pitrou commented on pull request #11101:
URL: https://github.com/apache/arrow/pull/11101#issuecomment-914287073
Updated numbers (locally) with the added heuristic:
* 4-byte bloom filter:
```
--------------------------------------------------------------------------------------------------------------------------------------------------------------------------------
Non-regressions: (2)
--------------------------------------------------------------------------------------------------------------------------------------------------------------------------------
benchmark baseline contender change %
counters
ParseCSVVehiclesExample 676.694 MiB/sec 1.463 GiB/sec 121.368 {'run_name':
'ParseCSVVehiclesExample', 'repetitions': 0, 'repetition_index': 0, 'threads':
1, 'iterations': 13}
ParseCSVStocksExample 949.070 MiB/sec 1.217 GiB/sec 31.354
{'run_name': 'ParseCSVStocksExample', 'repetitions': 0, 'repetition_index': 0,
'threads': 1, 'iterations': 103}
---------------------------------------------------------------------------------------------------------------------------------------------------------------------------------
Regressions: (3)
---------------------------------------------------------------------------------------------------------------------------------------------------------------------------------
benchmark baseline contender change %
counters
ParseCSVQuotedBlock 787.366 MiB/sec 682.948 MiB/sec -13.262
{'run_name': 'ParseCSVQuotedBlock', 'repetitions': 0, 'repetition_index': 1,
'threads': 1, 'iterations': 1072}
ParseCSVEscapedBlock 806.420 MiB/sec 639.499 MiB/sec -20.699
{'run_name': 'ParseCSVEscapedBlock', 'repetitions': 0, 'repetition_index': 0,
'threads': 1, 'iterations': 1234}
ParseCSVFlightsExample 639.818 MiB/sec 468.004 MiB/sec -26.854
{'run_name': 'ParseCSVFlightsExample', 'repetitions': 0, 'repetition_index': 2,
'threads': 1, 'iterations': 156}
```
* 8-byte SSE4.2 filter:
```
--------------------------------------------------------------------------------------------------------------------------------------------------------------------------------
Non-regressions: (2)
--------------------------------------------------------------------------------------------------------------------------------------------------------------------------------
benchmark baseline contender change %
counters
ParseCSVVehiclesExample 655.679 MiB/sec 2.795 GiB/sec 336.446 {'run_name':
'ParseCSVVehiclesExample', 'repetitions': 0, 'repetition_index': 2, 'threads':
1, 'iterations': 10}
ParseCSVStocksExample 925.001 MiB/sec 1.627 GiB/sec 80.154
{'run_name': 'ParseCSVStocksExample', 'repetitions': 0, 'repetition_index': 0,
'threads': 1, 'iterations': 103}
---------------------------------------------------------------------------------------------------------------------------------------------------------------------------------
Regressions: (3)
---------------------------------------------------------------------------------------------------------------------------------------------------------------------------------
benchmark baseline contender change %
counters
ParseCSVFlightsExample 625.680 MiB/sec 562.829 MiB/sec -10.045
{'run_name': 'ParseCSVFlightsExample', 'repetitions': 0, 'repetition_index': 1,
'threads': 1, 'iterations': 150}
ParseCSVQuotedBlock 824.559 MiB/sec 708.487 MiB/sec -14.077
{'run_name': 'ParseCSVQuotedBlock', 'repetitions': 0, 'repetition_index': 2,
'threads': 1, 'iterations': 1156}
ParseCSVEscapedBlock 763.441 MiB/sec 626.461 MiB/sec -17.942
{'run_name': 'ParseCSVEscapedBlock', 'repetitions': 0, 'repetition_index': 0,
'threads': 1, 'iterations': 1176}
```
--
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]
