neilconway commented on PR #26377:
URL: https://github.com/apache/datafusion/pull/26377#issuecomment-6195395686

   ARM benchmark results on an M4 Max (omitted from the PR description for 
brevity):
   
   upper
   - Some non-ASCII rows:
     - Utf8View, 4096 rows: 475.3 µs → 143.4 µs (−69.8%)
     - Utf8View, 4096 rows, 10% nulls: 447.0 µs → 147.8 µs (−66.9%)
     - Utf8View, 8192 rows: 967.0 µs → 315.0 µs (−67.4%)
     - Utf8View, 8192 rows, 10% nulls: 902.2 µs → 312.8 µs (−65.3%)
   - Non-ASCII words:
     - Utf8: 386.7 µs → 330.7 µs (−14.5%)
     - LargeUtf8: 389.1 µs → 329.3 µs (−15.4%)
     - Utf8View: 383.5 µs → 332.4 µs (−13.3%)
   - Mostly ASCII:
     - Utf8, len 8: 100.6 µs → 20.1 µs (−80.1%)
     - Utf8View, len 8: 107.3 µs → 21.1 µs (−80.4%)
     - Utf8, len 32: 370.4 µs → 29.9 µs (−91.9%)
     - Utf8View, len 32: 371.3 µs → 34.2 µs (−90.8%)
     - Utf8, len 128: 1.41 ms → 31.4 µs (−97.8%)
     - Utf8View, len 128: 1.34 ms → 34.9 µs (−97.4%)
   - Late non-ASCII:
     - Utf8, len 8: 125.2 µs → 83.4 µs (−33.4%)
     - Utf8View, len 8: 151.0 µs → 78.8 µs (−47.8%)
     - Utf8, len 32: 402.4 µs → 85.3 µs (−78.8%)
     - Utf8View, len 32: 402.9 µs → 86.4 µs (−78.6%)
     - Utf8, len 128: 1.45 ms → 90.6 µs (−93.7%)
     - Utf8View, len 128: 1.37 ms → 92.3 µs (−93.3%)
   
   lower
   - Some non-ASCII rows:
     - Utf8View, 4096 rows: 178.6 µs → 130.6 µs (−26.9%)
     - Utf8View, 4096 rows, 10% nulls: 173.9 µs → 133.2 µs (−23.4%)
     - Utf8View, 8192 rows: 381.3 µs → 285.3 µs (−25.2%)
     - Utf8View, 8192 rows, 10% nulls: 367.6 µs → 286.9 µs (−21.9%)
   - Non-ASCII words:
     - Utf8: 289.4 µs → 295.5 µs (+2.1%)
     - LargeUtf8: 291.1 µs → 294.7 µs (+1.2%)
     - Utf8View: 295.6 µs → 300.4 µs (+1.6%)
   - Mostly ASCII:
     - Utf8, len 8: 52.3 µs → 17.6 µs (−66.3%)
     - Utf8View, len 8: 55.1 µs → 19.5 µs (−64.5%)
     - Utf8, len 32: 55.2 µs → 29.0 µs (−47.5%)
     - Utf8View, len 32: 61.9 µs → 32.7 µs (−47.1%)
     - Utf8, len 128: 69.0 µs → 30.5 µs (−55.8%)
     - Utf8View, len 128: 71.1 µs → 36.8 µs (−48.2%)
   - Late non-ASCII:
     - Utf8, len 8: 81.2 µs → 85.7 µs (+5.6%)
     - Utf8View, len 8: 78.0 µs → 80.2 µs (+2.8%)
     - Utf8, len 32: 80.7 µs → 87.5 µs (+8.5%)
     - Utf8View, len 32: 85.6 µs → 88.3 µs (+3.2%)
     - Utf8, len 128: 86.9 µs → 92.0 µs (+5.9%)
     - Utf8View, len 128: 89.8 µs → 92.3 µs (+2.7%)
   
   translate (ASCII scalar from/to, no deletions)
   - Utf8:
     - 1024 rows, len 8: 7.1 µs → 7.0 µs (−0.3%)
     - 1024 rows, len 32: 13.6 µs → 12.0 µs (−11.6%)
     - 1024 rows, len 128: 43.6 µs → 28.0 µs (−35.6%)
     - 1024 rows, len 1024: 287.2 µs → 195.3 µs (−32.0%)
     - 4096 rows, len 8: 26.8 µs → 25.5 µs (−4.7%)
     - 4096 rows, len 32: 53.0 µs → 43.7 µs (−17.6%)
     - 4096 rows, len 128: 172.5 µs → 110.5 µs (−36.0%)
     - 4096 rows, len 1024: 1.15 ms → 778.7 µs (−32.4%)
   - Utf8View:
     - 1024 rows, len 8: 5.9 µs → 5.7 µs (−3.6%)
     - 1024 rows, len 32: 11.1 µs → 10.6 µs (−4.9%)
     - 1024 rows, len 128: 28.7 µs → 28.5 µs (−0.7%)
     - 1024 rows, len 1024: 188.9 µs → 188.7 µs (−0.1%)
     - 4096 rows, len 8: 22.6 µs → 21.6 µs (−4.2%)
     - 4096 rows, len 32: 44.3 µs → 41.9 µs (−5.5%)
     - 4096 rows, len 128: 110.8 µs → 109.1 µs (−1.6%)
     - 4096 rows, len 1024: 756.3 µs → 753.4 µs (−0.4%)
   
   replace (single ASCII byte, Utf8)
   - 1024 rows, len 32: 23.6 µs → 7.7 µs (−67.6%)
   - 1024 rows, len 128: 87.1 µs → 10.4 µs (−88.0%)
   - 1024 rows, len 32, 20% nulls: 20.3 µs → 8.0 µs (−60.8%)
   - 1024 rows, len 128, 20% nulls: 70.9 µs → 9.6 µs (−86.5%)
   - 4096 rows, len 32: 93.3 µs → 28.8 µs (−69.1%)
   - 4096 rows, len 128: 348.3 µs → 38.4 µs (−89.0%)
   - 4096 rows, len 32, 20% nulls: 81.2 µs → 30.9 µs (−62.0%)
   - 4096 rows, len 128, 20% nulls: 284.3 µs → 36.3 µs (−87.2%)
   
   x86 (AMD EPYC Milan, Hetzner ccx33)
   
   upper
   - Some non-ASCII rows:
     - Utf8View, 4096 rows: 952.7 µs → 254.5 µs (−73.3%)
     - Utf8View, 4096 rows, 10% nulls: 881.7 µs → 270.6 µs (−69.3%)
     - Utf8View, 8192 rows: 1.94 ms → 539.7 µs (−72.2%)
     - Utf8View, 8192 rows, 10% nulls: 1.79 ms → 559.4 µs (−68.7%)
   - Non-ASCII words:
     - Utf8: 825.5 µs → 726.6 µs (−12.0%)
     - LargeUtf8: 824.7 µs → 726.3 µs (−11.9%)
     - Utf8View: 759.5 µs → 727.6 µs (−4.2%)
   - Mostly ASCII:
     - Utf8, len 8: 204.0 µs → 19.6 µs (−90.4%)
     - Utf8View, len 8: 214.6 µs → 36.5 µs (−83.0%)
     - Utf8, len 32: 789.6 µs → 23.2 µs (−97.1%)
     - Utf8View, len 32: 757.4 µs → 36.2 µs (−95.2%)
     - Utf8, len 128: 3.13 ms → 51.7 µs (−98.3%)
     - Utf8View, len 128: 2.89 ms → 47.5 µs (−98.4%)
   - Late non-ASCII:
     - Utf8, len 8: 268.2 µs → 154.1 µs (−42.5%)
     - Utf8View, len 8: 259.4 µs → 156.5 µs (−39.7%)
     - Utf8, len 32: 881.8 µs → 136.4 µs (−84.5%)
     - Utf8View, len 32: 798.5 µs → 138.1 µs (−82.7%)
     - Utf8, len 128: 3.20 ms → 164.5 µs (−94.9%)
     - Utf8View, len 128: 2.95 ms → 163.9 µs (−94.4%)
   
   lower
   - Some non-ASCII rows:
     - Utf8View, 4096 rows: 270.3 µs → 232.8 µs (−13.9%)
     - Utf8View, 4096 rows, 10% nulls: 271.0 µs → 235.8 µs (−13.0%)
     - Utf8View, 8192 rows: 568.8 µs → 495.8 µs (−12.8%)
     - Utf8View, 8192 rows, 10% nulls: 572.5 µs → 502.4 µs (−12.3%)
   - Non-ASCII words:
     - Utf8: 514.2 µs → 511.0 µs (−0.6%)
     - LargeUtf8: 507.8 µs → 516.4 µs (+1.7%)
     - Utf8View: 510.2 µs → 509.4 µs (−0.2%)
   - Mostly ASCII:
     - Utf8, len 8: 67.8 µs → 18.6 µs (−72.5%)
     - Utf8View, len 8: 73.9 µs → 36.5 µs (−50.7%)
     - Utf8, len 32: 60.5 µs → 22.9 µs (−62.2%)
   - Some non-ASCII rows:
     - Utf8View, 4096 rows: 952.7 µs → 254.5 µs (−73.3%)
     - Utf8View, 4096 rows, 10% nulls: 881.7 µs → 270.6 µs (−69.3%)
     - Utf8View, 8192 rows: 1.94 ms → 539.7 µs (−72.2%)
     - Utf8View, 8192 rows, 10% nulls: 1.79 ms → 559.4 µs (−68.7%)
   - Non-ASCII words:
     - Utf8: 825.5 µs → 726.6 µs (−12.0%)
     - LargeUtf8: 824.7 µs → 726.3 µs (−11.9%)
     - Utf8View: 759.5 µs → 727.6 µs (−4.2%)
   - Mostly ASCII:
     - Utf8, len 8: 204.0 µs → 19.6 µs (−90.4%)
     - Utf8View, len 8: 214.6 µs → 36.5 µs (−83.0%)
     - Utf8, len 32: 789.6 µs → 23.2 µs (−97.1%)
     - Utf8View, len 32: 757.4 µs → 36.2 µs (−95.2%)
     - Utf8, len 128: 3.13 ms → 51.7 µs (−98.3%)
     - Utf8View, len 128: 2.89 ms → 47.5 µs (−98.4%)
   - Late non-ASCII:
     - Utf8, len 8: 268.2 µs → 154.1 µs (−42.5%)
     - Utf8View, len 8: 259.4 µs → 156.5 µs (−39.7%)
     - Utf8, len 32: 881.8 µs → 136.4 µs (−84.5%)
     - Utf8View, len 32: 798.5 µs → 138.1 µs (−82.7%)
     - Utf8, len 128: 3.20 ms → 164.5 µs (−94.9%)
     - Utf8View, len 128: 2.95 ms → 163.9 µs (−94.4%)
   


-- 
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]


---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]

Reply via email to