx.x) * 72) + 4613)]));
+    conv2d_nchw[11] = (conv2d_nchw[11] + (pad_temp_shared[15] * kernel_shared[((((int)threadIdx.x) * 72) + 4613)]));
+    conv2d_nchw[12] = (conv2d_nchw[12] + (pad_temp_shared[16] * kernel_shared[((((int)threadIdx.x) * 72) + 4613)]));
+    conv2d_nchw[13] = (conv2d_nchw[13] + (pad_temp_shared[17] * kernel_shared[((((int)threadIdx.x) * 72) + 4613)]));
+    conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[18] * kernel_shared[((((int)threadIdx.x) * 72) + 6)]));
+    conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[19] * kernel_shared[((((int)threadIdx.x) * 72) + 6)]));
+    conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[20] * kernel_shared[((((int)threadIdx.x) * 72) + 6)]));
+    conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[21] * kernel_shared[((((int)threadIdx.x) * 72) + 6)]));
+    conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[22] * kernel_shared[((((int)threadIdx.x) * 72) + 6)]));
+    conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[23] * kernel_shared[((((int)threadIdx.x) * 72) + 6)]));
+    conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[24] * kernel_shared[((((int)threadIdx.x) * 72) + 6)]));
+    conv2d_nchw[7] = (conv2d_nchw[7] + (pad_temp_shared[18] * kernel_shared[((((int)threadIdx.x) * 72) + 4614)]));
+    conv2d_nchw[8] = (conv2d_nchw[8] + (pad_temp_shared[19] * kernel_shared[((((int)threadIdx.x) * 72) + 4614)]));
+    conv2d_nchw[9] = (conv2d_nchw[9] + (pad_temp_shared[20] * kernel_shared[((((int)threadIdx.x) * 72) + 4614)]));
+    conv2d_nchw[10] = (conv2d_nchw[10] + (pad_temp_shared[21] * kernel_shared[((((int)threadIdx.x) * 72) + 4614)]));
+    conv2d_nchw[11] = (conv2d_nchw[11] + (pad_temp_shared[22] * kernel_shared[((((int)threadIdx.x) * 72) + 4614)]));
+    conv2d_nchw[12] = (conv2d_nchw[12] + (pad_temp_shared[23] * kernel_shared[((((int)threadIdx.x) * 72) + 4614)]));
+    conv2d_nchw[13] = (conv2d_nchw[13] + (pad_temp_shared[24] * kernel_shared[((((int)threadIdx.x) * 72) + 4614)]));
+    conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[19] * kernel_shared[((((int)threadIdx.x) * 72) + 7)]));
+    conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[20] * kernel_shared[((((int)threadIdx.x) * 72) + 7)]));
+    conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[21] * kernel_shared[((((int)threadIdx.x) * 72) + 7)]));
+    conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[22] * kernel_shared[((((int)threadIdx.x) * 72) + 7)]));
+    conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[23] * kernel_shared[((((int)threadIdx.x) * 72) + 7)]));
+    conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[24] * kernel_shared[((((int)threadIdx.x) * 72) + 7)]));
+    conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[25] * kernel_shared[((((int)threadIdx.x) * 72) + 7)]));
+    conv2d_nchw[7] = (conv2d_nchw[7] + (pad_temp_shared[19] * kernel_shared[((((int)threadIdx.x) * 72) + 4615)]));
+    conv2d_nchw[8] = (conv2d_nchw[8] + (pad_temp_shared[20] * kernel_shared[((((int)threadIdx.x) * 72) + 4615)]));
+    conv2d_nchw[9] = (conv2d_nchw[9] + (pad_temp_shared[21] * kernel_shared[((((int)threadIdx.x) * 72) + 4615)]));
+    conv2d_nchw[10] = (conv2d_nchw[10] + (pad_temp_shared[22] * kernel_shared[((((int)threadIdx.x) * 72) + 4615)]));
+    conv2d_nchw[11] = (conv2d_nchw[11] + (pad_temp_shared[23] * kernel_shared[((((int)threadIdx.x) * 72) + 4615)]));
+    conv2d_nchw[12] = (conv2d_nchw[12] + (pad_temp_shared[24] * kernel_shared[((((int)threadIdx.x) * 72) + 4615)]));
+    conv2d_nchw[13] = (conv2d_nchw[13] + (pad_temp_shared[25] * kernel_shared[((((int)threadIdx.x) * 72) + 4615)]));
+    conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[20] * kernel_shared[((((int)threadIdx.x) * 72) + 8)]));
+    conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[21] * kernel_shared[((((int)threadIdx.x) * 72) + 8)]));
+    conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[22] * kernel_shared[((((int)threadIdx.x) * 72) + 8)]));
+    conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[23] * kernel_shared[((((int)threadIdx.x) * 72) + 8)]));
+    conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[24] * kernel_shared[((((int)threadIdx.x) * 72) + 8)]));
+    conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[25] * kernel_shared[((((int)threadIdx.x) * 72) + 8)]));
+    conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[26] * kernel_shared[((((int)threadIdx.x) * 72) + 8)]));
+    conv2d_nchw[7] = (conv2d_nchw[7] + (pad_temp_shared[20] * kernel_shared[((((int)threadIdx.x) * 72) + 4616)]));
+    conv2d_nchw[8] = (conv2d_nchw[8] + (pad_temp_shared[21] * kernel_shared[((((int)threadIdx.x) * 72) + 4616)]));
+    conv2d_nchw[9] = (conv2d_nchw[9] + (pad_temp_shared[22] * kernel_shared[((((int)threadIdx.x) * 72) + 4616)]));
+    conv2d_nchw[10] = (conv2d_nchw[10] + (pad_temp_shared[23] * kernel_shared[((((int)threadIdx.x) * 72) + 4616)]));
+    conv2d_nchw[11] = (conv2d_nchw[11] + (pad_temp_shared[24] * kernel_shared[((((int)threadIdx.x) * 72) + 4616)]));
+    conv2d_nchw[12] = (conv2d_nchw[12] + (pad_temp_shared[25] * kernel_shared[((((int)threadIdx.x) * 72) + 4616)]));
+    conv2d_nchw[13] = (conv2d_nchw[13] + (pad_temp_shared[26] * kernel_shared[((((int)threadIdx.x) * 72) + 4616)]));
+    conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[27] * kernel_shared[((((int)threadIdx.x) * 72) + 9)]));
+    conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[28] * kernel_shared[((((int)threadIdx.x) * 72) + 9)]));
+    conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[29] * kernel_shared[((((int)threadIdx.x) * 72) + 9)]));
+    conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[30] * kernel_shared[((((int)threadIdx.x) * 72) + 9)]));
+    conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[31] * kernel_shared[((((int)threadIdx.x) * 72) + 9)]));
+    conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[32] * kernel_shared[((((int)threadIdx.x) * 72) + 9)]));
+    conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[33] * kernel_shared[((((int)threadIdx.x) * 72) + 9)]));
+    conv2d_nchw[7] = (conv2d_nchw[7] + (pad_temp_shared[27] * kernel_shared[((((int)threadIdx.x) * 72) + 4617)]));
+    conv2d_nchw[8] = (conv2d_nchw[8] + (pad_temp_shared[28] * kernel_shared[((((int)threadIdx.x) * 72) + 4617)]));
+    conv2d_nchw[9] = (conv2d_nchw[9] + (pad_temp_shared[29] * kernel_shared[((((int)threadIdx.x) * 72) + 4617)]));
+    conv2d_nchw[10] = (conv2d_nchw[10] + (pad_temp_shared[30] * kernel_shared[((((int)threadIdx.x) * 72) + 4617)]));
+    conv2d_nchw[11] = (conv2d_nchw[11] + (pad_temp_shared[31] * kernel_shared[((((int)threadIdx.x) * 72) + 4617)]));
+    conv2d_nchw[12] = (conv2d_nchw[12] + (pad_temp_shared[32] * kernel_shared[((((int)threadIdx.x) * 72) + 4617)]));
+    conv2d_nchw[13] = (conv2d_nchw[13] + (pad_temp_shared[33] * kernel_shared[((((int)threadIdx.x) * 72) + 4617)]));
+    conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[28] * kernel_shared[((((int)threadIdx.x) * 72) + 10)]));
+    conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[29] * kernel_shared[((((int)threadIdx.x) * 72) + 10)]));
+    conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[30] * kernel_shared[((((int)threadIdx.x) * 72) + 10)]));
+    conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[31] * kernel_shared[((((int)threadIdx.x) * 72) + 10)]));
+    conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[32] * kernel_shared[((((int)threadIdx.x) * 72) + 10)]));
+    conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[33] * kernel_shared[((((int)threadIdx.x) * 72) + 10)]));
+    conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[34] * kernel_shared[((((int)threadIdx.x) * 72) + 10)]));
+    conv2d_nchw[7] = (conv2d_nchw[7] + (pad_temp_shared[28] * kernel_shared[((((int)threadIdx.x) * 72) + 4618)]));
+    conv2d_nchw[8] = (conv2d_nchw[8] + (pad_temp_shared[29] * kernel_shared[((((int)threadIdx.x) * 72) + 4618)]));
+    conv2d_nchw[9] = (conv2d_nchw[9] + (pad_temp_shared[30] * kernel_shared[((((int)threadIdx.x) * 72) + 4618)]));
+    conv2d_nchw[10] = (conv2d_nchw[10] + (pad_temp_shared[31] * kernel_shared[((((int)threadIdx.x) * 72) + 4618)]));
+    conv2d_nchw[11] = (conv2d_nchw[11] + (pad_temp_shared[32] * kernel_shared[((((int)threadIdx.x) * 72) + 4618)]));
+    conv2d_nchw[12] = (conv2d_nchw[12] + (pad_temp_shared[33] * kernel_shared[((((int)threadIdx.x) * 72) + 4618)]));
+    conv2d_nchw[13] = (conv2d_nchw[13] + (pad_temp_shared[34] * kernel_shared[((((int)threadIdx.x) * 72) + 4618)]));
+    conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[29] * kernel_shared[((((int)threadIdx.x) * 72) + 11)]));
+    conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[30] * kernel_shared[((((int)threadIdx.x) * 72) + 11)]));
+    conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[31] * kernel_shared[((((int)threadIdx.x) * 72) + 11)]));
+    conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[32] * kernel_shared[((((int)threadIdx.x) * 72) + 11)]));
+    conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[33] * kernel_shared[((((int)threadIdx.x) * 72) + 11)]));
+    conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[34] * kernel_shared[((((int)threadIdx.x) * 72) + 11)]));
+    conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[35] * kernel_shared[((((int)threadIdx.x) * 72) + 11)]));
+    conv2d_nchw[7] = (conv2d_nchw[7] + (pad_temp_shared[29] * kernel_shared[((((int)threadIdx.x) * 72) + 4619)]));
+    conv2d_nchw[8] = (conv2d_nchw[8] + (pad_temp_shared[30] * kernel_shared[((((int)threadIdx.x) * 72) + 4619)]));
+    conv2d_nchw[9] = (conv2d_nchw[9] + (pad_temp_shared[31] * kernel_shared[((((int)threadIdx.x) * 72) + 4619)]));
+    conv2d_nchw[10] = (conv2d_nchw[10] + (pad_temp_shared[32] * kernel_shared[((((int)threadIdx.x) * 72) + 4619)]));
+    conv2d_nchw[11] = (conv2d_nchw[11] + (pad_temp_shared[33] * kernel_shared[((((int)threadIdx.x) * 72) + 4619)]));
+    conv2d_nchw[12] = (conv2d_nchw[12] + (pad_temp_shared[34] * kernel_shared[((((int)threadIdx.x) * 72) + 4619)]));
+    conv2d_nchw[13] = (conv2d_nchw[13] + (pad_temp_shared[35] * kernel_shared[((((int)threadIdx.x) * 72) + 4619)]));
+    conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[36] * kernel_shared[((((int)threadIdx.x) * 72) + 12)]));
+    conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[37] * kernel_shared[((((int)threadIdx.x) * 72) + 12)]));
+    conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[38] * kernel_shared[((((int)threadIdx.x) * 72) + 12)]));
+    conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[39] * kernel_shared[((((int)threadIdx.x) * 72) + 12)]));
+    conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[40] * kernel_shared[((((int)threadIdx.x) * 72) + 12)]));
+    conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[41] * kernel_shared[((((int)threadIdx.x) * 72) + 12)]));
+    conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[42] * kernel_shared[((((int)threadIdx.x) * 72) + 12)]));
+    conv2d_nchw[7] = (conv2d_nchw[7] + (pad_temp_shared[36] * kernel_shared[((((int)threadIdx.x) * 72) + 4620)]));
+    conv2d_nchw[8] = (conv2d_nchw[8] + (pad_temp_shared[37] * kernel_shared[((((int)threadIdx.x) * 72) + 4620)]));
+    conv2d_nchw[9] = (conv2d_nchw[9] + (pad_temp_shared[38] * kernel_shared[((((int)threadIdx.x) * 72) + 4620)]));
+    conv2d_nchw[10] = (conv2d_nchw[10] + (pad_temp_shared[39] * kernel_shared[((((int)threadIdx.x) * 72) + 4620)]));
+    conv2d_nchw[11] = (conv2d_nchw[11] + (pad_temp_shared[40] * kernel_shared[((((int)threadIdx.x) * 72) + 4620)]));
+    conv2d_nchw[12] = (conv2d_nchw[12] + (pad_temp_shared[41] * kernel_shared[((((int)threadIdx.x) * 72) + 4620)]));
+    conv2d_nchw[13] = (conv2d_nchw[13] + (pad_temp_shared[42] * kernel_shared[((((int)threadIdx.x) * 72) + 4620)]));
+    conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[37] * kernel_shared[((((int)threadIdx.x) * 72) + 13)]));
+    conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[38] * kernel_shared[((((int)threadIdx.x) * 72) + 13)]));
+    conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[39] * kernel_shared[((((int)threadIdx.x) * 72) + 13)]));
+    conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[40] * kernel_shared[((((int)threadIdx.x) * 72) + 13)]));
+    conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[41] * kernel_shared[((((int)threadIdx.x) * 72) + 13)]));
+    conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[42] * kernel_shared[((((int)threadIdx.x) * 72) + 13)]));
+    conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[43] * kernel_shared[((((int)threadIdx.x) * 72) + 13)]));
+    conv2d_nchw[7] = (conv2d_nchw[7] + (pad_temp_shared[37] * kernel_shared[((((int)threadIdx.x) * 72) + 4621)]));
+    conv2d_nchw[8] = (conv2d_nchw[8] + (pad_temp_shared[38] * kernel_shared[((((int)threadIdx.x) * 72) + 4621)]));
+    conv2d_nchw[9] = (conv2d_nchw[9] + (pad_temp_shared[39] * kernel_shared[((((int)threadIdx.x) * 72) + 4621)]));
+    conv2d_nchw[10] = (conv2d_nchw[10] + (pad_temp_shared[40] * kernel_shared[((((int)threadIdx.x) * 72) + 4621)]));
+    conv2d_nchw[11] = (conv2d_nchw[11] + (pad_temp_shared[41] * kernel_shared[((((int)threadIdx.x) * 72) + 4621)]));
+    conv2d_nchw[12] = (conv2d_nchw[12] + (pad_temp_shared[42] * kernel_shared[((((int)threadIdx.x) * 72) + 4621)]));
+    conv2d_nchw[13] = (conv2d_nchw[13] + (pad_temp_shared[43] * kernel_shared[((((int)threadIdx.x) * 72) + 4621)]));
+    conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[38] * kernel_shared[((((int)threadIdx.x) * 72) + 14)]));
+    conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[39] * kernel_shared[((((int)threadIdx.x) * 72) + 14)]));
+    conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[40] * kernel_shared[((((int)threadIdx.x) * 72) + 14)]));
+    conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[41] * kernel_shared[((((int)threadIdx.x) * 72) + 14)]));
+    conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[42] * kernel_shared[((((int)threadIdx.x) * 72) + 14)]));
+    conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[43] * kernel_shared[((((int)threadIdx.x) * 72) + 14)]));
+    conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[44] * kernel_shared[((((int)threadIdx.x) * 72) + 14)]));
+    conv2d_nchw[7] = (conv2d_nchw[7] + (pad_temp_shared[38] * kernel_shared[((((int)threadIdx.x) * 72) + 4622)]));
+    conv2d_nchw[8] = (conv2d_nchw[8] + (pad_temp_shared[39] * kernel_shared[((((int)threadIdx.x) * 72) + 4622)]));
+    conv2d_nchw[9] = (conv2d_nchw[9] + (pad_temp_shared[40] * kernel_shared[((((int)threadIdx.x) * 72) + 4622)]));
+    conv2d_nchw[10] = (conv2d_nchw[10] + (pad_temp_shared[41] * kernel_shared[((((int)threadIdx.x) * 72) + 4622)]));
+    conv2d_nchw[11] = (conv2d_nchw[11] + (pad_temp_shared[42] * kernel_shared[((((int)threadIdx.x) * 72) + 4622)]));
+    conv2d_nchw[12] = (conv2d_nchw[12] + (pad_temp_shared[43] * kernel_shared[((((int)threadIdx.x) * 72) + 4622)]));
+    conv2d_nchw[13] = (conv2d_nchw[13] + (pad_temp_shared[44] * kernel_shared[((((int)threadIdx.x) * 72) + 4622)]));
+    conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[45] * kernel_shared[((((int)threadIdx.x) * 72) + 15)]));
+    conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[46] * kernel_shared[((((int)threadIdx.x) * 72) + 15)]));
+    conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[47] * kernel_shared[((((int)threadIdx.x) * 72) + 15)]));
+    conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[48] * kernel_shared[((((int)threadIdx.x) * 72) + 15)]));
+    conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[49] * kernel_shared[((((int)threadIdx.x) * 72) + 15)]));
+    conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[50] * kernel_shared[((((int)threadIdx.x) * 72) + 15)]));
+    conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[51] * kernel_shared[((((int)threadIdx.x) * 72) + 15)]));
+    conv2d_nchw[7] = (conv2d_nchw[7] + (pad_temp_shared[45] * kernel_shared[((((int)threadIdx.x) * 72) + 4623)]));
+    conv2d_nchw[8] = (conv2d_nchw[8] + (pad_temp_shared[46] * kernel_shared[((((int)threadIdx.x) * 72) + 4623)]));
+    conv2d_nchw[9] = (conv2d_nchw[9] + (pad_temp_shared[47] * kernel_shared[((((int)threadIdx.x) * 72) + 4623)]));
+    conv2d_nchw[10] = (conv2d_nchw[10] + (pad_temp_shared[48] * kernel_shared[((((int)threadIdx.x) * 72) + 4623)]));
+    conv2d_nchw[11] = (conv2d_nchw[11] + (pad_temp_shared[49] * kernel_shared[((((int)threadIdx.x) * 72) + 4623)]));
+    conv2d_nchw[12] = (conv2d_nchw[12] + (pad_temp_shared[50] * kernel_shared[((((int)threadIdx.x) * 72) + 4623)]));
+    conv2d_nchw[13] = (conv2d_nchw[13] + (pad_temp_shared[51] * kernel_shared[((((int)threadIdx.x) * 72) + 4623)]));
+    conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[46] * kernel_shared[((((int)threadIdx.x) * 72) + 16)]));
+    conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[47] * kernel_shared[((((int)threadIdx.x) * 72) + 16)]));
+    conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[48] * kernel_shared[((((int)threadIdx.x) * 72) + 16)]));
+    conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[49] * kernel_shared[((((int)threadIdx.x) * 72) + 16)]));
+    conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[50] * kernel_shared[((((int)threadIdx.x) * 72) + 16)]));
+    conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[51] * kernel_shared[((((int)threadIdx.x) * 72) + 16)]));
+    conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[52] * kernel_shared[((((int)threadIdx.x) * 72) + 16)]));
+    conv2d_nchw[7] = (conv2d_nchw[7] + (pad_temp_shared[46] * kernel_shared[((((int)threadIdx.x) * 72) + 4624)]));
+    conv2d_nchw[8] = (conv2d_nchw[8] + (pad_temp_shared[47] * kernel_shared[((((int)threadIdx.x) * 72) + 4624)]));
+    conv2d_nchw[9] = (conv2d_nchw[9] + (pad_temp_shared[48] * kernel_shared[((((int)threadIdx.x) * 72) + 4624)]));
+    conv2d_nchw[10] = (conv2d_nchw[10] + (pad_temp_shared[49] * kernel_shared[((((int)threadIdx.x) * 72) + 4624)]));
+    conv2d_nchw[11] = (conv2d_nchw[11] + (pad_temp_shared[50] * kernel_shared[((((int)threadIdx.x) * 72) + 4624)]));
+    conv2d_nchw[12] = (conv2d_nchw[12] + (pad_temp_shared[51] * kernel_shared[((((int)threadIdx.x) * 72) + 4624)]));
+    conv2d_nchw[13] = (conv2d_nchw[13] + (pad_temp_shared[52] * kernel_shared[((((int)threadIdx.x) * 72) + 4624)]));
+    conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[47] * kernel_shared[((((int)threadIdx.x) * 72) + 17)]));
+    conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[48] * kernel_shared[((((int)threadIdx.x) * 72) + 17)]));
+    conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[49] * kernel_shared[((((int)threadIdx.x) * 72) + 17)]));
+    conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[50] * kernel_shared[((((int)threadIdx.x) * 72) + 17)]));
+    conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[51] * kernel_shared[((((int)threadIdx.x) * 72) + 17)]));
+    conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[52] * kernel_shared[((((int)threadIdx.x) * 72) + 17)]));
+    conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[53] * kernel_shared[((((int)threadIdx.x) * 72) + 17)]));
+    conv2d_nchw[7] = (conv2d_nchw[7] + (pad_temp_shared[47] * kernel_shared[((((int)threadIdx.x) * 72) + 4625)]));
+    conv2d_nchw[8] = (conv2d_nchw[8] + (pad_temp_shared[48] * kernel_shared[((((int)threadIdx.x) * 72) + 4625)]));
+    conv2d_nchw[9] = (conv2d_nchw[9] + (pad_temp_shared[49] * kernel_shared[((((int)threadIdx.x) * 72) + 4625)]));
+    conv2d_nchw[10] = (conv2d_nchw[10] + (pad_temp_shared[50] * kernel_shared[((((int)threadIdx.x) * 72) + 4625)]));
+    conv2d_nchw[11] = (conv2d_nchw[11] + (pad_temp_shared[51] * kernel_shared[((((int)threadIdx.x) * 72) + 4625)]));
+    conv2d_nchw[12] = (conv2d_nchw[12] + (pad_temp_shared[52] * kernel_shared[((((int)threadIdx.x) * 72) + 4625)]));
+    conv2d_nchw[13] = (conv2d_nchw[13] + (pad_temp_shared[53] * kernel_shared[((((int)threadIdx.x) * 72) + 4625)]));
+    conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[54] * kernel_shared[((((int)threadIdx.x) * 72) + 18)]));
+    conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[55] * kernel_shared[((((int)threadIdx.x) * 72) + 18)]));
+    conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[56] * kernel_shared[((((int)threadIdx.x) * 72) + 18)]));
+    conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[57] * kernel_shared[((((int)threadIdx.x) * 72) + 18)]));
+    conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[58] * kernel_shared[((((int)threadIdx.x) * 72) + 18)]));
+    conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[59] * kernel_shared[((((int)threadIdx.x) * 72) + 18)]));
+    conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[60] * kernel_shared[((((int)threadIdx.x) * 72) + 18)]));
+    conv2d_nchw[7] = (conv2d_nchw[7] + (pad_temp_shared[54] * kernel_shared[((((int)threadIdx.x) * 72) + 4626)]));
+    conv2d_nchw[8] = (conv2d_nchw[8] + (pad_temp_shared[55] * kernel_shared[((((int)threadIdx.x) * 72) + 4626)]));
+    conv2d_nchw[9] = (conv2d_nchw[9] + (pad_temp_shared[56] * kernel_shared[((((int)threadIdx.x) * 72) + 4626)]));
+    conv2d_nchw[10] = (conv2d_nchw[10] + (pad_temp_shared[57] * kernel_shared[((((int)threadIdx.x) * 72) + 4626)]));
+    conv2d_nchw[11] = (conv2d_nchw[11] + (pad_temp_shared[58] * kernel_shared[((((int)threadIdx.x) * 72) + 4626)]));
+    conv2d_nchw[12] = (conv2d_nchw[12] + (pad_temp_shared[59] * kernel_shared[((((int)threadIdx.x) * 72) + 4626)]));
+    conv2d_nchw[13] = (conv2d_nchw[13] + (pad_temp_shared[60] * kernel_shared[((((int)threadIdx.x) * 72) + 4626)]));
+    conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[55] * kernel_shared[((((int)threadIdx.x) * 72) + 19)]));
+    conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[56] * kernel_shared[((((int)threadIdx.x) * 72) + 19)]));
+    conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[57] * kernel_shared[((((int)threadIdx.x) * 72) + 19)]));
+    conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[58] * kernel_shared[((((int)threadIdx.x) * 72) + 19)]));
+    conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[59] * kernel_shared[((((int)threadIdx.x) * 72) + 19)]));
+    conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[60] * kernel_shared[((((int)threadIdx.x) * 72) + 19)]));
+    conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[61] * kernel_shared[((((int)threadIdx.x) * 72) + 19)]));
+    conv2d_nchw[7] = (conv2d_nchw[7] + (pad_temp_shared[55] * kernel_shared[((((int)threadIdx.x) * 72) + 4627)]));
+    conv2d_nchw[8] = (conv2d_nchw[8] + (pad_temp_shared[56] * kernel_shared[((((int)threadIdx.x) * 72) + 4627)]));
+    conv2d_nchw[9] = (conv2d_nchw[9] + (pad_temp_shared[57] * kernel_shared[((((int)threadIdx.x) * 72) + 4627)]));
+    conv2d_nchw[10] = (conv2d_nchw[10] + (pad_temp_shared[58] * kernel_shared[((((int)threadIdx.x) * 72) + 4627)]));
+    conv2d_nchw[11] = (conv2d_nchw[11] + (pad_temp_shared[59] * kernel_shared[((((int)threadIdx.x) * 72) + 4627)]));
+    conv2d_nchw[12] = (conv2d_nchw[12] + (pad_temp_shared[60] * kernel_shared[((((int)threadIdx.x) * 72) + 4627)]));
+    conv2d_nchw[13] = (conv2d_nchw[13] + (pad_temp_shared[61] * kernel_shared[((((int)threadIdx.x) * 72) + 4627)]));
+    conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[56] * kernel_shared[((((int)threadIdx.x) * 72) + 20)]));
+    conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[57] * kernel_shared[((((int)threadIdx.x) * 72) + 20)]));
+    conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[58] * kernel_shared[((((int)threadIdx.x) * 72) + 20)]));
+    conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[59] * kernel_shared[((((int)threadIdx.x) * 72) + 20)]));
+    conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[60] * kernel_shared[((((int)threadIdx.x) * 72) + 20)]));
+    conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[61] * kernel_shared[((((int)threadIdx.x) * 72) + 20)]));
+    conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[62] * kernel_shared[((((int)threadIdx.x) * 72) + 20)]));
+    conv2d_nchw[7] = (conv2d_nchw[7] + (pad_temp_shared[56] * kernel_shared[((((int)threadIdx.x) * 72) + 4628)]));
+    conv2d_nchw[8] = (conv2d_nchw[8] + (pad_temp_shared[57] * kernel_shared[((((int)threadIdx.x) * 72) + 4628)]));
+    conv2d_nchw[9] = (conv2d_nchw[9] + (pad_temp_shared[58] * kernel_shared[((((int)threadIdx.x) * 72) + 4628)]));
+    conv2d_nchw[10] = (conv2d_nchw[10] + (pad_temp_shared[59] * kernel_shared[((((int)threadIdx.x) * 72) + 4628)]));
+    conv2d_nchw[11] = (conv2d_nchw[11] + (pad_temp_shared[60] * kernel_shared[((((int)threadIdx.x) * 72) + 4628)]));
+    conv2d_nchw[12] = (conv2d_nchw[12] + (pad_temp_shared[61] * kernel_shared[((((int)threadIdx.x) * 72) + 4628)]));
+    conv2d_nchw[13] = (conv2d_nchw[13] + (pad_temp_shared[62] * kernel_shared[((((int)threadIdx.x) * 72) + 4628)]));
+    conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[63] * kernel_shared[((((int)threadIdx.x) * 72) + 21)]));
+    conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[64] * kernel_shared[((((int)threadIdx.x) * 72) + 21)]));
+    conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[65] * kernel_shared[((((int)threadIdx.x) * 72) + 21)]));
+    conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[66] * kernel_shared[((((int)threadIdx.x) * 72) + 21)]));
+    conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[67] * kernel_shared[((((int)threadIdx.x) * 72) + 21)]));
+    conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[68] * kernel_shared[((((int)threadIdx.x) * 72) + 21)]));
+    conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[69] * kernel_shared[((((int)threadIdx.x) * 72) + 21)]));
+    conv2d_nchw[7] = (conv2d_nchw[7] + (pad_temp_shared[63] * kernel_shared[((((int)threadIdx.x) * 72) + 4629)]));
+    conv2d_nchw[8] = (conv2d_nchw[8] + (pad_temp_shared[64] * kernel_shared[((((int)threadIdx.x) * 72) + 4629)]));
+    conv2d_nchw[9] = (conv2d_nchw[9] + (pad_temp_shared[65] * kernel_shared[((((int)threadIdx.x) * 72) + 4629)]));
+    conv2d_nchw[10] = (conv2d_nchw[10] + (pad_temp_shared[66] * kernel_shared[((((int)threadIdx.x) * 72) + 4629)]));
+    conv2d_nchw[11] = (conv2d_nchw[11] + (pad_temp_shared[67] * kernel_shared[((((int)threadIdx.x) * 72) + 4629)]));
+    conv2d_nchw[12] = (conv2d_nchw[12] + (pad_temp_shared[68] * kernel_shared[((((int)threadIdx.x) * 72) + 4629)]));
+    conv2d_nchw[13] = (conv2d_nchw[13] + (pad_temp_shared[69] * kernel_shared[((((int)threadIdx.x) * 72) + 4629)]));
+    conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[64] * kernel_shared[((((int)threadIdx.x) * 72) + 22)]));
+    conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[65] * kernel_shared[((((int)threadIdx.x) * 72) + 22)]));
+    conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[66] * kernel_shared[((((int)threadIdx.x) * 72) + 22)]));
+    conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[67] * kernel_shared[((((int)threadIdx.x) * 72) + 22)]));
+    conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[68] * kernel_shared[((((int)threadIdx.x) * 72) + 22)]));
+    conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[69] * kernel_shared[((((int)threadIdx.x) * 72) + 22)]));
+    conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[70] * kernel_shared[((((int)threadIdx.x) * 72) + 22)]));
+    conv2d_nchw[7] = (conv2d_nchw[7] + (pad_temp_shared[64] * kernel_shared[((((int)threadIdx.x) * 72) + 4630)]));
+    conv2d_nchw[8] = (conv2d_nchw[8] + (pad_temp_shared[65] * kernel_shared[((((int)threadIdx.x) * 72) + 4630)]));
+    conv2d_nchw[9] = (conv2d_nchw[9] + (pad_temp_shared[66] * kernel_shared[((((int)threadIdx.x) * 72) + 4630)]));
+    conv2d_nchw[10] = (conv2d_nchw[10] + (pad_temp_shared[67] * kernel_shared[((((int)threadIdx.x) * 72) + 4630)]));
+    conv2d_nchw[11] = (conv2d_nchw[11] + (pad_temp_shared[68] * kernel_shared[((((int)threadIdx.x) * 72) + 4630)]));
+    conv2d_nchw[12] = (conv2d_nchw[12] + (pad_temp_shared[69] * kernel_shared[((((int)threadIdx.x) * 72) + 4630)]));
+    conv2d_nchw[13] = (conv2d_nchw[13] + (pad_temp_shared[70] * kernel_shared[((((int)threadIdx.x) * 72) + 4630)]));
+    conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[65] * kernel_shared[((((int)threadIdx.x) * 72) + 23)]));
+    conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[66] * kernel_shared[((((int)threadIdx.x) * 72) + 23)]));
+    conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[67] * kernel_shared[((((int)threadIdx.x) * 72) + 23)]));
+    conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[68] * kernel_shared[((((int)threadIdx.x) * 72) + 23)]));
+    conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[69] * kernel_shared[((((int)threadIdx.x) * 72) + 23)]));
+    conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[70] * kernel_shared[((((int)threadIdx.x) * 72) + 23)]));
+    conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[71] * kernel_shared[((((int)threadIdx.x) * 72) + 23)]));
+    conv2d_nchw[7] = (conv2d_nchw[7] + (pad_temp_shared[65] * kernel_shared[((((int)threadIdx.x) * 72) + 4631)]));
+    conv2d_nchw[8] = (conv2d_nchw[8] + (pad_temp_shared[66] * kernel_shared[((((int)threadIdx.x) * 72) + 4631)]));
+    conv2d_nchw[9] = (conv2d_nchw[9] + (pad_temp_shared[67] * kernel_shared[((((int)threadIdx.x) * 72) + 4631)]));
+    conv2d_nchw[10] = (conv2d_nchw[10] + (pad_temp_shared[68] * kernel_shared[((((int)threadIdx.x) * 72) + 4631)]));
+    conv2d_nchw[11] = (conv2d_nchw[11] + (pad_temp_shared[69] * kernel_shared[((((int)threadIdx.x) * 72) + 4631)]));
+    conv2d_nchw[12] = (conv2d_nchw[12] + (pad_temp_shared[70] * kernel_shared[((((int)threadIdx.x) * 72) + 4631)]));
+    conv2d_nchw[13] = (conv2d_nchw[13] + (pad_temp_shared[71] * kernel_shared[((((int)threadIdx.x) * 72) + 4631)]));
+    conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[72] * kernel_shared[((((int)threadIdx.x) * 72) + 24)]));
+    conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[73] * kernel_shared[((((int)threadIdx.x) * 72) + 24)]));
+    conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[74] * kernel_shared[((((int)threadIdx.x) * 72) + 24)]));
+    conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[75] * kernel_shared[((((int)threadIdx.x) * 72) + 24)]));
+    conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[76] * kernel_shared[((((int)threadIdx.x) * 72) + 24)]));
+    conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[77] * kernel_shared[((((int)threadIdx.x) * 72) + 24)]));
+    conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[78] * kernel_shared[((((int)threadIdx.x) * 72) + 24)]));
+    conv2d_nchw[7] = (conv2d_nchw[7] + (pad_temp_shared[72] * kernel_shared[((((int)threadIdx.x) * 72) + 4632)]));
+    conv2d_nchw[8] = (conv2d_nchw[8] + (pad_temp_shared[73] * kernel_shared[((((int)threadIdx.x) * 72) + 4632)]));
+    conv2d_nchw[9] = (conv2d_nchw[9] + (pad_temp_shared[74] * kernel_shared[((((int)threadIdx.x) * 72) + 4632)]));
+    conv2d_nchw[10] = (conv2d_nchw[10] + (pad_temp_shared[75] * kernel_shared[((((int)threadIdx.x) * 72) + 4632)]));
+    conv2d_nchw[11] = (conv2d_nchw[11] + (pad_temp_shared[76] * kernel_shared[((((int)threadIdx.x) * 72) + 4632)]));
+    conv2d_nchw[12] = (conv2d_nchw[12] + (pad_temp_shared[77] * kernel_shared[((((int)threadIdx.x) * 72) + 4632)]));
+    conv2d_nchw[13] = (conv2d_nchw[13] + (pad_temp_shared[78] * kernel_shared[((((int)threadIdx.x) * 72) + 4632)]));
+    conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[73] * kernel_shared[((((int)threadIdx.x) * 72) + 25)]));
+    conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[74] * kernel_shared[((((int)threadIdx.x) * 72) + 25)]));
+    conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[75] * kernel_shared[((((int)threadIdx.x) * 72) + 25)]));
+    conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[76] * kernel_shared[((((int)threadIdx.x) * 72) + 25)]));
+    conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[77] * kernel_shared[((((int)threadIdx.x) * 72) + 25)]));
+    conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[78] * kernel_shared[((((int)threadIdx.x) * 72) + 25)]));
+    conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[79] * kernel_shared[((((int)threadIdx.x) * 72) + 25)]));
+    conv2d_nchw[7] = (conv2d_nchw[7] + (pad_temp_shared[73] * kernel_shared[((((int)threadIdx.x) * 72) + 4633)]));
+    conv2d_nchw[8] = (conv2d_nchw[8] + (pad_temp_shared[74] * kernel_shared[((((int)threadIdx.x) * 72) + 4633)]));
+    conv2d_nchw[9] = (conv2d_nchw[9] + (pad_temp_shared[75] * kernel_shared[((((int)threadIdx.x) * 72) + 4633)]));
+    conv2d_nchw[10] = (conv2d_nchw[10] + (pad_temp_shared[76] * kernel_shared[((((int)threadIdx.x) * 72) + 4633)]));
+    conv2d_nchw[11] = (conv2d_nchw[11] + (pad_temp_shared[77] * kernel_shared[((((int)threadIdx.x) * 72) + 4633)]));
+    conv2d_nchw[12] = (conv2d_nchw[12] + (pad_temp_shared[78] * kernel_shared[((((int)threadIdx.x) * 72) + 4633)]));
+    conv2d_nchw[13] = (conv2d_nchw[13] + (pad_temp_shared[79] * kernel_shared[((((int)threadIdx.x) * 72) + 4633)]));
+    conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[74] * kernel_shared[((((int)threadIdx.x) * 72) + 26)]));
+    conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[75] * kernel_shared[((((int)threadIdx.x) * 72) + 26)]));
+    conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[76] * kernel_shared[((((int)threadIdx.x) * 72) + 26)]));
+    conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[77] * kernel_shared[((((int)threadIdx.x) * 72) + 26)]));
+    conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[78] * kernel_shared[((((int)threadIdx.x) * 72) + 26)]));
+    conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[79] * kernel_shared[((((int)threadIdx.x) * 72) + 26)]));
+    conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[80] * kernel_shared[((((int)threadIdx.x) * 72) + 26)]));
+    conv2d_nchw[7] = (conv2d_nchw[7] + (pad_temp_shared[74] * kernel_shared[((((int)threadIdx.x) * 72) + 4634)]));
+    conv2d_nchw[8] = (conv2d_nchw[8] + (pad_temp_shared[75] * kernel_shared[((((int)threadIdx.x) * 72) + 4634)]));
+    conv2d_nchw[9] = (conv2d_nchw[9] + (pad_temp_shared[76] * kernel_shared[((((int)threadIdx.x) * 72) + 4634)]));
+    conv2d_nchw[10] = (conv2d_nchw[10] + (pad_temp_shared[77] * kernel_shared[((((int)threadIdx.x) * 72) + 4634)]));
+    conv2d_nchw[11] = (conv2d_nchw[11] + (pad_temp_shared[78] * kernel_shared[((((int)threadIdx.x) * 72) + 4634)]));
+    conv2d_nchw[12] = (conv2d_nchw[12] + (pad_temp_shared[79] * kernel_shared[((((int)threadIdx.x) * 72) + 4634)]));
+    conv2d_nchw[13] = (conv2d_nchw[13] + (pad_temp_shared[80] * kernel_shared[((((int)threadIdx.x) * 72) + 4634)]));
+    conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[81] * kernel_shared[((((int)threadIdx.x) * 72) + 27)]));
+    conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[82] * kernel_shared[((((int)threadIdx.x) * 72) + 27)]));
+    conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[83] * kernel_shared[((((int)threadIdx.x) * 72) + 27)]));
+    conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[84] * kernel_shared[((((int)threadIdx.x) * 72) + 27)]));
+    conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[85] * kernel_shared[((((int)threadIdx.x) * 72) + 27)]));
+    conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[86] * kernel_shared[((((int)threadIdx.x) * 72) + 27)]));
+    conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[87] * kernel_shared[((((int)threadIdx.x) * 72) + 27)]));
+    conv2d_nchw[7] = (conv2d_nchw[7] + (pad_temp_shared[81] * kernel_shared[((((int)threadIdx.x) * 72) + 4635)]));
+    conv2d_nchw[8] = (conv2d_nchw[8] + (pad_temp_shared[82] * kernel_shared[((((int)threadIdx.x) * 72) + 4635)]));
+    conv2d_nchw[9] = (conv2d_nchw[9] + (pad_temp_shared[83] * kernel_shared[((((int)threadIdx.x) * 72) + 4635)]));
+    conv2d_nchw[10] = (conv2d_nchw[10] + (pad_temp_shared[84] * kernel_shared[((((int)threadIdx.x) * 72) + 4635)]));
+    conv2d_nchw[11] = (conv2d_nchw[11] + (pad_temp_shared[85] * kernel_shared[((((int)threadIdx.x) * 72) + 4635)]));
+    conv2d_nchw[12] = (conv2d_nchw[12] + (pad_temp_shared[86] * kernel_shared[((((int)threadIdx.x) * 72) + 4635)]));
+    conv2d_nchw[13] = (conv2d_nchw[13] + (pad_temp_shared[87] * kernel_shared[((((int)threadIdx.x) * 72) + 4635)]));
+    conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[82] * kernel_shared[((((int)threadIdx.x) * 72) + 28)]));
+    conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[83] * kernel_shared[((((int)threadIdx.x) * 72) + 28)]));
+    conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[84] * kernel_shared[((((int)threadIdx.x) * 72) + 28)]));
+    conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[85] * kernel_shared[((((int)threadIdx.x) * 72) + 28)]));
+    conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[86] * kernel_shared[((((int)threadIdx.x) * 72) + 28)]));
+    conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[87] * kernel_shared[((((int)threadIdx.x) * 72) + 28)]));
+    conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[88] * kernel_shared[((((int)threadIdx.x) * 72) + 28)]));
+    conv2d_nchw[7] = (conv2d_nchw[7] + (pad_temp_shared[82] * kernel_shared[((((int)threadIdx.x) * 72) + 4636)]));
+    conv2d_nchw[8] = (conv2d_nchw[8] + (pad_temp_shared[83] * kernel_shared[((((int)threadIdx.x) * 72) + 4636)]));
+    conv2d_nchw[9] = (conv2d_nchw[9] + (pad_temp_shared[84] * kernel_shared[((((int)threadIdx.x) * 72) + 4636)]));
+    conv2d_nchw[10] = (conv2d_nchw[10] + (pad_temp_shared[85] * kernel_shared[((((int)threadIdx.x) * 72) + 4636)]));
+    conv2d_nchw[11] = (conv2d_nchw[11] + (pad_temp_shared[86] * kernel_shared[((((int)threadIdx.x) * 72) + 4636)]));
+    conv2d_nchw[12] = (conv2d_nchw[12] + (pad_temp_shared[87] * kernel_shared[((((int)threadIdx.x) * 72) + 4636)]));
+    conv2d_nchw[13] = (conv2d_nchw[13] + (pad_temp_shared[88] * kernel_shared[((((int)threadIdx.x) * 72) + 4636)]));
+    conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[83] * kernel_shared[((((int)threadIdx.x) * 72) + 29)]));
+    conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[84] * kernel_shared[((((int)threadIdx.x) * 72) + 29)]));
+    conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[85] * kernel_shared[((((int)threadIdx.x) * 72) + 29)]));
+    conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[86] * kernel_shared[((((int)threadIdx.x) * 72) + 29)]));
+    conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[87] * kernel_shared[((((int)threadIdx.x) * 72) + 29)]));
+    conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[88] * kernel_shared[((((int)threadIdx.x) * 72) + 29)]));
+    conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[89] * kernel_shared[((((int)threadIdx.x) * 72) + 29)]));
+    conv2d_nchw[7] = (conv2d_nchw[7] + (pad_temp_shared[83] * kernel_shared[((((int)threadIdx.x) * 72) + 4637)]));
+    conv2d_nchw[8] = (conv2d_nchw[8] + (pad_temp_shared[84] * kernel_shared[((((int)threadIdx.x) * 72) + 4637)]));
+    conv2d_nchw[9] = (conv2d_nchw[9] + (pad_temp_shared[85] * kernel_shared[((((int)threadIdx.x) * 72) + 4637)]));
+    conv2d_nchw[10] = (conv2d_nchw[10] + (pad_temp_shared[86] * kernel_shared[((((int)threadIdx.x) * 72) + 4637)]));
+    conv2d_nchw[11] = (conv2d_nchw[11] + (pad_temp_shared[87] * kernel_shared[((((int)threadIdx.x) * 72) + 4637)]));
+    conv2d_nchw[12] = (conv2d_nchw[12] + (pad_temp_shared[88] * kernel_shared[((((int)threadIdx.x) * 72) + 4637)]));
+    conv2d_nchw[13] = (conv2d_nchw[13] + (pad_temp_shared[89] * kernel_shared[((((int)threadIdx.x) * 72) + 4637)]));
+    conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[90] * kernel_shared[((((int)threadIdx.x) * 72) + 30)]));
+    conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[91] * kernel_shared[((((int)threadIdx.x) * 72) + 30)]));
+    conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[92] * kernel_shared[((((int)threadIdx.x) * 72) + 30)]));
+    conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[93] * kernel_shared[((((int)threadIdx.x) * 72) + 30)]));
+    conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[94] * kernel_shared[((((int)threadIdx.x) * 72) + 30)]));
+    conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[95] * kernel_shared[((((int)threadIdx.x) * 72) + 30)]));
+    conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[96] * kernel_shared[((((int)threadIdx.x) * 72) + 30)]));
+    conv2d_nchw[7] = (conv2d_nchw[7] + (pad_temp_shared[90] * kernel_shared[((((int)threadIdx.x) * 72) + 4638)]));
+    conv2d_nchw[8] = (conv2d_nchw[8] + (pad_temp_shared[91] * kernel_shared[((((int)threadIdx.x) * 72) + 4638)]));
+    conv2d_nchw[9] = (conv2d_nchw[9] + (pad_temp_shared[92] * kernel_shared[((((int)threadIdx.x) * 72) + 4638)]));
+    conv2d_nchw[10] = (conv2d_nchw[10] + (pad_temp_shared[93] * kernel_shared[((((int)threadIdx.x) * 72) + 4638)]));
+    conv2d_nchw[11] = (conv2d_nchw[11] + (pad_temp_shared[94] * kernel_shared[((((int)threadIdx.x) * 72) + 4638)]));
+    conv2d_nchw[12] = (conv2d_nchw[12] + (pad_temp_shared[95] * kernel_shared[((((int)threadIdx.x) * 72) + 4638)]));
+    conv2d_nchw[13] = (conv2d_nchw[13] + (pad_temp_shared[96] * kernel_shared[((((int)threadIdx.x) * 72) + 4638)]));
+    conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[91] * kernel_shared[((((int)threadIdx.x) * 72) + 31)]));
+    conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[92] * kernel_shared[((((int)threadIdx.x) * 72) + 31)]));
+    conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[93] * kernel_shared[((((int)threadIdx.x) * 72) + 31)]));
+    conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[94] * kernel_shared[((((int)threadIdx.x) * 72) + 31)]));
+    conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[95] * kernel_shared[((((int)threadIdx.x) * 72) + 31)]));
+    conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[96] * kernel_shared[((((int)threadIdx.x) * 72) + 31)]));
+    conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[97] * kernel_shared[((((int)threadIdx.x) * 72) + 31)]));
+    conv2d_nchw[7] = (conv2d_nchw[7] + (pad_temp_shared[91] * kernel_shared[((((int)threadIdx.x) * 72) + 4639)]));
+    conv2d_nchw[8] = (conv2d_nchw[8] + (pad_temp_shared[92] * kernel_shared[((((int)threadIdx.x) * 72) + 4639)]));
+    conv2d_nchw[9] = (conv2d_nchw[9] + (pad_temp_shared[93] * kernel_shared[((((int)threadIdx.x) * 72) + 4639)]));
+    conv2d_nchw[10] = (conv2d_nchw[10] + (pad_temp_shared[94] * kernel_shared[((((int)threadIdx.x) * 72) + 4639)]));
+    conv2d_nchw[11] = (conv2d_nchw[11] + (pad_temp_shared[95] * kernel_shared[((((int)threadIdx.x) * 72) + 4639)]));
+    conv2d_nchw[12] = (conv2d_nchw[12] + (pad_temp_shared[96] * kernel_shared[((((int)threadIdx.x) * 72) + 4639)]));
+    conv2d_nchw[13] = (conv2d_nchw[13] + (pad_temp_shared[97] * kernel_shared[((((int)threadIdx.x) * 72) + 4639)]));
+    conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[92] * kernel_shared[((((int)threadIdx.x) * 72) + 32)]));
+    conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[93] * kernel_shared[((((int)threadIdx.x) * 72) + 32)]));
+    conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[94] * kernel_shared[((((int)threadIdx.x) * 72) + 32)]));
+    conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[95] * kernel_shared[((((int)threadIdx.x) * 72) + 32)]));
+    conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[96] * kernel_shared[((((int)threadIdx.x) * 72) + 32)]));
+    conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[97] * kernel_shared[((((int)threadIdx.x) * 72) + 32)]));
+    conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[98] * kernel_shared[((((int)threadIdx.x) * 72) + 32)]));
+    conv2d_nchw[7] = (conv2d_nchw[7] + (pad_temp_shared[92] * kernel_shared[((((int)threadIdx.x) * 72) + 4640)]));
+    conv2d_nchw[8] = (conv2d_nchw[8] + (pad_temp_shared[93] * kernel_shared[((((int)threadIdx.x) * 72) + 4640)]));
+    conv2d_nchw[9] = (conv2d_nchw[9] + (pad_temp_shared[94] * kernel_shared[((((int)threadIdx.x) * 72) + 4640)]));
+    conv2d_nchw[10] = (conv2d_nchw[10] + (pad_temp_shared[95] * kernel_shared[((((int)threadIdx.x) * 72) + 4640)]));
+    conv2d_nchw[11] = (conv2d_nchw[11] + (pad_temp_shared[96] * kernel_shared[((((int)threadIdx.x) * 72) + 4640)]));
+    conv2d_nchw[12] = (conv2d_nchw[12] + (pad_temp_shared[97] * kernel_shared[((((int)threadIdx.x) * 72) + 4640)]));
+    conv2d_nchw[13] = (conv2d_nchw[13] + (pad_temp_shared[98] * kernel_shared[((((int)threadIdx.x) * 72) + 4640)]));
+    conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[99] * kernel_shared[((((int)threadIdx.x) * 72) + 33)]));
+    conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[100] * kernel_shared[((((int)threadIdx.x) * 72) + 33)]));
+    conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[101] * kernel_shared[((((int)threadIdx.x) * 72) + 33)]));
+    conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[102] * kernel_shared[((((int)threadIdx.x) * 72) + 33)]));
+    conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[103] * kernel_shared[((((int)threadIdx.x) * 72) + 33)]));
+    conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[104] * kernel_shared[((((int)threadIdx.x) * 72) + 33)]));
+    conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[105] * kernel_shared[((((int)threadIdx.x) * 72) + 33)]));
+    conv2d_nchw[7] = (conv2d_nchw[7] + (pad_temp_shared[99] * kernel_shared[((((int)threadIdx.x) * 72) + 4641)]));
+    conv2d_nchw[8] = (conv2d_nchw[8] + (pad_temp_shared[100] * kernel_shared[((((int)threadIdx.x) * 72) + 4641)]));
+    conv2d_nchw[9] = (conv2d_nchw[9] + (pad_temp_shared[101] * kernel_shared[((((int)threadIdx.x) * 72) + 4641)]));
+    conv2d_nchw[10] = (conv2d_nchw[10] + (pad_temp_shared[102] * kernel_shared[((((int)threadIdx.x) * 72) + 4641)]));
+    conv2d_nchw[11] = (conv2d_nchw[11] + (pad_temp_shared[103] * kernel_shared[((((int)threadIdx.x) * 72) + 4641)]));
+    conv2d_nchw[12] = (conv2d_nchw[12] + (pad_temp_shared[104] * kernel_shared[((((int)threadIdx.x) * 72) + 4641)]));
+    conv2d_nchw[13] = (conv2d_nchw[13] + (pad_temp_shared[105] * kernel_shared[((((int)threadIdx.x) * 72) + 4641)]));
+    conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[100] * kernel_shared[((((int)threadIdx.x) * 72) + 34)]));
+    conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[101] * kernel_shared[((((int)threadIdx.x) * 72) + 34)]));
+    conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[102] * kernel_shared[((((int)threadIdx.x) * 72) + 34)]));
+    conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[103] * kernel_shared[((((int)threadIdx.x) * 72) + 34)]));
+    conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[104] * kernel_shared[((((int)threadIdx.x) * 72) + 34)]));
+    conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[105] * kernel_shared[((((int)threadIdx.x) * 72) + 34)]));
+    conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[106] * kernel_shared[((((int)threadIdx.x) * 72) + 34)]));
+    conv2d_nchw[7] = (conv2d_nchw[7] + (pad_temp_shared[100] * kernel_shared[((((int)threadIdx.x) * 72) + 4642)]));
+    conv2d_nchw[8] = (conv2d_nchw[8] + (pad_temp_shared[101] * kernel_shared[((((int)threadIdx.x) * 72) + 4642)]));
+    conv2d_nchw[9] = (conv2d_nchw[9] + (pad_temp_shared[102] * kernel_shared[((((int)threadIdx.x) * 72) + 4642)]));
+    conv2d_nchw[10] = (conv2d_nchw[10] + (pad_temp_shared[103] * kernel_shared[((((int)threadIdx.x) * 72) + 4642)]));
+    conv2d_nchw[11] = (conv2d_nchw[11] + (pad_temp_shared[104] * kernel_shared[((((int)threadIdx.x) * 72) + 4642)]));
+    conv2d_nchw[12] = (conv2d_nchw[12] + (pad_temp_shared[105] * kernel_shared[((((int)threadIdx.x) * 72) + 4642)]));
+    conv2d_nchw[13] = (conv2d_nchw[13] + (pad_temp_shared[106] * kernel_shared[((((int)threadIdx.x) * 72) + 4642)]));
+    conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[101] * kernel_shared[((((int)threadIdx.x) * 72) + 35)]));
+    conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[102] * kernel_shared[((((int)threadIdx.x) * 72) + 35)]));
+    conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[103] * kernel_shared[((((int)threadIdx.x) * 72) + 35)]));
+    conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[104] * kernel_shared[((((int)threadIdx.x) * 72) + 35)]));
+    conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[105] * kernel_shared[((((int)threadIdx.x) * 72) + 35)]));
+    conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[106] * kernel_shared[((((int)threadIdx.x) * 72) + 35)]));
+    conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[107] * kernel_shared[((((int)threadIdx.x) * 72) + 35)]));
+    conv2d_nchw[7] = (conv2d_nchw[7] + (pad_temp_shared[101] * kernel_shared[((((int)threadIdx.x) * 72) + 4643)]));
+    conv2d_nchw[8] = (conv2d_nchw[8] + (pad_temp_shared[102] * kernel_shared[((((int)threadIdx.x) * 72) + 4643)]));
+    conv2d_nchw[9] = (conv2d_nchw[9] + (pad_temp_shared[103] * kernel_shared[((((int)threadIdx.x) * 72) + 4643)]));
+    conv2d_nchw[10] = (conv2d_nchw[10] + (pad_temp_shared[104] * kernel_shared[((((int)threadIdx.x) * 72) + 4643)]));
+    conv2d_nchw[11] = (conv2d_nchw[11] + (pad_temp_shared[105] * kernel_shared[((((int)threadIdx.x) * 72) + 4643)]));
+    conv2d_nchw[12] = (conv2d_nchw[12] + (pad_temp_shared[106] * kernel_shared[((((int)threadIdx.x) * 72) + 4643)]));
+    conv2d_nchw[13] = (conv2d_nchw[13] + (pad_temp_shared[107] * kernel_shared[((((int)threadIdx.x) * 72) + 4643)]));
+    conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[108] * kernel_shared[((((int)threadIdx.x) * 72) + 36)]));
+    conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[109] * kernel_shared[((((int)threadIdx.x) * 72) + 36)]));
+    conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[110] * kernel_shared[((((int)threadIdx.x) * 72) + 36)]));
+    conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[111] * kernel_shared[((((int)threadIdx.x) * 72) + 36)]));
+    conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[112] * kernel_shared[((((int)threadIdx.x) * 72) + 36)]));
+    conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[113] * kernel_shared[((((int)threadIdx.x) * 72) + 36)]));
+    conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[114] * kernel_shared[((((int)threadIdx.x) * 72) + 36)]));
+    conv2d_nchw[7] = (conv2d_nchw[7] + (pad_temp_shared[108] * kernel_shared[((((int)threadIdx.x) * 72) + 4644)]));
+    conv2d_nchw[8] = (conv2d_nchw[8] + (pad_temp_shared[109] * kernel_shared[((((int)threadIdx.x) * 72) + 4644)]));
+    conv2d_nchw[9] = (conv2d_nchw[9] + (pad_temp_shared[110] * kernel_shared[((((int)threadIdx.x) * 72) + 4644)]));
+    conv2d_nchw[10] = (conv2d_nchw[10] + (pad_temp_shared[111] * kernel_shared[((((int)threadIdx.x) * 72) + 4644)]));
+    conv2d_nchw[11] = (conv2d_nchw[11] + (pad_temp_shared[112] * kernel_shared[((((int)threadIdx.x) * 72) + 4644)]));
+    conv2d_nchw[12] = (conv2d_nchw[12] + (pad_temp_shared[113] * kernel_shared[((((int)threadIdx.x) * 72) + 4644)]));
+    conv2d_nchw[13] = (conv2d_nchw[13] + (pad_temp_shared[114] * kernel_shared[((((int)threadIdx.x) * 72) + 4644)]));
+    conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[109] * kernel_shared[((((int)threadIdx.x) * 72) + 37)]));
+    conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[110] * kernel_shared[((((int)threadIdx.x) * 72) + 37)]));
+    conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[111] * kernel_shared[((((int)threadIdx.x) * 72) + 37)]));
+    conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[112] * kernel_shared[((((int)threadIdx.x) * 72) + 37)]));
+    conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[113] * kernel_shared[((((int)threadIdx.x) * 72) + 37)]));
+    conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[114] * kernel_shared[((((int)threadIdx.x) * 72) + 37)]));
+    conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[115] * kernel_shared[((((int)threadIdx.x) * 72) + 37)]));
+    conv2d_nchw[7] = (conv2d_nchw[7] + (pad_temp_shared[109] * kernel_shared[((((int)threadIdx.x) * 72) + 4645)]));
+    conv2d_nchw[8] = (conv2d_nchw[8] + (pad_temp_shared[110] * kernel_shared[((((int)threadIdx.x) * 72) + 4645)]));
+    conv2d_nchw[9] = (conv2d_nchw[9] + (pad_temp_shared[111] * kernel_shared[((((int)threadIdx.x) * 72) + 4645)]));
+    conv2d_nchw[10] = (conv2d_nchw[10] + (pad_temp_shared[112] * kernel_shared[((((int)threadIdx.x) * 72) + 4645)]));
+    conv2d_nchw[11] = (conv2d_nchw[11] + (pad_temp_shared[113] * kernel_shared[((((int)threadIdx.x) * 72) + 4645)]));
+    conv2d_nchw[12] = (conv2d_nchw[12] + (pad_temp_shared[114] * kernel_shared[((((int)threadIdx.x) * 72) + 4645)]));
+    conv2d_nchw[13] = (conv2d_nchw[13] + (pad_temp_shared[115] * kernel_shared[((((int)threadIdx.x) * 72) + 4645)]));
+    conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[110] * kernel_shared[((((int)threadIdx.x) * 72) + 38)]));
+    conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[111] * kernel_shared[((((int)threadIdx.x) * 72) + 38)]));
+    conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[112] * kernel_shared[((((int)threadIdx.x) * 72) + 38)]));
+    conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[113] * kernel_shared[((((int)threadIdx.x) * 72) + 38)]));
+    conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[114] * kernel_shared[((((int)threadIdx.x) * 72) + 38)]));
+    conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[115] * kernel_shared[((((int)threadIdx.x) * 72) + 38)]));
+    conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[116] * kernel_shared[((((int)threadIdx.x) * 72) + 38)]));
+    conv2d_nchw[7] = (conv2d_nchw[7] + (pad_temp_shared[110] * kernel_shared[((((int)threadIdx.x) * 72) + 4646)]));
+    conv2d_nchw[8] = (conv2d_nchw[8] + (pad_temp_shared[111] * kernel_shared[((((int)threadIdx.x) * 72) + 4646)]));
+    conv2d_nchw[9] = (conv2d_nchw[9] + (pad_temp_shared[112] * kernel_shared[((((int)threadIdx.x) * 72) + 4646)]));
+    conv2d_nchw[10] = (conv2d_nchw[10] + (pad_temp_shared[113] * kernel_shared[((((int)threadIdx.x) * 72) + 4646)]));
+    conv2d_nchw[11] = (conv2d_nchw[11] + (pad_temp_shared[114] * kernel_shared[((((int)threadIdx.x) * 72) + 4646)]));
+    conv2d_nchw[12] = (conv2d_nchw[12] + (pad_temp_shared[115] * kernel_shared[((((int)threadIdx.x) * 72) + 4646)]));
+    conv2d_nchw[13] = (conv2d_nchw[13] + (pad_temp_shared[116] * kernel_shared[((((int)threadIdx.x) * 72) + 4646)]));
+    conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[117] * kernel_shared[((((int)threadIdx.x) * 72) + 39)]));
+    conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[118] * kernel_shared[((((int)threadIdx.x) * 72) + 39)]));
+    conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[119] * kernel_shared[((((int)threadIdx.x) * 72) + 39)]));
+    conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[120] * kernel_shared[((((int)threadIdx.x) * 72) + 39)]));
+    conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[121] * kernel_shared[((((int)threadIdx.x) * 72) + 39)]));
+    conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[122] * kernel_shared[((((int)threadIdx.x) * 72) + 39)]));
+    conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[123] * kernel_shared[((((int)threadIdx.x) * 72) + 39)]));
+    conv2d_nchw[7] = (conv2d_nchw[7] + (pad_temp_shared[117] * kernel_shared[((((int)threadIdx.x) * 72) + 4647)]));
+    conv2d_nchw[8] = (conv2d_nchw[8] + (pad_temp_shared[118] * kernel_shared[((((int)threadIdx.x) * 72) + 4647)]));
+    conv2d_nchw[9] = (conv2d_nchw[9] + (pad_temp_shared[119] * kernel_shared[((((int)threadIdx.x) * 72) + 4647)]));
+    conv2d_nchw[10] = (conv2d_nchw[10] + (pad_temp_shared[120] * kernel_shared[((((int)threadIdx.x) * 72) + 4647)]));
+    conv2d_nchw[11] = (conv2d_nchw[11] + (pad_temp_shared[121] * kernel_shared[((((int)threadIdx.x) * 72) + 4647)]));
+    conv2d_nchw[12] = (conv2d_nchw[12] + (pad_temp_shared[122] * kernel_shared[((((int)threadIdx.x) * 72) + 4647)]));
+    conv2d_nchw[13] = (conv2d_nchw[13] + (pad_temp_shared[123] * kernel_shared[((((int)threadIdx.x) * 72) + 4647)]));
+    conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[118] * kernel_shared[((((int)threadIdx.x) * 72) + 40)]));
+    conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[119] * kernel_shared[((((int)threadIdx.x) * 72) + 40)]));
+    conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[120] * kernel_shared[((((int)threadIdx.x) * 72) + 40)]));
+    conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[121] * kernel_shared[((((int)threadIdx.x) * 72) + 40)]));
+    conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[122] * kernel_shared[((((int)threadIdx.x) * 72) + 40)]));
+    conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[123] * kernel_shared[((((int)threadIdx.x) * 72) + 40)]));
+    conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[124] * kernel_shared[((((int)threadIdx.x) * 72) + 40)]));
+    conv2d_nchw[7] = (conv2d_nchw[7] + (pad_temp_shared[118] * kernel_shared[((((int)threadIdx.x) * 72) + 4648)]));
+    conv2d_nchw[8] = (conv2d_nchw[8] + (pad_temp_shared[119] * kernel_shared[((((int)threadIdx.x) * 72) + 4648)]));
+    conv2d_nchw[9] = (conv2d_nchw[9] + (pad_temp_shared[120] * kernel_shared[((((int)threadIdx.x) * 72) + 4648)]));
+    conv2d_nchw[10] = (conv2d_nchw[10] + (pad_temp_shared[121] * kernel_shared[((((int)threadIdx.x) * 72) + 4648)]));
+    conv2d_nchw[11] = (conv2d_nchw[11] + (pad_temp_shared[122] * kernel_shared[((((int)threadIdx.x) * 72) + 4648)]));
+    conv2d_nchw[12] = (conv2d_nchw[12] + (pad_temp_shared[123] * kernel_shared[((((int)threadIdx.x) * 72) + 4648)]));
+    conv2d_nchw[13] = (conv2d_nchw[13] + (pad_temp_shared[124] * kernel_shared[((((int)threadIdx.x) * 72) + 4648)]));
+    conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[119] * kernel_shared[((((int)threadIdx.x) * 72) + 41)]));
+    conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[120] * kernel_shared[((((int)threadIdx.x) * 72) + 41)]));
+    conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[121] * kernel_shared[((((int)threadIdx.x) * 72) + 41)]));
+    conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[122] * kernel_shared[((((int)threadIdx.x) * 72) + 41)]));
+    conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[123] * kernel_shared[((((int)threadIdx.x) * 72) + 41)]));
+    conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[124] * kernel_shared[((((int)threadIdx.x) * 72) + 41)]));
+    conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[125] * kernel_shared[((((int)threadIdx.x) * 72) + 41)]));
+    conv2d_nchw[7] = (conv2d_nchw[7] + (pad_temp_shared[119] * kernel_shared[((((int)threadIdx.x) * 72) + 4649)]));
+    conv2d_nchw[8] = (conv2d_nchw[8] + (pad_temp_shared[120] * kernel_shared[((((int)threadIdx.x) * 72) + 4649)]));
+    conv2d_nchw[9] = (conv2d_nchw[9] + (pad_temp_shared[121] * kernel_shared[((((int)threadIdx.x) * 72) + 4649)]));
+    conv2d_nchw[10] = (conv2d_nchw[10] + (pad_temp_shared[122] * kernel_shared[((((int)threadIdx.x) * 72) + 4649)]));
+    conv2d_nchw[11] = (conv2d_nchw[11] + (pad_temp_shared[123] * kernel_shared[((((int)threadIdx.x) * 72) + 4649)]));
+    conv2d_nchw[12] = (conv2d_nchw[12] + (pad_temp_shared[124] * kernel_shared[((((int)threadIdx.x) * 72) + 4649)]));
+    conv2d_nchw[13] = (conv2d_nchw[13] + (pad_temp_shared[125] * kernel_shared[((((int)threadIdx.x) * 72) + 4649)]));
+    conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[126] * kernel_shared[((((int)threadIdx.x) * 72) + 42)]));
+    conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[127] * kernel_shared[((((int)threadIdx.x) * 72) + 42)]));
+    conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[128] * kernel_shared[((((int)threadIdx.x) * 72) + 42)]));
+    conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[129] * kernel_shared[((((int)threadIdx.x) * 72) + 42)]));
+    conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[130] * kernel_shared[((((int)threadIdx.x) * 72) + 42)]));
+    conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[131] * kernel_shared[((((int)threadIdx.x) * 72) + 42)]));
+    conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[132] * kernel_shared[((((int)threadIdx.x) * 72) + 42)]));
+    conv2d_nchw[7] = (conv2d_nchw[7] + (pad_temp_shared[126] * kernel_shared[((((int)threadIdx.x) * 72) + 4650)]));
+    conv2d_nchw[8] = (conv2d_nchw[8] + (pad_temp_shared[127] * kernel_shared[((((int)threadIdx.x) * 72) + 4650)]));
+    conv2d_nchw[9] = (conv2d_nchw[9] + (pad_temp_shared[128] * kernel_shared[((((int)threadIdx.x) * 72) + 4650)]));
+    conv2d_nchw[10] = (conv2d_nchw[10] + (pad_temp_shared[129] * kernel_shared[((((int)threadIdx.x) * 72) + 4650)]));
+    conv2d_nchw[11] = (conv2d_nchw[11] + (pad_temp_shared[130] * kernel_shared[((((int)threadIdx.x) * 72) + 4650)]));
+    conv2d_nchw[12] = (conv2d_nchw[12] + (pad_temp_shared[131] * kernel_shared[((((int)threadIdx.x) * 72) + 4650)]));
+    conv2d_nchw[13] = (conv2d_nchw[13] + (pad_temp_shared[132] * kernel_shared[((((int)threadIdx.x) * 72) + 4650)]));
+    conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[127] * kernel_shared[((((int)threadIdx.x) * 72) + 43)]));
+    conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[128] * kernel_shared[((((int)threadIdx.x) * 72) + 43)]));
+    conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[129] * kernel_shared[((((int)threadIdx.x) * 72) + 43)]));
+    conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[130] * kernel_shared[((((int)threadIdx.x) * 72) + 43)]));
+    conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[131] * kernel_shared[((((int)threadIdx.x) * 72) + 43)]));
+    conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[132] * kernel_shared[((((int)threadIdx.x) * 72) + 43)]));
+    conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[133] * kernel_shared[((((int)threadIdx.x) * 72) + 43)]));
+    conv2d_nchw[7] = (conv2d_nchw[7] + (pad_temp_shared[127] * kernel_shared[((((int)threadIdx.x) * 72) + 4651)]));
+    conv2d_nchw[8] = (conv2d_nchw[8] + (pad_temp_shared[128] * kernel_shared[((((int)threadIdx.x) * 72) + 4651)]));
+    conv2d_nchw[9] = (conv2d_nchw[9] + (pad_temp_shared[129] * kernel_shared[((((int)threadIdx.x) * 72) + 4651)]));
+    conv2d_nchw[10] = (conv2d_nchw[10] + (pad_temp_shared[130] * kernel_shared[((((int)threadIdx.x) * 72) + 4651)]));
+    conv2d_nchw[11] = (conv2d_nchw[11] + (pad_temp_shared[131] * kernel_shared[((((int)threadIdx.x) * 72) + 4651)]));
+    conv2d_nchw[12] = (conv2d_nchw[12] + (pad_temp_shared[132] * kernel_shared[((((int)threadIdx.x) * 72) + 4651)]));
+    conv2d_nchw[13] = (conv2d_nchw[13] + (pad_temp_shared[133] * kernel_shared[((((int)threadIdx.x) * 72) + 4651)]));
+    conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[128] * kernel_shared[((((int)threadIdx.x) * 72) + 44)]));
+    conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[129] * kernel_shared[((((int)threadIdx.x) * 72) + 44)]));
+    conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[130] * kernel_shared[((((int)threadIdx.x) * 72) + 44)]));
+    conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[131] * kernel_shared[((((int)threadIdx.x) * 72) + 44)]));
+    conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[132] * kernel_shared[((((int)threadIdx.x) * 72) + 44)]));
+    conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[133] * kernel_shared[((((int)threadIdx.x) * 72) + 44)]));
+    conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[134] * kernel_shared[((((int)threadIdx.x) * 72) + 44)]));
+    conv2d_nchw[7] = (conv2d_nchw[7] + (pad_temp_shared[128] * kernel_shared[((((int)threadIdx.x) * 72) + 4652)]));
+    conv2d_nchw[8] = (conv2d_nchw[8] + (pad_temp_shared[129] * kernel_shared[((((int)threadIdx.x) * 72) + 4652)]));
+    conv2d_nchw[9] = (conv2d_nchw[9] + (pad_temp_shared[130] * kernel_shared[((((int)threadIdx.x) * 72) + 4652)]));
+    conv2d_nchw[10] = (conv2d_nchw[10] + (pad_temp_shared[131] * kernel_shared[((((int)threadIdx.x) * 72) + 4652)]));
+    conv2d_nchw[11] = (conv2d_nchw[11] + (pad_temp_shared[132] * kernel_shared[((((int)threadIdx.x) * 72) + 4652)]));
+    conv2d_nchw[12] = (conv2d_nchw[12] + (pad_temp_shared[133] * kernel_shared[((((int)threadIdx.x) * 72) + 4652)]));
+    conv2d_nchw[13] = (conv2d_nchw[13] + (pad_temp_shared[134] * kernel_shared[((((int)threadIdx.x) * 72) + 4652)]));
+    conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[135] * kernel_shared[((((int)threadIdx.x) * 72) + 45)]));
+    conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[136] * kernel_shared[((((int)threadIdx.x) * 72) + 45)]));
+    conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[137] * kernel_shared[((((int)threadIdx.x) * 72) + 45)]));
+    conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[138] * kernel_shared[((((int)threadIdx.x) * 72) + 45)]));
+    conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[139] * kernel_shared[((((int)threadIdx.x) * 72) + 45)]));
+    conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[140] * kernel_shared[((((int)threadIdx.x) * 72) + 45)]));
+    conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[141] * kernel_shared[((((int)threadIdx.x) * 72) + 45)]));
+    conv2d_nchw[7] = (conv2d_nchw[7] + (pad_temp_shared[135] * kernel_shared[((((int)threadIdx.x) * 72) + 4653)]));
+    conv2d_nchw[8] = (conv2d_nchw[8] + (pad_temp_shared[136] * kernel_shared[((((int)threadIdx.x) * 72) + 4653)]));
+    conv2d_nchw[9] = (conv2d_nchw[9] + (pad_temp_shared[137] * kernel_shared[((((int)threadIdx.x) * 72) + 4653)]));
+    conv2d_nchw[10] = (conv2d_nchw[10] + (pad_temp_shared[138] * kernel_shared[((((int)threadIdx.x) * 72) + 4653)]));
+    conv2d_nchw[11] = (conv2d_nchw[11] + (pad_temp_shared[139] * kernel_shared[((((int)threadIdx.x) * 72) + 4653)]));
+    conv2d_nchw[12] = (conv2d_nchw[12] + (pad_temp_shared[140] * kernel_shared[((((int)threadIdx.x) * 72) + 4653)]));
+    conv2d_nchw[13] = (conv2d_nchw[13] + (pad_temp_shared[141] * kernel_shared[((((int)threadIdx.x) * 72) + 4653)]));
+    conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[136] * kernel_shared[((((int)threadIdx.x) * 72) + 46)]));
+    conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[137] * kernel_shared[((((int)threadIdx.x) * 72) + 46)]));
+    conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[138] * kernel_shared[((((int)threadIdx.x) * 72) + 46)]));
+    conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[139] * kernel_shared[((((int)threadIdx.x) * 72) + 46)]));
+    conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[140] * kernel_shared[((((int)threadIdx.x) * 72) + 46)]));
+    conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[141] * kernel_shared[((((int)threadIdx.x) * 72) + 46)]));
+    conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[142] * kernel_shared[((((int)threadIdx.x) * 72) + 46)]));
+    conv2d_nchw[7] = (conv2d_nchw[7] + (pad_temp_shared[136] * kernel_shared[((((int)threadIdx.x) * 72) + 4654)]));
+    conv2d_nchw[8] = (conv2d_nchw[8] + (pad_temp_shared[137] * kernel_shared[((((int)threadIdx.x) * 72) + 4654)]));
+    conv2d_nchw[9] = (conv2d_nchw[9] + (pad_temp_shared[138] * kernel_shared[((((int)threadIdx.x) * 72) + 4654)]));
+    conv2d_nchw[10] = (conv2d_nchw[10] + (pad_temp_shared[139] * kernel_shared[((((int)threadIdx.x) * 72) + 4654)]));
+    conv2d_nchw[11] = (conv2d_nchw[11] + (pad_temp_shared[140] * kernel_shared[((((int)threadIdx.x) * 72) + 4654)]));
+    conv2d_nchw[12] = (conv2d_nchw[12] + (pad_temp_shared[141] * kernel_shared[((((int)threadIdx.x) * 72) + 4654)]));
+    conv2d_nchw[13] = (conv2d_nchw[13] + (pad_temp_shared[142] * kernel_shared[((((int)threadIdx.x) * 72) + 4654)]));
+    conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[137] * kernel_shared[((((int)threadIdx.x) * 72) + 47)]));
+    conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[138] * kernel_shared[((((int)threadIdx.x) * 72) + 47)]));
+    conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[139] * kernel_shared[((((int)threadIdx.x) * 72) + 47)]));
+    conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[140] * kernel_shared[((((int)threadIdx.x) * 72) + 47)]));
+    conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[141] * kernel_shared[((((int)threadIdx.x) * 72) + 47)]));
+    conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[142] * kernel_shared[((((int)threadIdx.x) * 72) + 47)]));
+    conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[143] * kernel_shared[((((int)threadIdx.x) * 72) + 47)]));
+    conv2d_nchw[7] = (conv2d_nchw[7] + (pad_temp_shared[137] * kernel_shared[((((int)threadIdx.x) * 72) + 4655)]));
+    conv2d_nchw[8] = (conv2d_nchw[8] + (pad_temp_shared[138] * kernel_shared[((((int)threadIdx.x) * 72) + 4655)]));
+    conv2d_nchw[9] = (conv2d_nchw[9] + (pad_temp_shared[139] * kernel_shared[((((int)threadIdx.x) * 72) + 4655)]));
+    conv2d_nchw[10] = (conv2d_nchw[10] + (pad_temp_shared[140] * kernel_shared[((((int)threadIdx.x) * 72) + 4655)]));
+    conv2d_nchw[11] = (conv2d_nchw[11] + (pad_temp_shared[141] * kernel_shared[((((int)threadIdx.x) * 72) + 4655)]));
+    conv2d_nchw[12] = (conv2d_nchw[12] + (pad_temp_shared[142] * kernel_shared[((((int)threadIdx.x) * 72) + 4655)]));
+    conv2d_nchw[13] = (conv2d_nchw[13] + (pad_temp_shared[143] * kernel_shared[((((int)threadIdx.x) * 72) + 4655)]));
+    conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[144] * kernel_shared[((((int)threadIdx.x) * 72) + 48)]));
+    conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[145] * kernel_shared[((((int)threadIdx.x) * 72) + 48)]));
+    conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[146] * kernel_shared[((((int)threadIdx.x) * 72) + 48)]));
+    conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[147] * kernel_shared[((((int)threadIdx.x) * 72) + 48)]));
+    conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[148] * kernel_shared[((((int)threadIdx.x) * 72) + 48)]));
+    conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[149] * kernel_shared[((((int)threadIdx.x) * 72) + 48)]));
+    conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[150] * kernel_shared[((((int)threadIdx.x) * 72) + 48)]));
+    conv2d_nchw[7] = (conv2d_nchw[7] + (pad_temp_shared[144] * kernel_shared[((((int)threadIdx.x) * 72) + 4656)]));
+    conv2d_nchw[8] = (conv2d_nchw[8] + (pad_temp_shared[145] * kernel_shared[((((int)threadIdx.x) * 72) + 4656)]));
+    conv2d_nchw[9] = (conv2d_nchw[9] + (pad_temp_shared[146] * kernel_shared[((((int)threadIdx.x) * 72) + 4656)]));
+    conv2d_nchw[10] = (conv2d_nchw[10] + (pad_temp_shared[147] * kernel_shared[((((int)threadIdx.x) * 72) + 4656)]));
+    conv2d_nchw[11] = (conv2d_nchw[11] + (pad_temp_shared[148] * kernel_shared[((((int)threadIdx.x) * 72) + 4656)]));
+    conv2d_nchw[12] = (conv2d_nchw[12] + (pad_temp_shared[149] * kernel_shared[((((int)threadIdx.x) * 72) + 4656)]));
+    conv2d_nchw[13] = (conv2d_nchw[13] + (pad_temp_shared[150] * kernel_shared[((((int)threadIdx.x) * 72) + 4656)]));
+    conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[145] * kernel_shared[((((int)threadIdx.x) * 72) + 49)]));
+    conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[146] * kernel_shared[((((int)threadIdx.x) * 72) + 49)]));
+    conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[147] * kernel_shared[((((int)threadIdx.x) * 72) + 49)]));
+    conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[148] * kernel_shared[((((int)threadIdx.x) * 72) + 49)]));
+    conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[149] * kernel_shared[((((int)threadIdx.x) * 72) + 49)]));
+    conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[150] * kernel_shared[((((int)threadIdx.x) * 72) + 49)]));
+    conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[151] * kernel_shared[((((int)threadIdx.x) * 72) + 49)]));
+    conv2d_nchw[7] = (conv2d_nchw[7] + (pad_temp_shared[145] * kernel_shared[((((int)threadIdx.x) * 72) + 4657)]));
+    conv2d_nchw[8] = (conv2d_nchw[8] + (pad_temp_shared[146] * kernel_shared[((((int)threadIdx.x) * 72) + 4657)]));
+    conv2d_nchw[9] = (conv2d_nchw[9] + (pad_temp_shared[147] * kernel_shared[((((int)threadIdx.x) * 72) + 4657)]));
+    conv2d_nchw[10] = (conv2d_nchw[10] + (pad_temp_shared[148] * kernel_shared[((((int)threadIdx.x) * 72) + 4657)]));
+    conv2d_nchw[11] = (conv2d_nchw[11] + (pad_temp_shared[149] * kernel_shared[((((int)threadIdx.x) * 72) + 4657)]));
+    conv2d_nchw[12] = (conv2d_nchw[12] + (pad_temp_shared[150] * kernel_shared[((((int)threadIdx.x) * 72) + 4657)]));
+    conv2d_nchw[13] = (conv2d_nchw[13] + (pad_temp_shared[151] * kernel_shared[((((int)threadIdx.x) * 72) + 4657)]));
+    conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[146] * kernel_shared[((((int)threadIdx.x) * 72) + 50)]));
+    conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[147] * kernel_shared[((((int)threadIdx.x) * 72) + 50)]));
+    conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[148] * kernel_shared[((((int)threadIdx.x) * 72) + 50)]));
+    conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[149] * kernel_shared[((((int)threadIdx.x) * 72) + 50)]));
+    conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[150] * kernel_shared[((((int)threadIdx.x) * 72) + 50)]));
+    conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[151] * kernel_shared[((((int)threadIdx.x) * 72) + 50)]));
+    conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[152] * kernel_shared[((((int)threadIdx.x) * 72) + 50)]));
+    conv2d_nchw[7] = (conv2d_nchw[7] + (pad_temp_shared[146] * kernel_shared[((((int)threadIdx.x) * 72) + 4658)]));
+    conv2d_nchw[8] = (conv2d_nchw[8] + (pad_temp_shared[147] * kernel_shared[((((int)threadIdx.x) * 72) + 4658)]));
+    conv2d_nchw[9] = (conv2d_nchw[9] + (pad_temp_shared[148] * kernel_shared[((((int)threadIdx.x) * 72) + 4658)]));
+    conv2d_nchw[10] = (conv2d_nchw[10] + (pad_temp_shared[149] * kernel_shared[((((int)threadIdx.x) * 72) + 4658)]));
+    conv2d_nchw[11] = (conv2d_nchw[11] + (pad_temp_shared[150] * kernel_shared[((((int)threadIdx.x) * 72) + 4658)]));
+    conv2d_nchw[12] = (conv2d_nchw[12] + (pad_temp_shared[151] * kernel_shared[((((int)threadIdx.x) * 72) + 4658)]));
+    conv2d_nchw[13] = (conv2d_nchw[13] + (pad_temp_shared[152] * kernel_shared[((((int)threadIdx.x) * 72) + 4658)]));
+    conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[153] * kernel_shared[((((int)threadIdx.x) * 72) + 51)]));
+    conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[154] * kernel_shared[((((int)threadIdx.x) * 72) + 51)]));
+    conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[155] * kernel_shared[((((int)threadIdx.x) * 72) + 51)]));
+    conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[156] * kernel_shared[((((int)threadIdx.x) * 72) + 51)]));
+    conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[157] * kernel_shared[((((int)threadIdx.x) * 72) + 51)]));
+    conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[158] * kernel_shared[((((int)threadIdx.x) * 72) + 51)]));
+    conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[159] * kernel_shared[((((int)threadIdx.x) * 72) + 51)]));
+    conv2d_nchw[7] = (conv2d_nchw[7] + (pad_temp_shared[153] * kernel_shared[((((int)threadIdx.x) * 72) + 4659)]));
+    conv2d_nchw[8] = (conv2d_nchw[8] + (pad_temp_shared[154] * kernel_shared[((((int)threadIdx.x) * 72) + 4659)]));
+    conv2d_nchw[9] = (conv2d_nchw[9] + (pad_temp_shared[155] * kernel_shared[((((int)threadIdx.x) * 72) + 4659)]));
+    conv2d_nchw[10] = (conv2d_nchw[10] + (pad_temp_shared[156] * kernel_shared[((((int)threadIdx.x) * 72) + 4659)]));
+    conv2d_nchw[11] = (conv2d_nchw[11] + (pad_temp_shared[157] * kernel_shared[((((int)threadIdx.x) * 72) + 4659)]));
+    conv2d_nchw[12] = (conv2d_nchw[12] + (pad_temp_shared[158] * kernel_shared[((((int)threadIdx.x) * 72) + 4659)]));
+    conv2d_nchw[13] = (conv2d_nchw[13] + (pad_temp_shared[159] * kernel_shared[((((int)threadIdx.x) * 72) + 4659)]));
+    conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[154] * kernel_shared[((((int)threadIdx.x) * 72) + 52)]));
+    conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[155] * kernel_shared[((((int)threadIdx.x) * 72) + 52)]));
+    conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[156] * kernel_shared[((((int)threadIdx.x) * 72) + 52)]));
+    conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[157] * kernel_shared[((((int)threadIdx.x) * 72) + 52)]));
+    conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[158] * kernel_shared[((((int)threadIdx.x) * 72) + 52)]));
+    conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[159] * kernel_shared[((((int)threadIdx.x) * 72) + 52)]));
+    conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[160] * kernel_shared[((((int)threadIdx.x) * 72) + 52)]));
+    conv2d_nchw[7] = (conv2d_nchw[7] + (pad_temp_shared[154] * kernel_shared[((((int)threadIdx.x) * 72) + 4660)]));
+    conv2d_nchw[8] = (conv2d_nchw[8] + (pad_temp_shared[155] * kernel_shared[((((int)threadIdx.x) * 72) + 4660)]));
+    conv2d_nchw[9] = (conv2d_nchw[9] + (pad_temp_shared[156] * kernel_shared[((((int)threadIdx.x) * 72) + 4660)]));
+    conv2d_nchw[10] = (conv2d_nchw[10] + (pad_temp_shared[157] * kernel_shared[((((int)threadIdx.x) * 72) + 4660)]));
+    conv2d_nchw[11] = (conv2d_nchw[11] + (pad_temp_shared[158] * kernel_shared[((((int)threadIdx.x) * 72) + 4660)]));
+    conv2d_nchw[12] = (conv2d_nchw[12] + (pad_temp_shared[159] * kernel_shared[((((int)threadIdx.x) * 72) + 4660)]));
+    conv2d_nchw[13] = (conv2d_nchw[13] + (pad_temp_shared[160] * kernel_shared[((((int)threadIdx.x) * 72) + 4660)]));
+    conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[155] * kernel_shared[((((int)threadIdx.x) * 72) + 53)]));
+    conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[156] * kernel_shared[((((int)threadIdx.x) * 72) + 53)]));
+    conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[157] * kernel_shared[((((int)threadIdx.x) * 72) + 53)]));
+    conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[158] * kernel_shared[((((int)threadIdx.x) * 72) + 53)]));
+    conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[159] * kernel_shared[((((int)threadIdx.x) * 72) + 53)]));
+    conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[160] * kernel_shared[((((int)threadIdx.x) * 72) + 53)]));
+    conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[161] * kernel_shared[((((int)threadIdx.x) * 72) + 53)]));
+    conv2d_nchw[7] = (conv2d_nchw[7] + (pad_temp_shared[155] * kernel_shared[((((int)threadIdx.x) * 72) + 4661)]));
+    conv2d_nchw[8] = (conv2d_nchw[8] + (pad_temp_shared[156] * kernel_shared[((((int)threadIdx.x) * 72) + 4661)]));
+    conv2d_nchw[9] = (conv2d_nchw[9] + (pad_temp_shared[157] * kernel_shared[((((int)threadIdx.x) * 72) + 4661)]));
+    conv2d_nchw[10] = (conv2d_nchw[10] + (pad_temp_shared[158] * kernel_shared[((((int)threadIdx.x) * 72) + 4661)]));
+    conv2d_nchw[11] = (conv2d_nchw[11] + (pad_temp_shared[159] * kernel_shared[((((int)threadIdx.x) * 72) + 4661)]));
+    conv2d_nchw[12] = (conv2d_nchw[12] + (pad_temp_shared[160] * kernel_shared[((((int)threadIdx.x) * 72) + 4661)]));
+    conv2d_nchw[13] = (conv2d_nchw[13] + (pad_temp_shared[161] * kernel_shared[((((int)threadIdx.x) * 72) + 4661)]));
+    conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[162] * kernel_shared[((((int)threadIdx.x) * 72) + 54)]));
+    conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[163] * kernel_shared[((((int)threadIdx.x) * 72) + 54)]));
+    conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[164] * kernel_shared[((((int)threadIdx.x) * 72) + 54)]));
+    conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[165] * kernel_shared[((((int)threadIdx.x) * 72) + 54)]));
+    conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[166] * kernel_shared[((((int)threadIdx.x) * 72) + 54)]));
+    conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[167] * kernel_shared[((((int)threadIdx.x) * 72) + 54)]));
+    conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[168] * kernel_shared[((((int)threadIdx.x) * 72) + 54)]));
+    conv2d_nchw[7] = (conv2d_nchw[7] + (pad_temp_shared[162] * kernel_shared[((((int)threadIdx.x) * 72) + 4662)]));
+    conv2d_nchw[8] = (conv2d_nchw[8] + (pad_temp_shared[163] * kernel_shared[((((int)threadIdx.x) * 72) + 4662)]));
+    conv2d_nchw[9] = (conv2d_nchw[9] + (pad_temp_shared[164] * kernel_shared[((((int)threadIdx.x) * 72) + 4662)]));
+    conv2d_nchw[10] = (conv2d_nchw[10] + (pad_temp_shared[165] * kernel_shared[((((int)threadIdx.x) * 72) + 4662)]));
+    conv2d_nchw[11] = (conv2d_nchw[11] + (pad_temp_shared[166] * kernel_shared[((((int)threadIdx.x) * 72) + 4662)]));
+    conv2d_nchw[12] = (conv2d_nchw[12] + (pad_temp_shared[167] * kernel_shared[((((int)threadIdx.x) * 72) + 4662)]));
+    conv2d_nchw[13] = (conv2d_nchw[13] + (pad_temp_shared[168] * kernel_shared[((((int)threadIdx.x) * 72) + 4662)]));
+    conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[163] * kernel_shared[((((int)threadIdx.x) * 72) + 55)]));
+    conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[164] * kernel_shared[((((int)threadIdx.x) * 72) + 55)]));
+    conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[165] * kernel_shared[((((int)threadIdx.x) * 72) + 55)]));
+    conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[166] * kernel_shared[((((int)threadIdx.x) * 72) + 55)]));
+    conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[167] * kernel_shared[((((int)threadIdx.x) * 72) + 55)]));
+    conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[168] * kernel_shared[((((int)threadIdx.x) * 72) + 55)]));
+    conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[169] * kernel_shared[((((int)threadIdx.x) * 72) + 55)]));
+    conv2d_nchw[7] = (conv2d_nchw[7] + (pad_temp_shared[163] * kernel_shared[((((int)threadIdx.x) * 72) + 4663)]));
+    conv2d_nchw[8] = (conv2d_nchw[8] + (pad_temp_shared[164] * kernel_shared[((((int)threadIdx.x) * 72) + 4663)]));
+    conv2d_nchw[9] = (conv2d_nchw[9] + (pad_temp_shared[165] * kernel_shared[((((int)threadIdx.x) * 72) + 4663)]));
+    conv2d_nchw[10] = (conv2d_nchw[10] + (pad_temp_shared[166] * kernel_shared[((((int)threadIdx.x) * 72) + 4663)]));
+    conv2d_nchw[11] = (conv2d_nchw[11] + (pad_temp_shared[167] * kernel_shared[((((int)threadIdx.x) * 72) + 4663)]));
+    conv2d_nchw[12] = (conv2d_nchw[12] + (pad_temp_shared[168] * kernel_shared[((((int)threadIdx.x) * 72) + 4663)]));
+    conv2d_nchw[13] = (conv2d_nchw[13] + (pad_temp_shared[169] * kernel_shared[((((int)threadIdx.x) * 72) + 4663)]));
+    conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[164] * kernel_shared[((((int)threadIdx.x) * 72) + 56)]));
+    conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[165] * kernel_shared[((((int)threadIdx.x) * 72) + 56)]));
+    conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[166] * kernel_shared[((((int)threadIdx.x) * 72) + 56)]));
+    conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[167] * kernel_shared[((((int)threadIdx.x) * 72) + 56)]));
+    conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[168] * kernel_shared[((((int)threadIdx.x) * 72) + 56)]));
+    conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[169] * kernel_shared[((((int)threadIdx.x) * 72) + 56)]));
+    conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[170] * kernel_shared[((((int)threadIdx.x) * 72) + 56)]));
+    conv2d_nchw[7] = (conv2d_nchw[7] + (pad_temp_shared[164] * kernel_shared[((((int)threadIdx.x) * 72) + 4664)]));
+    conv2d_nchw[8] = (conv2d_nchw[8] + (pad_temp_shared[165] * kernel_shared[((((int)threadIdx.x) * 72) + 4664)]));
+    conv2d_nchw[9] = (conv2d_nchw[9] + (pad_temp_shared[166] * kernel_shared[((((int)threadIdx.x) * 72) + 4664)]));
+    conv2d_nchw[10] = (conv2d_nchw[10] + (pad_temp_shared[167] * kernel_shared[((((int)threadIdx.x) * 72) + 4664)]));
+    conv2d_nchw[11] = (conv2d_nchw[11] + (pad_temp_shared[168] * kernel_shared[((((int)threadIdx.x) * 72) + 4664)]));
+    conv2d_nchw[12] = (conv2d_nchw[12] + (pad_temp_shared[169] * kernel_shared[((((int)threadIdx.x) * 72) + 4664)]));
+    conv2d_nchw[13] = (conv2d_nchw[13] + (pad_temp_shared[170] * kernel_shared[((((int)threadIdx.x) * 72) + 4664)]));
+    conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[171] * kernel_shared[((((int)threadIdx.x) * 72) + 57)]));
+    conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[172] * kernel_shared[((((int)threadIdx.x) * 72) + 57)]));
+    conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[173] * kernel_shared[((((int)threadIdx.x) * 72) + 57)]));
+    conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[174] * kernel_shared[((((int)threadIdx.x) * 72) + 57)]));
+    conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[175] * kernel_shared[((((int)threadIdx.x) * 72) + 57)]));
+    conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[176] * kernel_shared[((((int)threadIdx.x) * 72) + 57)]));
+    conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[177] * kernel_shared[((((int)threadIdx.x) * 72) + 57)]));
+    conv2d_nchw[7] = (conv2d_nchw[7] + (pad_temp_shared[171] * kernel_shared[((((int)threadIdx.x) * 72) + 4665)]));
+    conv2d_nchw[8] = (conv2d_nchw[8] + (pad_temp_shared[172] * kernel_shared[((((int)threadIdx.x) * 72) + 4665)]));
+    conv2d_nchw[9] = (conv2d_nchw[9] + (pad_temp_shared[173] * kernel_shared[((((int)threadIdx.x) * 72) + 4665)]));
+    conv2d_nchw[10] = (conv2d_nchw[10] + (pad_temp_shared[174] * kernel_shared[((((int)threadIdx.x) * 72) + 4665)]));
+    conv2d_nchw[11] = (conv2d_nchw[11] + (pad_temp_shared[175] * kernel_shared[((((int)threadIdx.x) * 72) + 4665)]));
+    conv2d_nchw[12] = (conv2d_nchw[12] + (pad_temp_shared[176] * kernel_shared[((((int)threadIdx.x) * 72) + 4665)]));
+    conv2d_nchw[13] = (conv2d_nchw[13] + (pad_temp_shared[177] * kernel_shared[((((int)threadIdx.x) * 72) + 4665)]));
+    conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[172] * kernel_shared[((((int)threadIdx.x) * 72) + 58)]));
+    conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[173] * kernel_shared[((((int)threadIdx.x) * 72) + 58)]));
+    conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[174] * kernel_shared[((((int)threadIdx.x) * 72) + 58)]));
+    conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[175] * kernel_shared[((((int)threadIdx.x) * 72) + 58)]));
+    conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[176] * kernel_shared[((((int)threadIdx.x) * 72) + 58)]));
+    conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[177] * kernel_shared[((((int)threadIdx.x) * 72) + 58)]));
+    conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[178] * kernel_shared[((((int)threadIdx.x) * 72) + 58)]));
+    conv2d_nchw[7] = (conv2d_nchw[7] + (pad_temp_shared[172] * kernel_shared[((((int)threadIdx.x) * 72) + 4666)]));
+    conv2d_nchw[8] = (conv2d_nchw[8] + (pad_temp_shared[173] * kernel_shared[((((int)threadIdx.x) * 72) + 4666)]));
+    conv2d_nchw[9] = (conv2d_nchw[9] + (pad_temp_shared[174] * kernel_shared[((((int)threadIdx.x) * 72) + 4666)]));
+    conv2d_nchw[10] = (conv2d_nchw[10] + (pad_temp_shared[175] * kernel_shared[((((int)threadIdx.x) * 72) + 4666)]));
+    conv2d_nchw[11] = (conv2d_nchw[11] + (pad_temp_shared[176] * kernel_shared[((((int)threadIdx.x) * 72) + 4666)]));
+    conv2d_nchw[12] = (conv2d_nchw[12] + (pad_temp_shared[177] * kernel_shared[((((int)threadIdx.x) * 72) + 4666)]));
+    conv2d_nchw[13] = (conv2d_nchw[13] + (pad_temp_shared[178] * kernel_shared[((((int)threadIdx.x) * 72) + 4666)]));
+    conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[173] * kernel_shared[((((int)threadIdx.x) * 72) + 59)]));
+    conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[174] * kernel_shared[((((int)threadIdx.x) * 72) + 59)]));
+    conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[175] * kernel_shared[((((int)threadIdx.x) * 72) + 59)]));
+    conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[176] * kernel_shared[((((int)threadIdx.x) * 72) + 59)]));
+    conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[177] * kernel_shared[((((int)threadIdx.x) * 72) + 59)]));
+    conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[178] * kernel_shared[((((int)threadIdx.x) * 72) + 59)]));
+    conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[179] * kernel_shared[((((int)threadIdx.x) * 72) + 59)]));
+    conv2d_nchw[7] = (conv2d_nchw[7] + (pad_temp_shared[173] * kernel_shared[((((int)threadIdx.x) * 72) + 4667)]));
+    conv2d_nchw[8] = (conv2d_nchw[8] + (pad_temp_shared[174] * kernel_shared[((((int)threadIdx.x) * 72) + 4667)]));
+    conv2d_nchw[9] = (conv2d_nchw[9] + (pad_temp_shared[175] * kernel_shared[((((int)threadIdx.x) * 72) + 4667)]));
+    conv2d_nchw[10] = (conv2d_nchw[10] + (pad_temp_shared[176] * kernel_shared[((((int)threadIdx.x) * 72) + 4667)]));
+    conv2d_nchw[11] = (conv2d_nchw[11] + (pad_temp_shared[177] * kernel_shared[((((int)threadIdx.x) * 72) + 4667)]));
+    conv2d_nchw[12] = (conv2d_nchw[12] + (pad_temp_shared[178] * kernel_shared[((((int)threadIdx.x) * 72) + 4667)]));
+    conv2d_nchw[13] = (conv2d_nchw[13] + (pad_temp_shared[179] * kernel_shared[((((int)threadIdx.x) * 72) + 4667)]));
+    conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[180] * kernel_shared[((((int)threadIdx.x) * 72) + 60)]));
+    conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[181] * kernel_shared[((((int)threadIdx.x) * 72) + 60)]));
+    conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[182] * kernel_shared[((((int)threadIdx.x) * 72) + 60)]));
+    conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[183] * kernel_shared[((((int)threadIdx.x) * 72) + 60)]));
+    conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[184] * kernel_shared[((((int)threadIdx.x) * 72) + 60)]));
+    conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[185] * kernel_shared[((((int)threadIdx.x) * 72) + 60)]));
+    conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[186] * kernel_shared[((((int)threadIdx.x) * 72) + 60)]));
+    conv2d_nchw[7] = (conv2d_nchw[7] + (pad_temp_shared[180] * kernel_shared[((((int)threadIdx.x) * 72) + 4668)]));
+    conv2d_nchw[8] = (conv2d_nchw[8] + (pad_temp_shared[181] * kernel_shared[((((int)threadIdx.x) * 72) + 4668)]));
+    conv2d_nchw[9] = (conv2d_nchw[9] + (pad_temp_shared[182] * kernel_shared[((((int)threadIdx.x) * 72) + 4668)]));
+    conv2d_nchw[10] = (conv2d_nchw[10] + (pad_temp_shared[183] * kernel_shared[((((int)threadIdx.x) * 72) + 4668)]));
+    conv2d_nchw[11] = (conv2d_nchw[11] + (pad_temp_shared[184] * kernel_shared[((((int)threadIdx.x) * 72) + 4668)]));
+    conv2d_nchw[12] = (conv2d_nchw[12] + (pad_temp_shared[185] * kernel_shared[((((int)threadIdx.x) * 72) + 4668)]));
+    conv2d_nchw[13] = (conv2d_nchw[13] + (pad_temp_shared[186] * kernel_shared[((((int)threadIdx.x) * 72) + 4668)]));
+    conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[181] * kernel_shared[((((int)threadIdx.x) * 72) + 61)]));
+    conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[182] * kernel_shared[((((int)threadIdx.x) * 72) + 61)]));
+    conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[183] * kernel_shared[((((int)threadIdx.x) * 72) + 61)]));
+    conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[184] * kernel_shared[((((int)threadIdx.x) * 72) + 61)]));
+    conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[185] * kernel_shared[((((int)threadIdx.x) * 72) + 61)]));
+    conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[186] * kernel_shared[((((int)threadIdx.x) * 72) + 61)]));
+    conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[187] * kernel_shared[((((int)threadIdx.x) * 72) + 61)]));
+    conv2d_nchw[7] = (conv2d_nchw[7] + (pad_temp_shared[181] * kernel_shared[((((int)threadIdx.x) * 72) + 4669)]));
+    conv2d_nchw[8] = (conv2d_nchw[8] + (pad_temp_shared[182] * kernel_shared[((((int)threadIdx.x) * 72) + 4669)]));
+    conv2d_nchw[9] = (conv2d_nchw[9] + (pad_temp_shared[183] * kernel_shared[((((int)threadIdx.x) * 72) + 4669)]));
+    conv2d_nchw[10] = (conv2d_nchw[10] + (pad_temp_shared[184] * kernel_shared[((((int)threadIdx.x) * 72) + 4669)]));
+    conv2d_nchw[11] = (conv2d_nchw[11] + (pad_temp_shared[185] * kernel_shared[((((int)threadIdx.x) * 72) + 4669)]));
+    conv2d_nchw[12] = (conv2d_nchw[12] + (pad_temp_shared[186] * kernel_shared[((((int)threadIdx.x) * 72) + 4669)]));
+    conv2d_nchw[13] = (conv2d_nchw[13] + (pad_temp_shared[187] * kernel_shared[((((int)threadIdx.x) * 72) + 4669)]));
+    conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[182] * kernel_shared[((((int)threadIdx.x) * 72) + 62)]));
+    conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[183] * kernel_shared[((((int)threadIdx.x) * 72) + 62)]));
+    conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[184] * kernel_shared[((((int)threadIdx.x) * 72) + 62)]));
+    conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[185] * kernel_shared[((((int)threadIdx.x) * 72) + 62)]));
+    conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[186] * kernel_shared[((((int)threadIdx.x) * 72) + 62)]));
+    conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[187] * kernel_shared[((((int)threadIdx.x) * 72) + 62)]));
+    conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[188] * kernel_shared[((((int)threadIdx.x) * 72) + 62)]));
+    conv2d_nchw[7] = (conv2d_nchw[7] + (pad_temp_shared[182] * kernel_shared[((((int)threadIdx.x) * 72) + 4670)]));
+    conv2d_nchw[8] = (conv2d_nchw[8] + (pad_temp_shared[183] * kernel_shared[((((int)threadIdx.x) * 72) + 4670)]));
+    conv2d_nchw[9] = (conv2d_nchw[9] + (pad_temp_shared[184] * kernel_shared[((((int)threadIdx.x) * 72) + 4670)]));
+    conv2d_nchw[10] = (conv2d_nchw[10] + (pad_temp_shared[185] * kernel_shared[((((int)threadIdx.x) * 72) + 4670)]));
+    conv2d_nchw[11] = (conv2d_nchw[11] + (pad_temp_shared[186] * kernel_shared[((((int)threadIdx.x) * 72) + 4670)]));
+    conv2d_nchw[12] = (conv2d_nchw[12] + (pad_temp_shared[187] * kernel_shared[((((int)threadIdx.x) * 72) + 4670)]));
+    conv2d_nchw[13] = (conv2d_nchw[13] + (pad_temp_shared[188] * kernel_shared[((((int)threadIdx.x) * 72) + 4670)]));
+    conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[189] * kernel_shared[((((int)threadIdx.x) * 72) + 63)]));
+    conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[190] * kernel_shared[((((int)threadIdx.x) * 72) + 63)]));
+    conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[191] * kernel_shared[((((int)threadIdx.x) * 72) + 63)]));
+    conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[192] * kernel_shared[((((int)threadIdx.x) * 72) + 63)]));
+    conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[193] * kernel_shared[((((int)threadIdx.x) * 72) + 63)]));
+    conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[194] * kernel_shared[((((int)threadIdx.x) * 72) + 63)]));
+    conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[195] * kernel_shared[((((int)threadIdx.x) * 72) + 63)]));
+    conv2d_nchw[7] = (conv2d_nchw[7] + (pad_temp_shared[189] * kernel_shared[((((int)threadIdx.x) * 72) + 4671)]));
+    conv2d_nchw[8] = (conv2d_nchw[8] + (pad_temp_shared[190] * kernel_shared[((((int)threadIdx.x) * 72) + 4671)]));
+    conv2d_nchw[9] = (conv2d_nchw[9] + (pad_temp_shared[191] * kernel_shared[((((int)threadIdx.x) * 72) + 4671)]));
+    conv2d_nchw[10] = (conv2d_nchw[10] + (pad_temp_shared[192] * kernel_shared[((((int)threadIdx.x) * 72) + 4671)]));
+    conv2d_nchw[11] = (conv2d_nchw[11] + (pad_temp_shared[193] * kernel_shared[((((int)threadIdx.x) * 72) + 4671)]));
+    conv2d_nchw[12] = (conv2d_nchw[12] + (pad_temp_shared[194] * kernel_shared[((((int)threadIdx.x) * 72) + 4671)]));
+    conv2d_nchw[13] = (conv2d_nchw[13] + (pad_temp_shared[195] * kernel_shared[((((int)threadIdx.x) * 72) + 4671)]));
+    conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[190] * kernel_shared[((((int)threadIdx.x) * 72) + 64)]));
+    conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[191] * kernel_shared[((((int)threadIdx.x) * 72) + 64)]));
+    conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[192] * kernel_shared[((((int)threadIdx.x) * 72) + 64)]));
+    conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[193] * kernel_shared[((((int)threadIdx.x) * 72) + 64)]));
+    conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[194] * kernel_shared[((((int)threadIdx.x) * 72) + 64)]));
+    conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[195] * kernel_shared[((((int)threadIdx.x) * 72) + 64)]));
+    conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[196] * kernel_shared[((((int)threadIdx.x) * 72) + 64)]));
+    conv2d_nchw[7] = (conv2d_nchw[7] + (pad_temp_shared[190] * kernel_shared[((((int)threadIdx.x) * 72) + 4672)]));
+    conv2d_nchw[8] = (conv2d_nchw[8] + (pad_temp_shared[191] * kernel_shared[((((int)threadIdx.x) * 72) + 4672)]));
+    conv2d_nchw[9] = (conv2d_nchw[9] + (pad_temp_shared[192] * kernel_shared[((((int)threadIdx.x) * 72) + 4672)]));
+    conv2d_nchw[10] = (conv2d_nchw[10] + (pad_temp_shared[193] * kernel_shared[((((int)threadIdx.x) * 72) + 4672)]));
+    conv2d_nchw[11] = (conv2d_nchw[11] + (pad_temp_shared[194] * kernel_shared[((((int)threadIdx.x) * 72) + 4672)]));
+    conv2d_nchw[12] = (conv2d_nchw[12] + (pad_temp_shared[195] * kernel_shared[((((int)threadIdx.x) * 72) + 4672)]));
+    conv2d_nchw[13] = (conv2d_nchw[13] + (pad_temp_shared[196] * kernel_shared[((((int)threadIdx.x) * 72) + 4672)]));
+    conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[191] * kernel_shared[((((int)threadIdx.x) * 72) + 65)]));
+    conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[192] * kernel_shared[((((int)threadIdx.x) * 72) + 65)]));
+    conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[193] * kernel_shared[((((int)threadIdx.x) * 72) + 65)]));
+    conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[194] * kernel_shared[((((int)threadIdx.x) * 72) + 65)]));
+    conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[195] * kernel_shared[((((int)threadIdx.x) * 72) + 65)]));
+    conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[196] * kernel_shared[((((int)threadIdx.x) * 72) + 65)]));
+    conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[197] * kernel_shared[((((int)threadIdx.x) * 72) + 65)]));
+    conv2d_nchw[7] = (conv2d_nchw[7] + (pad_temp_shared[191] * kernel_shared[((((int)threadIdx.x) * 72) + 4673)]));
+    conv2d_nchw[8] = (conv2d_nchw[8] + (pad_temp_shared[192] * kernel_shared[((((int)threadIdx.x) * 72) + 4673)]));
+    conv2d_nchw[9] = (conv2d_nchw[9] + (pad_temp_shared[193] * kernel_shared[((((int)threadIdx.x) * 72) + 4673)]));
+    conv2d_nchw[10] = (conv2d_nchw[10] + (pad_temp_shared[194] * kernel_shared[((((int)threadIdx.x) * 72) + 4673)]));
+    conv2d_nchw[11] = (conv2d_nchw[11] + (pad_temp_shared[195] * kernel_shared[((((int)threadIdx.x) * 72) + 4673)]));
+    conv2d_nchw[12] = (conv2d_nchw[12] + (pad_temp_shared[196] * kernel_shared[((((int)threadIdx.x) * 72) + 4673)]));
+    conv2d_nchw[13] = (conv2d_nchw[13] + (pad_temp_shared[197] * kernel_shared[((((int)threadIdx.x) * 72) + 4673)]));
+    conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[198] * kernel_shared[((((int)threadIdx.x) * 72) + 66)]));
+    conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[199] * kernel_shared[((((int)threadIdx.x) * 72) + 66)]));
+    conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[200] * kernel_shared[((((int)threadIdx.x) * 72) + 66)]));
+    conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[201] * kernel_shared[((((int)threadIdx.x) * 72) + 66)]));
+    conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[202] * kernel_shared[((((int)threadIdx.x) * 72) + 66)]));
+    conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[203] * kernel_shared[((((int)threadIdx.x) * 72) + 66)]));
+    conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[204] * kernel_shared[((((int)threadIdx.x) * 72) + 66)]));
+    conv2d_nchw[7] = (conv2d_nchw[7] + (pad_temp_shared[198] * kernel_shared[((((int)threadIdx.x) * 72) + 4674)]));
+    conv2d_nchw[8] = (conv2d_nchw[8] + (pad_temp_shared[199] * kernel_shared[((((int)threadIdx.x) * 72) + 4674)]));
+    conv2d_nchw[9] = (conv2d_nchw[9] + (pad_temp_shared[200] * kernel_shared[((((int)threadIdx.x) * 72) + 4674)]));
+    conv2d_nchw[10] = (conv2d_nchw[10] + (pad_temp_shared[201] * kernel_shared[((((int)threadIdx.x) * 72) + 4674)]));
+    conv2d_nchw[11] = (conv2d_nchw[11] + (pad_temp_shared[202] * kernel_shared[((((int)threadIdx.x) * 72) + 4674)]));
+    conv2d_nchw[12] = (conv2d_nchw[12] + (pad_temp_shared[203] * kernel_shared[((((int)threadIdx.x) * 72) + 4674)]));
+    conv2d_nchw[13] = (conv2d_nchw[13] + (pad_temp_shared[204] * kernel_shared[((((int)threadIdx.x) * 72) + 4674)]));
+    conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[199] * kernel_shared[((((int)threadIdx.x) * 72) + 67)]));
+    conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[200] * kernel_shared[((((int)threadIdx.x) * 72) + 67)]));
+    conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[201] * kernel_shared[((((int)threadIdx.x) * 72) + 67)]));
+    conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[202] * kernel_shared[((((int)threadIdx.x) * 72) + 67)]));
+    conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[203] * kernel_shared[((((int)threadIdx.x) * 72) + 67)]));
+    conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[204] * kernel_shared[((((int)threadIdx.x) * 72) + 67)]));
+    conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[205] * kernel_shared[((((int)threadIdx.x) * 72) + 67)]));
+    conv2d_nchw[7] = (conv2d_nchw[7] + (pad_temp_shared[199] * kernel_shared[((((int)threadIdx.x) * 72) + 4675)]));
+    conv2d_nchw[8] = (conv2d_nchw[8] + (pad_temp_shared[200] * kernel_shared[((((int)threadIdx.x) * 72) + 4675)]));
+    conv2d_nchw[9] = (conv2d_nchw[9] + (pad_temp_shared[201] * kernel_shared[((((int)threadIdx.x) * 72) + 4675)]));
+    conv2d_nchw[10] = (conv2d_nchw[10] + (pad_temp_shared[202] * kernel_shared[((((int)threadIdx.x) * 72) + 4675)]));
+    conv2d_nchw[11] = (conv2d_nchw[11] + (pad_temp_shared[203] * kernel_shared[((((int)threadIdx.x) * 72) + 4675)]));
+    conv2d_nchw[12] = (conv2d_nchw[12] + (pad_temp_shared[204] * kernel_shared[((((int)threadIdx.x) * 72) + 4675)]));
+    conv2d_nchw[13] = (conv2d_nchw[13] + (pad_temp_shared[205] * kernel_shared[((((int)threadIdx.x) * 72) + 4675)]));
+    conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[200] * kernel_shared[((((int)threadIdx.x) * 72) + 68)]));
+    conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[201] * kernel_shared[((((int)threadIdx.x) * 72) + 68)]));
+    conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[202] * kernel_shared[((((int)threadIdx.x) * 72) + 68)]));
+    conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[203] * kernel_shared[((((int)threadIdx.x) * 72) + 68)]));
+    conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[204] * kernel_shared[((((int)threadIdx.x) * 72) + 68)]));
+    conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[205] * kernel_shared[((((int)threadIdx.x) * 72) + 68)]));
+    conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[206] * kernel_shared[((((int)threadIdx.x) * 72) + 68)]));
+    conv2d_nchw[7] = (conv2d_nchw[7] + (pad_temp_shared[200] * kernel_shared[((((int)threadIdx.x) * 72) + 4676)]));
+    conv2d_nchw[8] = (conv2d_nchw[8] + (pad_temp_shared[201] * kernel_shared[((((int)threadIdx.x) * 72) + 4676)]));
+    conv2d_nchw[9] = (conv2d_nchw[9] + (pad_temp_shared[202] * kernel_shared[((((int)threadIdx.x) * 72) + 4676)]));
+    conv2d_nchw[10] = (conv2d_nchw[10] + (pad_temp_shared[203] * kernel_shared[((((int)threadIdx.x) * 72) + 4676)]));
+    conv2d_nchw[11] = (conv2d_nchw[11] + (pad_temp_shared[204] * kernel_shared[((((int)threadIdx.x) * 72) + 4676)]));
+    conv2d_nchw[12] = (conv2d_nchw[12] + (pad_temp_shared[205] * kernel_shared[((((int)threadIdx.x) * 72) + 4676)]));
+    conv2d_nchw[13] = (conv2d_nchw[13] + (pad_temp_shared[206] * kernel_shared[((((int)threadIdx.x) * 72) + 4676)]));
+    conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[207] * kernel_shared[((((int)threadIdx.x) * 72) + 69)]));
+    conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[208] * kernel_shared[((((int)threadIdx.x) * 72) + 69)]));
+    conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[209] * kernel_shared[((((int)threadIdx.x) * 72) + 69)]));
+    conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[210] * kernel_shared[((((int)threadIdx.x) * 72) + 69)]));
+    conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[211] * kernel_shared[((((int)threadIdx.x) * 72) + 69)]));
+    conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[212] * kernel_shared[((((int)threadIdx.x) * 72) + 69)]));
+    conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[213] * kernel_shared[((((int)threadIdx.x) * 72) + 69)]));
+    conv2d_nchw[7] = (conv2d_nchw[7] + (pad_temp_shared[207] * kernel_shared[((((int)threadIdx.x) * 72) + 4677)]));
+    conv2d_nchw[8] = (conv2d_nchw[8] + (pad_temp_shared[208] * kernel_shared[((((int)threadIdx.x) * 72) + 4677)]));
+    conv2d_nchw[9] = (conv2d_nchw[9] + (pad_temp_shared[209] * kernel_shared[((((int)threadIdx.x) * 72) + 4677)]));
+    conv2d_nchw[10] = (conv2d_nchw[10] + (pad_temp_shared[210] * kernel_shared[((((int)threadIdx.x) * 72) + 4677)]));
+    conv2d_nchw[11] = (conv2d_nchw[11] + (pad_temp_shared[211] * kernel_shared[((((int)threadIdx.x) * 72) + 4677)]));
+    conv2d_nchw[12] = (conv2d_nchw[12] + (pad_temp_shared[212] * kernel_shared[((((int)threadIdx.x) * 72) + 4677)]));
+    conv2d_nchw[13] = (conv2d_nchw[13] + (pad_temp_shared[213] * kernel_shared[((((int)threadIdx.x) * 72) + 4677)]));
+    conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[208] * kernel_shared[((((int)threadIdx.x) * 72) + 70)]));
+    conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[209] * kernel_shared[((((int)threadIdx.x) * 72) + 70)]));
+    conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[210] * kernel_shared[((((int)threadIdx.x) * 72) + 70)]));
+    conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[211] * kernel_shared[((((int)threadIdx.x) * 72) + 70)]));
+    conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[212] * kernel_shared[((((int)threadIdx.x) * 72) + 70)]));
+    conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[213] * kernel_shared[((((int)threadIdx.x) * 72) + 70)]));
+    conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[214] * kernel_shared[((((int)threadIdx.x) * 72) + 70)]));
+    conv2d_nchw[7] = (conv2d_nchw[7] + (pad_temp_shared[208] * kernel_shared[((((int)threadIdx.x) * 72) + 4678)]));
+    conv2d_nchw[8] = (conv2d_nchw[8] + (pad_temp_shared[209] * kernel_shared[((((int)threadIdx.x) * 72) + 4678)]));
+    conv2d_nchw[9] = (conv2d_nchw[9] + (pad_temp_shared[210] * kernel_shared[((((int)threadIdx.x) * 72) + 4678)]));
+    conv2d_nchw[10] = (conv2d_nchw[10] + (pad_temp_shared[211] * kernel_shared[((((int)threadIdx.x) * 72) + 4678)]));
+    conv2d_nchw[11] = (conv2d_nchw[11] + (pad_temp_shared[212] * kernel_shared[((((int)threadIdx.x) * 72) + 4678)]));
+    conv2d_nchw[12] = (conv2d_nchw[12] + (pad_temp_shared[213] * kernel_shared[((((int)threadIdx.x) * 72) + 4678)]));
+    conv2d_nchw[13] = (conv2d_nchw[13] + (pad_temp_shared[214] * kernel_shared[((((int)threadIdx.x) * 72) + 4678)]));
+    conv2d_nchw[0] = (conv2d_nchw[0] + (pad_temp_shared[209] * kernel_shared[((((int)threadIdx.x) * 72) + 71)]));
+    conv2d_nchw[1] = (conv2d_nchw[1] + (pad_temp_shared[210] * kernel_shared[((((int)threadIdx.x) * 72) + 71)]));
+    conv2d_nchw[2] = (conv2d_nchw[2] + (pad_temp_shared[211] * kernel_shared[((((int)threadIdx.x) * 72) + 71)]));
+    conv2d_nchw[3] = (conv2d_nchw[3] + (pad_temp_shared[212] * kernel_shared[((((int)threadIdx.x) * 72) + 71)]));
+    conv2d_nchw[4] = (conv2d_nchw[4] + (pad_temp_shared[213] * kernel_shared[((((int)threadIdx.x) * 72) + 71)]));
+    conv2d_nchw[5] = (conv2d_nchw[5] + (pad_temp_shared[214] * kernel_shared[((((int)threadIdx.x) * 72) + 71)]));
+    conv2d_nchw[6] = (conv2d_nchw[6] + (pad_temp_shared[215] * kernel_shared[((((int)threadIdx.x) * 72) + 71)]));
+    conv2d_nchw[7] = (conv2d_nchw[7] + (pad_temp_shared[209] * kernel_shared[((((int)threadIdx.x) * 72) + 4679)]));
+    conv2d_nchw[8] = (conv2d_nchw[8] + (pad_temp_shared[210] * kernel_shared[((((int)threadIdx.x) * 72) + 4679)]));
+    conv2d_nchw[9] = (conv2d_nchw[9] + (pad_temp_shared[211] * kernel_shared[((((int)threadIdx.x) * 72) + 4679)]));
+    conv2d_nchw[10] = (conv2d_nchw[10] + (pad_temp_shared[212] * kernel_shared[((((int)threadIdx.x) * 72) + 4679)]));
+    conv2d_nchw[11] = (conv2d_nchw[11] + (pad_temp_shared[213] * kernel_shared[((((int)threadIdx.x) * 72) + 4679)]));
+    conv2d_nchw[12] = (conv2d_nchw[12] + (pad_temp_shared[214] * kernel_shared[((((int)threadIdx.x) * 72) + 4679)]));
+    conv2d_nchw[13] = (conv2d_nchw[13] + (pad_temp_shared[215] * kernel_shared[((((int)threadIdx.x) * 72) + 4679)]));
   }
+  compute[((((((int)blockIdx.x) / 7) * 6272) + (((int)threadIdx.x) * 49)) + ((((int)blockIdx.x) % 7) * 7))] = max((conv2d_nchw[0] + bias[(((((int)blockIdx.x) / 7) * 128) + ((int)threadIdx.x))]), 0.000000e+00f);
+  compute[(((((((int)blockIdx.x) / 7) * 6272) + (((int)threadIdx.x) * 49)) + ((((int)blockIdx.x) % 7) * 7)) + 1)] = max((conv2d_nchw[1] + bias[(((((int)blockIdx.x) / 7) * 128) + ((int)threadIdx.x))]), 0.000000e+00f);
+  compute[(((((((int)blockIdx.x) / 7) * 6272) + (((int)threadIdx.x) * 49)) + ((((int)blockIdx.x) % 7) * 7)) + 2)] = max((conv2d_nchw[2] + bias[(((((int)blockIdx.x) / 7) * 128) + ((int)threadIdx.x))]), 0.000000e+00f);
+  compute[(((((((int)blockIdx.x) / 7) * 6272) + (((int)threadIdx.x) * 49)) + ((((int)blockIdx.x) % 7) * 7)) + 3)] = max((conv2d_nchw[3] + bias[(((((int)blockIdx.x) / 7) * 128) + ((int)threadIdx.x))]), 0.000000e+00f);
+  compute[(((((((int)blockIdx.x) / 7) * 6272) + (((int)threadIdx.x) * 49)) + ((((int)blockIdx.x) % 7) * 7)) + 4)] = max((conv2d_nchw[4] + bias[(((((int)blockIdx.x) / 7) * 128) + ((int)threadIdx.x))]), 0.000000e+00f);
+  compute[(((((((int)blockIdx.x) / 7) * 6272) + (((int)threadIdx.x) * 49)) + ((((int)blockIdx.x) % 7) * 7)) + 5)] = max((conv2d_nchw[5] + bias[(((((int)blockIdx.x) / 7) * 128) + ((int)threadIdx.x))]), 0.000000e+00f);
+  compute[(((((((int)blockIdx.x) / 7) * 6272) + (((int)threadIdx.x) * 49)) + ((((int)blockIdx.x) % 7) * 7)) + 6)] = max((conv2d_nchw[6] + bias[(((((int)blockIdx.x) / 7) * 128) + ((int)threadIdx.x))]), 0.000000e+00f);
+  compute[(((((((int)blockIdx.x) / 7) * 6272) + (((int)threadIdx.x) * 49)) + ((((int)blockIdx.x) % 7) * 7)) + 3136)] = max((conv2d_nchw[7] + bias[((((((int)blockIdx.x) / 7) * 128) + ((int)threadIdx.x)) + 64)]), 0.000000e+00f);
+  compute[(((((((int)blockIdx.x) / 7) * 6272) + (((int)threadIdx.x) * 49)) + ((((int)blockIdx.x) % 7) * 7)) + 3137)] = max((conv2d_nchw[8] + bias[((((((int)blockIdx.x) / 7) * 128) + ((int)threadIdx.x)) + 64)]), 0.000000e+00f);
+  compute[(((((((int)blockIdx.x) / 7) * 6272) + (((int)threadIdx.x) * 49)) + ((((int)blockIdx.x) % 7) * 7)) + 3138)] = max((conv2d_nchw[9] + bias[((((((int)blockIdx.x) / 7) * 128) + ((int)threadIdx.x)) + 64)]), 0.000000e+00f);
+  compute[(((((((int)blockIdx.x) / 7) * 6272) + (((int)threadIdx.x) * 49)) + ((((int)blockIdx.x) % 7) * 7)) + 3139)] = max((conv2d_nchw[10] + bias[((((((int)blockIdx.x) / 7) * 128) + ((int)threadIdx.x)) + 64)]), 0.000000e+00f);
+  compute[(((((((int)blockIdx.x) / 7) * 6272) + (((int)threadIdx.x) * 49)) + ((((int)blockIdx.x) % 7) * 7)) + 3140)] = max((conv2d_nchw[11] + bias[((((((int)blockIdx.x) / 7) * 128) + ((int)threadIdx.x)) + 64)]), 0.000000e+00f);
+  compute[(((((((int)blockIdx.x) / 7) * 6272) + (((int)threadIdx.x) * 49)) + ((((int)blockIdx.x) % 7) * 7)) + 3141)] = max((conv2d_nchw[12] + bias[((((((int)blockIdx.x) / 7) * 128) + ((int)threadIdx.x)) + 64)]), 0.000000e+00f);
+  compute[(((((((int)blockIdx.x) / 7) * 6272) + (((int)threadIdx.x) * 49)) + ((((int)blockIdx.x) % 7) * 7)) + 3142)] = max((conv2d_nchw[13] + bias[((((((int)blockIdx.x) / 7) * 128) + ((int)threadIdx.x)) + 64)]), 0.000000e+00f);
 }
 </pre></div>
 </div>
@@ -1583,7 +3216,7 @@ In the example below we resume the status and do more 5 trials.</p>
 Get devices for measurement successfully!
 </pre></div>
 </div>
-<p class="sphx-glr-timing"><strong>Total running time of the script:</strong> ( 5 minutes  43.471 seconds)</p>
+<p class="sphx-glr-timing"><strong>Total running time of the script:</strong> ( 5 minutes  37.856 seconds)</p>
 <div class="sphx-glr-footer sphx-glr-footer-example docutils container" id="sphx-glr-download-how-to-tune-with-autoscheduler-tune-conv2d-layer-cuda-py">
 <div class="sphx-glr-download sphx-glr-download-python docutils container">
 <p><a class="reference download internal" download="" href="../../_downloads/e3e540f3b477c0c52d8eb73e674e8ffd/tune_conv2d_layer_cuda.py"><code class="xref download docutils literal notranslate"><span class="pre">Download</span> <span class="pre">Python</span> <span class="pre">source</span> <span class="pre">code:</span> <span class="pre">tune_conv2d_layer_cuda.py</span></code></a></p>
diff --git a/docs/how_to/tune_with_autoscheduler/tune_network_cuda.html b/docs/how_to/tune_with_autoscheduler/tune_network_cuda.html
index 3fe13512c2..d9a33aeb9c 100644
--- a/docs/how_to/tune_with_autoscheduler/tune_network_cuda.html
+++ b/docs/how_to/tune_with_autoscheduler/tune_network_cuda.html
@@ -916,7 +916,7 @@ so we can read the log file and load the best schedules.</p>
 Evaluate inference time cost...
 Execution time summary:
... 3757 lines suppressed ...