pitrou commented on PR #41014:
URL: https://github.com/apache/arrow/pull/41014#issuecomment-2042893122
The numbers here seem non-conclusive: some improvements (up to 10%), some
regressions (up to -10%) (on Ubuntu 22.04, gcc 12.3.0, AMD Zen 2):
```
--------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------
Non-regressions: (29)
--------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------
benchmark baseline contender change %
counters
BM_DecodeArrowBooleanPlain/DecodeArrowWithNull/num_values:65536/null_in_ten_thousand:10000
409.995 MiB/sec 454.790 MiB/sec 10.926 {'family_index': 1,
'per_family_instance_index': 19, 'run_name':
'BM_DecodeArrowBooleanPlain/DecodeArrowWithNull/num_values:65536/null_in_ten_thousand:10000',
'repetitions': 1, 'repetition_index': 0, 'threads': 1, 'iterations': 36599}
BM_DecodeArrowBooleanPlain/DecodeArrowWithNull/num_values:16384/null_in_ten_thousand:10000
386.709 MiB/sec 425.690 MiB/sec 10.080 {'family_index': 1,
'per_family_instance_index': 18, 'run_name':
'BM_DecodeArrowBooleanPlain/DecodeArrowWithNull/num_values:16384/null_in_ten_thousand:10000',
'repetitions': 1, 'repetition_index': 0, 'threads': 1, 'iterations': 132312}
BM_DecodeArrowBooleanPlain/DecodeArrowWithNull/num_values:4096/null_in_ten_thousand:10000
316.576 MiB/sec 337.519 MiB/sec 6.615 {'family_index': 1,
'per_family_instance_index': 17, 'run_name':
'BM_DecodeArrowBooleanPlain/DecodeArrowWithNull/num_values:4096/null_in_ten_thousand:10000',
'repetitions': 1, 'repetition_index': 0, 'threads': 1, 'iterations': 454428}
BM_DecodeArrowBooleanPlain/DecodeArrowWithNull/num_values:1024/null_in_ten_thousand:10000
184.019 MiB/sec 188.734 MiB/sec 2.562 {'family_index': 1,
'per_family_instance_index': 16, 'run_name':
'BM_DecodeArrowBooleanPlain/DecodeArrowWithNull/num_values:1024/null_in_ten_thousand:10000',
'repetitions': 1, 'repetition_index': 0, 'threads': 1, 'iterations': 1058022}
BM_DecodeArrowBooleanRle/DecodeArrowWithNull/num_values:16384/null_in_ten_thousand:1
23.019 MiB/sec 23.179 MiB/sec 0.699 {'family_index': 0,
'per_family_instance_index': 2, 'run_name':
'BM_DecodeArrowBooleanRle/DecodeArrowWithNull/num_values:16384/null_in_ten_thousand:1',
'repetitions': 1, 'repetition_index': 0, 'threads': 1, 'iterations': 8281}
BM_DecodeArrowBooleanRle/DecodeArrowWithNull/num_values:16384/null_in_ten_thousand:5000
19.326 MiB/sec 19.455 MiB/sec 0.669 {'family_index': 0,
'per_family_instance_index': 14, 'run_name':
'BM_DecodeArrowBooleanRle/DecodeArrowWithNull/num_values:16384/null_in_ten_thousand:5000',
'repetitions': 1, 'repetition_index': 0, 'threads': 1, 'iterations': 6841}
BM_DecodeArrowBooleanRle/DecodeArrowWithNull/num_values:65536/null_in_ten_thousand:1
21.078 MiB/sec 20.929 MiB/sec -0.708 {'family_index': 0,
'per_family_instance_index': 3, 'run_name':
'BM_DecodeArrowBooleanRle/DecodeArrowWithNull/num_values:65536/null_in_ten_thousand:1',
'repetitions': 1, 'repetition_index': 0, 'threads': 1, 'iterations': 1881}
BM_DecodeArrowBooleanRle/DecodeArrowWithNull/num_values:16384/null_in_ten_thousand:100
22.455 MiB/sec 22.230 MiB/sec -1.000 {'family_index': 0,
'per_family_instance_index': 6, 'run_name':
'BM_DecodeArrowBooleanRle/DecodeArrowWithNull/num_values:16384/null_in_ten_thousand:100',
'repetitions': 1, 'repetition_index': 0, 'threads': 1, 'iterations': 8040}
BM_DecodeArrowBooleanPlain/DecodeArrowWithNull/num_values:16384/null_in_ten_thousand:1
108.683 MiB/sec 106.542 MiB/sec -1.969 {'family_index': 1,
'per_family_instance_index': 2, 'run_name':
'BM_DecodeArrowBooleanPlain/DecodeArrowWithNull/num_values:16384/null_in_ten_thousand:1',
'repetitions': 1, 'repetition_index': 0, 'threads': 1, 'iterations': 38869}
BM_DecodeArrowBooleanPlain/DecodeArrowWithNull/num_values:65536/null_in_ten_thousand:1
93.772 MiB/sec 91.677 MiB/sec -2.233 {'family_index': 1,
'per_family_instance_index': 3, 'run_name':
'BM_DecodeArrowBooleanPlain/DecodeArrowWithNull/num_values:65536/null_in_ten_thousand:1',
'repetitions': 1, 'repetition_index': 0, 'threads': 1, 'iterations': 8401}
BM_DecodeArrowBooleanRle/DecodeArrowWithNull/num_values:16384/null_in_ten_thousand:1000
21.499 MiB/sec 21.012 MiB/sec -2.262 {'family_index': 0,
'per_family_instance_index': 10, 'run_name':
'BM_DecodeArrowBooleanRle/DecodeArrowWithNull/num_values:16384/null_in_ten_thousand:1000',
'repetitions': 1, 'repetition_index': 0, 'threads': 1, 'iterations': 7652}
BM_DecodeArrowBooleanRle/DecodeArrowWithNull/num_values:1024/null_in_ten_thousand:100
34.267 MiB/sec 33.465 MiB/sec -2.342 {'family_index': 0,
'per_family_instance_index': 4, 'run_name':
'BM_DecodeArrowBooleanRle/DecodeArrowWithNull/num_values:1024/null_in_ten_thousand:100',
'repetitions': 1, 'repetition_index': 0, 'threads': 1, 'iterations': 196090}
BM_DecodeArrowBooleanRle/DecodeArrowWithNull/num_values:4096/null_in_ten_thousand:10000
1.247 GiB/sec 1.216 GiB/sec -2.492 {'family_index': 0,
'per_family_instance_index': 17, 'run_name':
'BM_DecodeArrowBooleanRle/DecodeArrowWithNull/num_values:4096/null_in_ten_thousand:10000',
'repetitions': 1, 'repetition_index': 0, 'threads': 1, 'iterations': 1827073}
BM_DecodeArrowBooleanRle/DecodeArrowWithNull/num_values:4096/null_in_ten_thousand:100
37.933 MiB/sec 36.914 MiB/sec -2.687 {'family_index': 0,
'per_family_instance_index': 5, 'run_name':
'BM_DecodeArrowBooleanRle/DecodeArrowWithNull/num_values:4096/null_in_ten_thousand:100',
'repetitions': 1, 'repetition_index': 0, 'threads': 1, 'iterations': 54274}
BM_DecodeArrowBooleanRle/DecodeArrowWithNull/num_values:1024/null_in_ten_thousand:10000
322.521 MiB/sec 313.735 MiB/sec -2.724 {'family_index': 0,
'per_family_instance_index': 16, 'run_name':
'BM_DecodeArrowBooleanRle/DecodeArrowWithNull/num_values:1024/null_in_ten_thousand:10000',
'repetitions': 1, 'repetition_index': 0, 'threads': 1, 'iterations': 1843736}
BM_DecodeArrowBooleanRle/DecodeArrowWithNull/num_values:1024/null_in_ten_thousand:1000
33.831 MiB/sec 32.709 MiB/sec -3.316 {'family_index': 0,
'per_family_instance_index': 8, 'run_name':
'BM_DecodeArrowBooleanRle/DecodeArrowWithNull/num_values:1024/null_in_ten_thousand:1000',
'repetitions': 1, 'repetition_index': 0, 'threads': 1, 'iterations': 193096}
BM_DecodeArrowBooleanPlain/DecodeArrowWithNull/num_values:4096/null_in_ten_thousand:5000
28.684 MiB/sec 27.693 MiB/sec -3.455 {'family_index': 1,
'per_family_instance_index': 13, 'run_name':
'BM_DecodeArrowBooleanPlain/DecodeArrowWithNull/num_values:4096/null_in_ten_thousand:5000',
'repetitions': 1, 'repetition_index': 0, 'threads': 1, 'iterations': 41078}
BM_DecodeArrowBooleanRle/DecodeArrowWithNull/num_values:1024/null_in_ten_thousand:5000
30.930 MiB/sec 29.855 MiB/sec -3.477 {'family_index': 0,
'per_family_instance_index': 12, 'run_name':
'BM_DecodeArrowBooleanRle/DecodeArrowWithNull/num_values:1024/null_in_ten_thousand:5000',
'repetitions': 1, 'repetition_index': 0, 'threads': 1, 'iterations': 177542}
BM_DecodeArrowBooleanRle/DecodeArrowWithNull/num_values:4096/null_in_ten_thousand:5000
33.578 MiB/sec 32.396 MiB/sec -3.522 {'family_index': 0,
'per_family_instance_index': 13, 'run_name':
'BM_DecodeArrowBooleanRle/DecodeArrowWithNull/num_values:4096/null_in_ten_thousand:5000',
'repetitions': 1, 'repetition_index': 0, 'threads': 1, 'iterations': 47770}
BM_DecodeArrowBooleanRle/DecodeArrowWithNull/num_values:65536/null_in_ten_thousand:1000
19.530 MiB/sec 18.825 MiB/sec -3.610 {'family_index': 0,
'per_family_instance_index': 11, 'run_name':
'BM_DecodeArrowBooleanRle/DecodeArrowWithNull/num_values:65536/null_in_ten_thousand:1000',
'repetitions': 1, 'repetition_index': 0, 'threads': 1, 'iterations': 1749}
BM_DecodeArrowBooleanRle/DecodeArrowWithNull/num_values:16384/null_in_ten_thousand:10000
4.551 GiB/sec 4.364 GiB/sec -4.107 {'family_index': 0,
'per_family_instance_index': 18, 'run_name':
'BM_DecodeArrowBooleanRle/DecodeArrowWithNull/num_values:16384/null_in_ten_thousand:10000',
'repetitions': 1, 'repetition_index': 0, 'threads': 1, 'iterations': 1653340}
BM_DecodeArrowBooleanRle/DecodeArrowWithNull/num_values:65536/null_in_ten_thousand:10000
12.488 GiB/sec 11.963 GiB/sec -4.202 {'family_index': 0,
'per_family_instance_index': 19, 'run_name':
'BM_DecodeArrowBooleanRle/DecodeArrowWithNull/num_values:65536/null_in_ten_thousand:10000',
'repetitions': 1, 'repetition_index': 0, 'threads': 1, 'iterations': 1150447}
BM_DecodeArrowBooleanPlain/DecodeArrowWithNull/num_values:4096/null_in_ten_thousand:1000
33.488 MiB/sec 32.029 MiB/sec -4.358 {'family_index': 1,
'per_family_instance_index': 9, 'run_name':
'BM_DecodeArrowBooleanPlain/DecodeArrowWithNull/num_values:4096/null_in_ten_thousand:1000',
'repetitions': 1, 'repetition_index': 0, 'threads': 1, 'iterations': 47876}
BM_DecodeArrowBooleanRle/DecodeArrowWithNull/num_values:65536/null_in_ten_thousand:5000
15.974 MiB/sec 15.275 MiB/sec -4.378 {'family_index': 0,
'per_family_instance_index': 15, 'run_name':
'BM_DecodeArrowBooleanRle/DecodeArrowWithNull/num_values:65536/null_in_ten_thousand:5000',
'repetitions': 1, 'repetition_index': 0, 'threads': 1, 'iterations': 1423}
BM_DecodeArrowBooleanPlain/DecodeArrowWithNull/num_values:1024/null_in_ten_thousand:1
320.887 MiB/sec 306.821 MiB/sec -4.384 {'family_index': 1,
'per_family_instance_index': 0, 'run_name':
'BM_DecodeArrowBooleanPlain/DecodeArrowWithNull/num_values:1024/null_in_ten_thousand:1',
'repetitions': 1, 'repetition_index': 0, 'threads': 1, 'iterations': 1835564}
BM_DecodeArrowBooleanRle/DecodeArrowWithNull/num_values:65536/null_in_ten_thousand:100
20.976 MiB/sec 20.043 MiB/sec -4.444 {'family_index': 0,
'per_family_instance_index': 7, 'run_name':
'BM_DecodeArrowBooleanRle/DecodeArrowWithNull/num_values:65536/null_in_ten_thousand:100',
'repetitions': 1, 'repetition_index': 0, 'threads': 1, 'iterations': 1864}
BM_DecodeArrowBooleanPlain/DecodeArrowWithNull/num_values:65536/null_in_ten_thousand:1000
16.754 MiB/sec 15.945 MiB/sec -4.832 {'family_index': 1,
'per_family_instance_index': 11, 'run_name':
'BM_DecodeArrowBooleanPlain/DecodeArrowWithNull/num_values:65536/null_in_ten_thousand:1000',
'repetitions': 1, 'repetition_index': 0, 'threads': 1, 'iterations': 1502}
BM_DecodeArrowBooleanRle/DecodeArrowWithNull/num_values:4096/null_in_ten_thousand:1000
36.834 MiB/sec 35.047 MiB/sec -4.850 {'family_index': 0,
'per_family_instance_index': 9, 'run_name':
'BM_DecodeArrowBooleanRle/DecodeArrowWithNull/num_values:4096/null_in_ten_thousand:1000',
'repetitions': 1, 'repetition_index': 0, 'threads': 1, 'iterations': 51976}
BM_DecodeArrowBooleanPlain/DecodeArrowWithNull/num_values:4096/null_in_ten_thousand:1
1.164 GiB/sec 1.106 GiB/sec -4.966 {'family_index': 1,
'per_family_instance_index': 1, 'run_name':
'BM_DecodeArrowBooleanPlain/DecodeArrowWithNull/num_values:4096/null_in_ten_thousand:1',
'repetitions': 1, 'repetition_index': 0, 'threads': 1, 'iterations': 1715926}
-----------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------
Regressions: (11)
-----------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------
benchmark baseline contender change %
counters
BM_DecodeArrowBooleanPlain/DecodeArrowWithNull/num_values:4096/null_in_ten_thousand:100
52.824 MiB/sec 50.082 MiB/sec -5.191 {'family_index': 1,
'per_family_instance_index': 5, 'run_name':
'BM_DecodeArrowBooleanPlain/DecodeArrowWithNull/num_values:4096/null_in_ten_thousand:100',
'repetitions': 1, 'repetition_index': 0, 'threads': 1, 'iterations': 74445}
BM_DecodeArrowBooleanPlain/DecodeArrowWithNull/num_values:1024/null_in_ten_thousand:5000
27.092 MiB/sec 25.616 MiB/sec -5.450 {'family_index': 1,
'per_family_instance_index': 12, 'run_name':
'BM_DecodeArrowBooleanPlain/DecodeArrowWithNull/num_values:1024/null_in_ten_thousand:5000',
'repetitions': 1, 'repetition_index': 0, 'threads': 1, 'iterations': 155156}
BM_DecodeArrowBooleanPlain/DecodeArrowWithNull/num_values:65536/null_in_ten_thousand:5000
14.632 MiB/sec 13.804 MiB/sec -5.665 {'family_index': 1,
'per_family_instance_index': 15, 'run_name':
'BM_DecodeArrowBooleanPlain/DecodeArrowWithNull/num_values:65536/null_in_ten_thousand:5000',
'repetitions': 1, 'repetition_index': 0, 'threads': 1, 'iterations': 1311}
BM_DecodeArrowBooleanPlain/DecodeArrowWithNull/num_values:65536/null_in_ten_thousand:100
29.288 MiB/sec 27.619 MiB/sec -5.699 {'family_index': 1,
'per_family_instance_index': 7, 'run_name':
'BM_DecodeArrowBooleanPlain/DecodeArrowWithNull/num_values:65536/null_in_ten_thousand:100',
'repetitions': 1, 'repetition_index': 0, 'threads': 1, 'iterations': 2617}
BM_DecodeArrowBooleanPlain/DecodeArrowWithNull/num_values:1024/null_in_ten_thousand:1000
31.006 MiB/sec 29.083 MiB/sec -6.200 {'family_index': 1,
'per_family_instance_index': 8, 'run_name':
'BM_DecodeArrowBooleanPlain/DecodeArrowWithNull/num_values:1024/null_in_ten_thousand:1000',
'repetitions': 1, 'repetition_index': 0, 'threads': 1, 'iterations': 177884}
BM_DecodeArrowBooleanPlain/DecodeArrowWithNull/num_values:16384/null_in_ten_thousand:5000
18.538 MiB/sec 17.293 MiB/sec -6.713 {'family_index': 1,
'per_family_instance_index': 14, 'run_name':
'BM_DecodeArrowBooleanPlain/DecodeArrowWithNull/num_values:16384/null_in_ten_thousand:5000',
'repetitions': 1, 'repetition_index': 0, 'threads': 1, 'iterations': 6521}
BM_DecodeArrowBooleanRle/DecodeArrowWithNull/num_values:1024/null_in_ten_thousand:1
125.710 MiB/sec 116.814 MiB/sec -7.076 {'family_index': 0,
'per_family_instance_index': 0, 'run_name':
'BM_DecodeArrowBooleanRle/DecodeArrowWithNull/num_values:1024/null_in_ten_thousand:1',
'repetitions': 1, 'repetition_index': 0, 'threads': 1, 'iterations': 716292}
BM_DecodeArrowBooleanPlain/DecodeArrowWithNull/num_values:1024/null_in_ten_thousand:100
41.014 MiB/sec 37.855 MiB/sec -7.700 {'family_index': 1,
'per_family_instance_index': 4, 'run_name':
'BM_DecodeArrowBooleanPlain/DecodeArrowWithNull/num_values:1024/null_in_ten_thousand:100',
'repetitions': 1, 'repetition_index': 0, 'threads': 1, 'iterations': 235482}
BM_DecodeArrowBooleanPlain/DecodeArrowWithNull/num_values:16384/null_in_ten_thousand:1000
19.242 MiB/sec 17.707 MiB/sec -7.979 {'family_index': 1,
'per_family_instance_index': 10, 'run_name':
'BM_DecodeArrowBooleanPlain/DecodeArrowWithNull/num_values:16384/null_in_ten_thousand:1000',
'repetitions': 1, 'repetition_index': 0, 'threads': 1, 'iterations': 6681}
BM_DecodeArrowBooleanRle/DecodeArrowWithNull/num_values:4096/null_in_ten_thousand:1
181.310 MiB/sec 165.584 MiB/sec -8.674 {'family_index': 0,
'per_family_instance_index': 1, 'run_name':
'BM_DecodeArrowBooleanRle/DecodeArrowWithNull/num_values:4096/null_in_ten_thousand:1',
'repetitions': 1, 'repetition_index': 0, 'threads': 1, 'iterations': 260095}
BM_DecodeArrowBooleanPlain/DecodeArrowWithNull/num_values:16384/null_in_ten_thousand:100
44.159 MiB/sec 39.911 MiB/sec -9.620 {'family_index': 1,
'per_family_instance_index': 6, 'run_name':
'BM_DecodeArrowBooleanPlain/DecodeArrowWithNull/num_values:16384/null_in_ten_thousand:100',
'repetitions': 1, 'repetition_index': 0, 'threads': 1, 'iterations': 13799}
```
--
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: [email protected]
For queries about this service, please contact Infrastructure at:
[email protected]
