You are viewing a plain text version of this content. The canonical link for it is here.
Posted to github@arrow.apache.org by GitBox <gi...@apache.org> on 2022/07/20 13:23:40 UTC
[GitHub] [arrow] pitrou commented on pull request #13654: ARROW-17135: [C++] Reduce code size in compute/kernels/scalar_compare.cc
pitrou commented on PR #13654:
URL: https://github.com/apache/arrow/pull/13654#issuecomment-1190282677
Benchmark results on my work machine (a AMD Zen 2 CPU):
```
-----------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------
Non-regressions: (46)
-----------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------
benchmark baseline contender change % counters
GreaterArrayScalarInt64/32768/0 1.519G items/sec 2.887G items/sec 89.993 {'family_index': 1, 'per_family_instance_index': 5, 'run_name': 'GreaterArrayScalarInt64/32768/0', 'repetitions': 1, 'repetition_index': 0, 'threads': 1, 'iterations': 32357, 'null_percent': 0.0}
GreaterArrayScalarInt64/32768/10000 1.493G items/sec 2.797G items/sec 87.275 {'family_index': 1, 'per_family_instance_index': 0, 'run_name': 'GreaterArrayScalarInt64/32768/10000', 'repetitions': 1, 'repetition_index': 0, 'threads': 1, 'iterations': 31872, 'null_percent': 0.01}
GreaterArrayScalarInt64/32768/2 1.495G items/sec 2.783G items/sec 86.147 {'family_index': 1, 'per_family_instance_index': 3, 'run_name': 'GreaterArrayScalarInt64/32768/2', 'repetitions': 1, 'repetition_index': 0, 'threads': 1, 'iterations': 31886, 'null_percent': 50.0}
GreaterArrayScalarInt64/32768/10 1.496G items/sec 2.755G items/sec 84.135 {'family_index': 1, 'per_family_instance_index': 2, 'run_name': 'GreaterArrayScalarInt64/32768/10', 'repetitions': 1, 'repetition_index': 0, 'threads': 1, 'iterations': 31902, 'null_percent': 10.0}
GreaterArrayScalarInt64/32768/100 1.496G items/sec 2.730G items/sec 82.564 {'family_index': 1, 'per_family_instance_index': 1, 'run_name': 'GreaterArrayScalarInt64/32768/100', 'repetitions': 1, 'repetition_index': 0, 'threads': 1, 'iterations': 31822, 'null_percent': 1.0}
GreaterArrayScalarInt64/32768/1 1.496G items/sec 2.726G items/sec 82.191 {'family_index': 1, 'per_family_instance_index': 4, 'run_name': 'GreaterArrayScalarInt64/32768/1', 'repetitions': 1, 'repetition_index': 0, 'threads': 1, 'iterations': 31870, 'null_percent': 100.0}
GreaterArrayArrayInt64/32768/0 1.459G items/sec 2.575G items/sec 76.419 {'family_index': 0, 'per_family_instance_index': 5, 'run_name': 'GreaterArrayArrayInt64/32768/0', 'repetitions': 1, 'repetition_index': 0, 'threads': 1, 'iterations': 31129, 'null_percent': 0.0}
GreaterArrayArrayInt64/32768/100 1.439G items/sec 2.494G items/sec 73.355 {'family_index': 0, 'per_family_instance_index': 1, 'run_name': 'GreaterArrayArrayInt64/32768/100', 'repetitions': 1, 'repetition_index': 0, 'threads': 1, 'iterations': 30590, 'null_percent': 1.0}
GreaterArrayArrayInt64/32768/10000 1.438G items/sec 2.489G items/sec 73.105 {'family_index': 0, 'per_family_instance_index': 0, 'run_name': 'GreaterArrayArrayInt64/32768/10000', 'repetitions': 1, 'repetition_index': 0, 'threads': 1, 'iterations': 30442, 'null_percent': 0.01}
GreaterArrayArrayInt64/32768/2 1.437G items/sec 2.487G items/sec 73.037 {'family_index': 0, 'per_family_instance_index': 3, 'run_name': 'GreaterArrayArrayInt64/32768/2', 'repetitions': 1, 'repetition_index': 0, 'threads': 1, 'iterations': 30627, 'null_percent': 50.0}
GreaterArrayArrayInt64/32768/10 1.435G items/sec 2.473G items/sec 72.307 {'family_index': 0, 'per_family_instance_index': 2, 'run_name': 'GreaterArrayArrayInt64/32768/10', 'repetitions': 1, 'repetition_index': 0, 'threads': 1, 'iterations': 30633, 'null_percent': 10.0}
GreaterArrayArrayInt64/32768/1 1.436G items/sec 2.453G items/sec 70.814 {'family_index': 0, 'per_family_instance_index': 4, 'run_name': 'GreaterArrayArrayInt64/32768/1', 'repetitions': 1, 'repetition_index': 0, 'threads': 1, 'iterations': 30647, 'null_percent': 100.0}
MaxElementWiseArrayScalarInt64/32768/10000 951.172M items/sec 1.091G items/sec 14.717 {'family_index': 5, 'per_family_instance_index': 0, 'run_name': 'MaxElementWiseArrayScalarInt64/32768/10000', 'repetitions': 1, 'repetition_index': 0, 'threads': 1, 'iterations': 20443, 'null_percent': 0.01}
MaxElementWiseArrayArrayInt64/32768/0 628.126M items/sec 716.152M items/sec 14.014 {'family_index': 4, 'per_family_instance_index': 5, 'run_name': 'MaxElementWiseArrayArrayInt64/32768/0', 'repetitions': 1, 'repetition_index': 0, 'threads': 1, 'iterations': 13342, 'null_percent': 0.0}
MaxElementWiseArrayScalarInt64/32768/0 1.129G items/sec 1.277G items/sec 13.103 {'family_index': 5, 'per_family_instance_index': 5, 'run_name': 'MaxElementWiseArrayScalarInt64/32768/0', 'repetitions': 1, 'repetition_index': 0, 'threads': 1, 'iterations': 23890, 'null_percent': 0.0}
MaxElementWiseArrayScalarInt64/32768/100 797.989M items/sec 861.147M items/sec 7.915 {'family_index': 5, 'per_family_instance_index': 1, 'run_name': 'MaxElementWiseArrayScalarInt64/32768/100', 'repetitions': 1, 'repetition_index': 0, 'threads': 1, 'iterations': 16979, 'null_percent': 1.0}
MaxElementWiseArrayScalarInt64/32768/1 4.785G items/sec 5.135G items/sec 7.310 {'family_index': 5, 'per_family_instance_index': 4, 'run_name': 'MaxElementWiseArrayScalarInt64/32768/1', 'repetitions': 1, 'repetition_index': 0, 'threads': 1, 'iterations': 102277, 'null_percent': 100.0}
MaxElementWiseArrayArrayInt64/32768/1 4.046G items/sec 4.326G items/sec 6.923 {'family_index': 4, 'per_family_instance_index': 4, 'run_name': 'MaxElementWiseArrayArrayInt64/32768/1', 'repetitions': 1, 'repetition_index': 0, 'threads': 1, 'iterations': 86088, 'null_percent': 100.0}
MaxElementWiseArrayScalarInt64/32768/2 230.920M items/sec 246.430M items/sec 6.717 {'family_index': 5, 'per_family_instance_index': 3, 'run_name': 'MaxElementWiseArrayScalarInt64/32768/2', 'repetitions': 1, 'repetition_index': 0, 'threads': 1, 'iterations': 4903, 'null_percent': 50.0}
MaxElementWiseArrayArrayInt64/32768/2 115.210M items/sec 122.406M items/sec 6.247 {'family_index': 4, 'per_family_instance_index': 3, 'run_name': 'MaxElementWiseArrayArrayInt64/32768/2', 'repetitions': 1, 'repetition_index': 0, 'threads': 1, 'iterations': 2454, 'null_percent': 50.0}
GreaterArrayArrayString/32768/0 123.254M items/sec 126.920M items/sec 2.974 {'family_index': 2, 'per_family_instance_index': 5, 'run_name': 'GreaterArrayArrayString/32768/0', 'repetitions': 1, 'repetition_index': 0, 'threads': 1, 'iterations': 2622, 'null_percent': 0.0}
MaxElementWiseArrayArrayInt64/32768/10000 455.482M items/sec 468.196M items/sec 2.791 {'family_index': 4, 'per_family_instance_index': 0, 'run_name': 'MaxElementWiseArrayArrayInt64/32768/10000', 'repetitions': 1, 'repetition_index': 0, 'threads': 1, 'iterations': 9743, 'null_percent': 0.01}
MaxElementWiseArrayArrayInt64/32768/100 350.201M items/sec 358.919M items/sec 2.490 {'family_index': 4, 'per_family_instance_index': 1, 'run_name': 'MaxElementWiseArrayArrayInt64/32768/100', 'repetitions': 1, 'repetition_index': 0, 'threads': 1, 'iterations': 7471, 'null_percent': 1.0}
GreaterArrayScalarString/32768/10000 780.553M items/sec 799.631M items/sec 2.444 {'family_index': 3, 'per_family_instance_index': 0, 'run_name': 'GreaterArrayScalarString/32768/10000', 'repetitions': 1, 'repetition_index': 0, 'threads': 1, 'iterations': 16635, 'null_percent': 0.01}
GreaterArrayScalarString/32768/0 786.099M items/sec 804.830M items/sec 2.383 {'family_index': 3, 'per_family_instance_index': 5, 'run_name': 'GreaterArrayScalarString/32768/0', 'repetitions': 1, 'repetition_index': 0, 'threads': 1, 'iterations': 16776, 'null_percent': 0.0}
GreaterArrayScalarString/32768/1 779.519M items/sec 797.986M items/sec 2.369 {'family_index': 3, 'per_family_instance_index': 4, 'run_name': 'GreaterArrayScalarString/32768/1', 'repetitions': 1, 'repetition_index': 0, 'threads': 1, 'iterations': 16620, 'null_percent': 100.0}
GreaterArrayScalarString/32768/10 780.164M items/sec 798.110M items/sec 2.300 {'family_index': 3, 'per_family_instance_index': 2, 'run_name': 'GreaterArrayScalarString/32768/10', 'repetitions': 1, 'repetition_index': 0, 'threads': 1, 'iterations': 16631, 'null_percent': 10.0}
MaxElementWiseArrayArrayInt64/32768/10 231.824M items/sec 236.831M items/sec 2.160 {'family_index': 4, 'per_family_instance_index': 2, 'run_name': 'MaxElementWiseArrayArrayInt64/32768/10', 'repetitions': 1, 'repetition_index': 0, 'threads': 1, 'iterations': 4934, 'null_percent': 10.0}
GreaterArrayScalarString/32768/2 780.730M items/sec 796.308M items/sec 1.995 {'family_index': 3, 'per_family_instance_index': 3, 'run_name': 'GreaterArrayScalarString/32768/2', 'repetitions': 1, 'repetition_index': 0, 'threads': 1, 'iterations': 16624, 'null_percent': 50.0}
GreaterArrayScalarString/32768/100 780.414M items/sec 795.846M items/sec 1.977 {'family_index': 3, 'per_family_instance_index': 1, 'run_name': 'GreaterArrayScalarString/32768/100', 'repetitions': 1, 'repetition_index': 0, 'threads': 1, 'iterations': 16636, 'null_percent': 1.0}
MaxElementWiseArrayScalarInt64/32768/10 603.255M items/sec 612.917M items/sec 1.602 {'family_index': 5, 'per_family_instance_index': 2, 'run_name': 'MaxElementWiseArrayScalarInt64/32768/10', 'repetitions': 1, 'repetition_index': 0, 'threads': 1, 'iterations': 12784, 'null_percent': 10.0}
MaxElementWiseArrayArrayString/32768/10000 41.111M items/sec 41.547M items/sec 1.061 {'family_index': 6, 'per_family_instance_index': 0, 'run_name': 'MaxElementWiseArrayArrayString/32768/10000', 'repetitions': 1, 'repetition_index': 0, 'threads': 1, 'iterations': 873, 'null_percent': 0.01}
MaxElementWiseArrayArrayString/32768/100 41.341M items/sec 41.639M items/sec 0.721 {'family_index': 6, 'per_family_instance_index': 1, 'run_name': 'MaxElementWiseArrayArrayString/32768/100', 'repetitions': 1, 'repetition_index': 0, 'threads': 1, 'iterations': 860, 'null_percent': 1.0}
MaxElementWiseArrayScalarString/32768/1 90.721M items/sec 90.225M items/sec -0.547 {'family_index': 7, 'per_family_instance_index': 4, 'run_name': 'MaxElementWiseArrayScalarString/32768/1', 'repetitions': 1, 'repetition_index': 0, 'threads': 1, 'iterations': 1935, 'null_percent': 100.0}
GreaterArrayArrayString/32768/1 653.076M items/sec 648.121M items/sec -0.759 {'family_index': 2, 'per_family_instance_index': 4, 'run_name': 'GreaterArrayArrayString/32768/1', 'repetitions': 1, 'repetition_index': 0, 'threads': 1, 'iterations': 13918, 'null_percent': 100.0}
GreaterArrayArrayString/32768/100 122.658M items/sec 121.365M items/sec -1.055 {'family_index': 2, 'per_family_instance_index': 1, 'run_name': 'GreaterArrayArrayString/32768/100', 'repetitions': 1, 'repetition_index': 0, 'threads': 1, 'iterations': 2629, 'null_percent': 1.0}
MaxElementWiseArrayArrayString/32768/2 42.217M items/sec 41.698M items/sec -1.230 {'family_index': 6, 'per_family_instance_index': 3, 'run_name': 'MaxElementWiseArrayArrayString/32768/2', 'repetitions': 1, 'repetition_index': 0, 'threads': 1, 'iterations': 887, 'null_percent': 50.0}
GreaterArrayArrayString/32768/10 121.931M items/sec 120.381M items/sec -1.271 {'family_index': 2, 'per_family_instance_index': 2, 'run_name': 'GreaterArrayArrayString/32768/10', 'repetitions': 1, 'repetition_index': 0, 'threads': 1, 'iterations': 2593, 'null_percent': 10.0}
MaxElementWiseArrayArrayString/32768/1 183.437M items/sec 181.061M items/sec -1.295 {'family_index': 6, 'per_family_instance_index': 4, 'run_name': 'MaxElementWiseArrayArrayString/32768/1', 'repetitions': 1, 'repetition_index': 0, 'threads': 1, 'iterations': 3919, 'null_percent': 100.0}
GreaterArrayArrayString/32768/10000 123.361M items/sec 121.560M items/sec -1.460 {'family_index': 2, 'per_family_instance_index': 0, 'run_name': 'GreaterArrayArrayString/32768/10000', 'repetitions': 1, 'repetition_index': 0, 'threads': 1, 'iterations': 2631, 'null_percent': 0.01}
GreaterArrayArrayString/32768/2 208.825M items/sec 205.202M items/sec -1.735 {'family_index': 2, 'per_family_instance_index': 3, 'run_name': 'GreaterArrayArrayString/32768/2', 'repetitions': 1, 'repetition_index': 0, 'threads': 1, 'iterations': 4440, 'null_percent': 50.0}
MaxElementWiseArrayArrayString/32768/10 41.613M items/sec 40.830M items/sec -1.881 {'family_index': 6, 'per_family_instance_index': 2, 'run_name': 'MaxElementWiseArrayArrayString/32768/10', 'repetitions': 1, 'repetition_index': 0, 'threads': 1, 'iterations': 881, 'null_percent': 10.0}
MaxElementWiseArrayArrayString/32768/0 42.734M items/sec 41.408M items/sec -3.104 {'family_index': 6, 'per_family_instance_index': 5, 'run_name': 'MaxElementWiseArrayArrayString/32768/0', 'repetitions': 1, 'repetition_index': 0, 'threads': 1, 'iterations': 901, 'null_percent': 0.0}
MaxElementWiseArrayScalarString/32768/10000 55.931M items/sec 53.645M items/sec -4.088 {'family_index': 7, 'per_family_instance_index': 0, 'run_name': 'MaxElementWiseArrayScalarString/32768/10000', 'repetitions': 1, 'repetition_index': 0, 'threads': 1, 'iterations': 1185, 'null_percent': 0.01}
MaxElementWiseArrayScalarString/32768/100 55.952M items/sec 53.643M items/sec -4.126 {'family_index': 7, 'per_family_instance_index': 1, 'run_name': 'MaxElementWiseArrayScalarString/32768/100', 'repetitions': 1, 'repetition_index': 0, 'threads': 1, 'iterations': 1171, 'null_percent': 1.0}
MaxElementWiseArrayScalarString/32768/10 55.265M items/sec 52.728M items/sec -4.592 {'family_index': 7, 'per_family_instance_index': 2, 'run_name': 'MaxElementWiseArrayScalarString/32768/10', 'repetitions': 1, 'repetition_index': 0, 'threads': 1, 'iterations': 1171, 'null_percent': 10.0}
-------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------
Regressions: (2)
-------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------
benchmark baseline contender change % counters
MaxElementWiseArrayScalarString/32768/2 53.362M items/sec 50.664M items/sec -5.055 {'family_index': 7, 'per_family_instance_index': 3, 'run_name': 'MaxElementWiseArrayScalarString/32768/2', 'repetitions': 1, 'repetition_index': 0, 'threads': 1, 'iterations': 1133, 'null_percent': 50.0}
MaxElementWiseArrayScalarString/32768/0 57.910M items/sec 53.893M items/sec -6.937 {'family_index': 7, 'per_family_instance_index': 5, 'run_name': 'MaxElementWiseArrayScalarString/32768/0', 'repetitions': 1, 'repetition_index': 0, 'threads': 1, 'iterations': 1208, 'null_percent': 0.0}
```
--
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.
To unsubscribe, e-mail: github-unsubscribe@arrow.apache.org
For queries about this service, please contact Infrastructure at:
users@infra.apache.org