You are viewing a plain text version of this content. The canonical link for it is here.
Posted to commits@mxnet.apache.org by jx...@apache.org on 2018/05/15 18:23:37 UTC
[incubator-mxnet] branch master updated: Update performance page.
(#10876)
This is an automated email from the ASF dual-hosted git repository.
jxie pushed a commit to branch master
in repository https://gitbox.apache.org/repos/asf/incubator-mxnet.git
The following commit(s) were added to refs/heads/master by this push:
new db4de44 Update performance page. (#10876)
db4de44 is described below
commit db4de44f170df08a1500feabccb55a63266d7ae2
Author: Da Zheng <zh...@gmail.com>
AuthorDate: Tue May 15 11:23:32 2018 -0700
Update performance page. (#10876)
* Update perf.
* add float16.
* update header.
---
docs/faq/perf.md | 56 ++++++++++++++++++++++++++++++++++++++++----------------
1 file changed, 40 insertions(+), 16 deletions(-)
diff --git a/docs/faq/perf.md b/docs/faq/perf.md
index ce74391..ad81b5d 100644
--- a/docs/faq/perf.md
+++ b/docs/faq/perf.md
@@ -36,7 +36,7 @@ to measure the performance on different AWS EC2 machines.
AWS EC2 C5.18xlarge:
-| Batch | Alexnet | VGG | Inception-BN | Inception-v3 | Resnet 50 | Resnet 152 |
+| Batch | Alexnet | VGG 16 | Inception-BN | Inception-v3 | Resnet 50 | Resnet 152 |
|-------|---------|--------|--------------|--------------|-----------|------------|
| 1 | 390.53 | 81.57 | 124.13 | 62.26 | 76.22 | 32.92 |
| 2 | 596.45 | 100.84 | 206.58 | 93.36 | 119.55 | 46.80 |
@@ -48,7 +48,7 @@ AWS EC2 C5.18xlarge:
AWS EC2 C5.9xlarge:
-| Batch | Alexnet | VGG | Inception-BN | Inception-v3 | Resnet 50 | Resnet 152 |
+| Batch | Alexnet | VGG 16 | Inception-BN | Inception-v3 | Resnet 50 | Resnet 152 |
|-------|---------|-------|--------------|--------------|-----------|------------|
| 1 | 257.77 | 50.61 | 130.99 | 66.95 | 75.38 | 32.33 |
| 2 | 410.60 | 63.02 | 195.14 | 87.84 | 102.67 | 41.57 |
@@ -60,7 +60,7 @@ AWS EC2 C5.9xlarge:
AWS EC2 C5.4xlarge:
-| Batch | Alexnet | VGG | Inception-BN | Inception-v3 | Resnet 50 | Resnet 152 |
+| Batch | Alexnet | VGG 16 | Inception-BN | Inception-v3 | Resnet 50 | Resnet 152 |
|-------|---------|-------|--------------|--------------|-----------|------------|
| 1 | 214.15 | 29.32 | 114.97 | 47.96 | 61.01 | 23.92 |
| 2 | 310.04 | 34.81 | 150.09 | 60.89 | 71.16 | 27.92 |
@@ -72,7 +72,7 @@ AWS EC2 C5.4xlarge:
AWS EC2 C5.2xlarge:
-| Batch | Alexnet | VGG | Inception-BN | Inception-v3 | Resnet 50 | Resnet 152 |
+| Batch | Alexnet | VGG 16 | Inception-BN | Inception-v3 | Resnet 50 | Resnet 152 |
|-------|---------|-------|--------------|--------------|-----------|------------|
| 1 | 131.01 | 15.67 | 78.75 | 31.12 | 37.30 | 14.75 |
| 2 | 182.29 | 18.01 | 98.59 | 39.13 | 45.98 | 17.84 |
@@ -84,7 +84,7 @@ AWS EC2 C5.2xlarge:
AWS EC2 C5.xlarge:
-| Batch | Alexnet | VGG | Inception-BN | Inception-v3 | Resnet 50 | Resnet 152 |
+| Batch | Alexnet | VGG 16 | Inception-BN | Inception-v3 | Resnet 50 | Resnet 152 |
|-------|---------|------|--------------|--------------|-----------|------------|
| 1 | 36.64 | 3.93 | 27.06 | 10.09 | 12.98 | 5.06 |
| 2 | 49.21 | 4.49 | 29.67 | 10.80 | 12.94 | 5.14 |
@@ -117,7 +117,7 @@ and [MXNet-1.2.0.rc1](https://github.com/apache/incubator-mxnet/releases/downlo
- K80 (single GPU)
-| Batch | Alexnet | VGG | Inception-BN | Inception-v3 | Resnet 50 | Resnet 152 |
+| Batch | Alexnet | VGG 16 | Inception-BN | Inception-v3 | Resnet 50 | Resnet 152 |
|-------|---------|--------|--------------|--------------|-----------|------------|
| 1 | 243.93 | 43.59 | 68.62 | 35.52 | 67.41 | 23.65 |
| 2 | 338.16 | 49.14 | 113.41 | 56.29 | 93.35 | 33.88 |
@@ -125,10 +125,13 @@ and [MXNet-1.2.0.rc1](https://github.com/apache/incubator-mxnet/releases/downlo
| 8 | 683.52 | 70.50 | 190.49 | 86.23 | 131.32 | 50.54 |
| 16 | 1004.66 | 109.01 | 254.20 | 105.70 | 155.40 | 62.55 |
| 32 | 1238.55 | 114.98 | 285.49 | 116.79 | 159.42 | 64.99 |
+| 64 | 1346.72 | 123.56 | 308.73 | 122.21 | 167.58 | 70.21 |
+| 128 | 1416.91 | OOM | 320.98 | 123.11 | 171.55 | 71.85 |
+| 256 | 1462.97 | OOM | 329.16 | 127.53 | 153.01 | 57.23 |
- M60
-| Batch | Alexnet | VGG | Inception-BN | Inception-v3 | Resnet 50 | Resnet 152 |
+| Batch | Alexnet | VGG 16 | Inception-BN | Inception-v3 | Resnet 50 | Resnet 152 |
|-------|---------|--------|--------------|--------------|-----------|------------|
| 1 | 243.49 | 59.95 | 101.97 | 48.30 | 95.46 | 39.29 |
| 2 | 491.04 | 69.14 | 170.35 | 80.27 | 142.61 | 60.17 |
@@ -136,19 +139,38 @@ and [MXNet-1.2.0.rc1](https://github.com/apache/incubator-mxnet/releases/downlo
| 8 | 1077.73 | 109.34 | 343.42 | 152.82 | 208.74 | 87.27 |
| 16 | 1447.21 | 144.93 | 390.25 | 166.32 | 220.73 | 92.41 |
| 32 | 1797.66 | 151.86 | 416.69 | 176.56 | 230.19 | 97.03 |
+| 64 | 1779.38 | 150.18 | 427.51 | 183.47 | 239.12 | 101.59 |
+| 128 | 1787.36 | OOM | 439.04 | 185.29 | 243.31 | 103.39 |
+| 256 | 1899.10 | OOM | 450.22 | 183.42 | 242.36 | 100.98 |
- V100
-| Batch | Alexnet | VGG | Inception-BN | Inception-v3 | Resnet 50 | Resnet 152 |
+| Batch | Alexnet | VGG 16 | Inception-BN | Inception-v3 | Resnet 50 | Resnet 152 |
|-------|---------|--------|--------------|--------------|-----------|------------|
-| 1 | 659.51 | 205.16 | 136.91 | 76.54 | 162.15 | 61.38 |
-| 2 | 1248.21 | 265.40 | 261.85 | 144.23 | 293.74 | 116.30 |
-| 4 | 2122.41 | 333.97 | 477.22 | 270.03 | 479.14 | 195.17 |
-| 8 | 3894.30 | 420.26 | 831.09 | 450.68 | 699.39 | 294.19 |
-| 16 | 5815.58 | 654.16 | 1332.26 | 658.97 | 947.45 | 398.79 |
-| 32 | 7906.09 | 708.43 | 1784.23 | 817.33 | 1076.81 | 451.82 |
-
+| 1 | 659.51 | 205.16 | 157.37 | 87.71 | 162.15 | 61.38 |
+| 2 | 1248.21 | 265.40 | 297.34 | 159.24 | 293.74 | 116.30 |
+| 4 | 2122.41 | 333.97 | 520.91 | 279.84 | 479.14 | 195.17 |
+| 8 | 3894.30 | 420.26 | 898.09 | 455.03 | 699.39 | 294.19 |
+| 16 | 5815.58 | 654.16 | 1430.97 | 672.54 | 947.45 | 398.79 |
+| 32 | 7906.09 | 708.43 | 1847.26 | 814.59 | 1076.81 | 451.82 |
+| 64 | 9486.26 | 701.59 | 2134.89 | 899.01 | 1168.37 | 480.44 |
+| 128 | 10177.84 | 703.30 | 2318.32 | 904.33 | 1233.15 | 511.79 |
+| 256 | 10990.46 | 473.62 | 2425.28 | 960.20 | 1155.07 | 449.35 |
+
+Below is the performance result on V100 using float 16.
+
+| Batch | VGG 16 | Inception-BN | Inception-v3 | Resnet 50 | Resnet 152 |
+| ----- | ------- | ------------ | ------------ | --------- | ---------- |
+| 1 | 276.29 | 155.53 | 150.99 | 270.89 | 96.79 |
+| 2 | 476.91 | 296.45 | 282.02 | 493.99 | 176.88 |
+| 4 | 711.92 | 525.05 | 492.45 | 851.15 | 321.52 |
+| 8 | 1047.11 | 900.26 | 807.94 | 1282.36 | 517.66 |
+| 16 | 1299.88 | 1441.41 | 1192.21 | 1722.97 | 724.57 |
+| 32 | 1486.63 | 1854.30 | 1512.08 | 2085.51 | 887.34 |
+| 64 | 1219.65 | 2138.61 | 1687.35 | 2341.67 | 1002.90 |
+| 128 | 1169.81 | 2317.39 | 1818.26 | 2355.04 | 1046.98 |
+| 256 | 764.16 | 2425.16 | 1653.74 | 1991.88 | 976.73 |
### Training results
@@ -160,7 +182,7 @@ where the batch size for Alexnet is increased by 16x.
- K80 (single GPU)
- | Batch | Alexnet(\*8) | Inception-v3 | Resnet 50 |
+ | Batch | Alexnet(\*16) | Inception-v3 | Resnet 50 |
| --- | --- | --- | --- |
| 1 | 300.30 | 10.48 | 15.61 |
| 2 | 406.08 | 16.00 | 23.88 |
@@ -190,6 +212,8 @@ where the batch size for Alexnet is increased by 16x.
| 8 | 2919.02 | 118.43 | 174.81 |
| 16 | 2994.32 | 173.15 | 251.22 |
| 32 | 2585.61 | 214.48 | 298.51 |
+ | 64 | 1984.21 | 247.43 | 343.19 |
+ | 128 | OOM | 253.68 | 363.69 |
## Multiple Devices
--
To stop receiving notification emails like this one, please contact
jxie@apache.org.