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 2020/06/02 01:40:18 UTC

[GitHub] [arrow] ursabot commented on pull request #7314: ARROW-8996: [C++] SSE runtime support for aggregate sum dense kernel

ursabot commented on pull request #7314:
URL: https://github.com/apache/arrow/pull/7314#issuecomment-637216678


   [AMD64 Ubuntu 18.04 C++ Benchmark (#108899)](https://ci.ursalabs.org/#builders/73/builds/70) builder has been succeeded.
   
   Revision: d789801a5e83b9717b8f36d72189198f5527b757
   
   ```diff
     ========================  ==============  ==============  ========
     benchmark                 baseline        contender       change
     ========================  ==============  ==============  ========
     SumKernelInt64/32768/1    10.639 GiB/sec  10.196 GiB/sec  -4.168%
     SumKernelInt8/32768/10    1.490 GiB/sec   1.445 GiB/sec   -3.007%
     SumKernelFloat/32768/0    6.407 GiB/sec   8.436 GiB/sec   31.670%
     SumKernelInt32/32768/0    5.307 GiB/sec   9.227 GiB/sec   73.863%
     SumKernelDouble/32768/10  8.489 GiB/sec   8.436 GiB/sec   -0.616%
     SumKernelInt32/32768/50   5.451 GiB/sec   5.242 GiB/sec   -3.841%
     SumKernelDouble/32768/1   9.690 GiB/sec   9.568 GiB/sec   -1.255%
     SumKernelInt8/32768/0     1.782 GiB/sec   3.726 GiB/sec   109.147%
   - SumKernelFloat/32768/1    4.899 GiB/sec   4.583 GiB/sec   -6.447%
   - SumKernelInt8/32768/1     2.111 GiB/sec   1.917 GiB/sec   -9.222%
     SumKernelInt32/32768/1    7.389 GiB/sec   7.281 GiB/sec   -1.470%
   - SumKernelInt16/32768/1    4.350 GiB/sec   3.855 GiB/sec   -11.363%
     SumKernelDouble/32768/0   10.647 GiB/sec  14.048 GiB/sec  31.949%
     SumKernelInt8/32768/50    1.641 GiB/sec   1.580 GiB/sec   -3.726%
     SumKernelFloat/32768/50   2.518 GiB/sec   2.545 GiB/sec   1.052%
     SumKernelInt16/32768/0    2.953 GiB/sec   6.038 GiB/sec   104.455%
     SumKernelInt64/32768/50   8.469 GiB/sec   8.439 GiB/sec   -0.349%
     SumKernelInt64/32768/0    11.243 GiB/sec  15.118 GiB/sec  34.462%
     SumKernelInt16/32768/50   3.070 GiB/sec   2.960 GiB/sec   -3.604%
     SumKernelInt64/32768/10   9.242 GiB/sec   9.336 GiB/sec   1.020%
   - SumKernelInt16/32768/10   3.638 GiB/sec   3.390 GiB/sec   -6.834%
     SumKernelInt32/32768/10   6.253 GiB/sec   6.012 GiB/sec   -3.844%
     SumKernelFloat/32768/10   4.111 GiB/sec   3.935 GiB/sec   -4.279%
     SumKernelDouble/32768/50  7.446 GiB/sec   7.455 GiB/sec   0.133%
     ========================  ==============  ==============  ========
   ```


----------------------------------------------------------------
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.

For queries about this service, please contact Infrastructure at:
users@infra.apache.org