Skip to content

[Superseded] perf: vectorize small u8 table take - #9567

Closed
joseph-isaacs wants to merge 1 commit into
ji/small-u8-table-take-benchmarkfrom
ji/small-u8-table-take
Closed

[Superseded] perf: vectorize small u8 table take#9567
joseph-isaacs wants to merge 1 commit into
ji/small-u8-table-take-benchmarkfrom
ji/small-u8-table-take

perf: vectorize small u8 table take

593e136
Select commit
Loading
Failed to load commit list.
CodSpeed HQ / CodSpeed Performance Analysis failed Aug 24, 2026 in 0s

3 benchmarks regressed

⚠️ Unknown Walltime execution environment detected

Using the Walltime instrument on standard Hosted Runners will lead to inconsistent data.

For the most accurate results, we recommend using CodSpeed Macro Runners: bare-metal machines fine-tuned for performance measurement consistency.

⚠️ Different runtime environments detected

Some benchmarks with significant performance changes were compared across different runtime environments,
which may affect the accuracy of the results.

Open the report in CodSpeed to investigate

⚡ 9 improved benchmarks
❌ 3 regressed benchmarks
✅ 1969 untouched benchmarks
🆕 6 new benchmarks
⏩ 54 skipped benchmarks1

Warning

Please fix the performance issues or acknowledge them on CodSpeed.

Performance Changes

Mode Benchmark BASE HEAD Efficiency
❌ WallTime words_gather_scalar_avx2[65536] 8.3 µs 9.4 µs -12.35%
❌ Simulation compact[(1024, 90)] 1.6 µs 1.8 µs -10.8%
❌ Simulation compact[(2048, 90)] 1.7 µs 1.9 µs -10.4%
⚡ Simulation decode_primitives[u8, (2000, 4)] 51.4 µs 41.1 µs +25.06%
⚡ Simulation decode_primitives[u8, (2000, 2)] 51.5 µs 41.3 µs +24.82%
⚡ Simulation decode_primitives[u8, (2000, 8)] 51.9 µs 41.7 µs +24.39%
⚡ Simulation take[small_m/shuffled/primitive/nonnull/chunks=2048/indices=64] 801.9 µs 711.9 µs +12.64%
⚡ Simulation decode_primitives[u8, (1000, 8)] 43.5 µs 38.8 µs +12.16%
⚡ Simulation decode_primitives[u8, (1000, 4)] 43.7 µs 39 µs +12.12%
⚡ Simulation take[small_m/shuffled/primitive/nonnull/chunks=256/indices=64] 611.3 µs 545.4 µs +12.07%
⚡ Simulation decode_primitives[u8, (1000, 2)] 44.6 µs 39.8 µs +12.04%
⚡ Simulation take[small_m/shuffled/primitive/nonnull/chunks=2048/indices=16] 376.2 µs 338.5 µs +11.16%
🆕 WallTime dict_canonicalize_gt_u8_neon[1000000] N/A 50.7 µs N/A
🆕 WallTime dict_canonicalize_gt_u8_neon[16000000] N/A 779.5 µs N/A
🆕 WallTime dict_canonicalize_gt_u8_avx2[1000000] N/A 60 µs N/A
🆕 WallTime dict_canonicalize_gt_u8_avx2[16000000] N/A 1.4 ms N/A
🆕 WallTime dict_canonicalize_gt_u8_avx512[1000000] N/A 57.2 µs N/A
🆕 WallTime dict_canonicalize_gt_u8_avx512[16000000] N/A 1.3 ms N/A

Tip

Investigate this regression by commenting @codspeedbot fix this regression on this PR, or directly use the CodSpeed MCP with your agent.

Comparing ji/small-u8-table-take (593e136) with develop (ef43456)2

Open in CodSpeed

Footnotes

  1. 54 benchmarks were skipped, so the baseline results were used instead. If they were deleted from the codebase, click here and archive them to remove them from the performance reports. ↩

  2. No successful run was found on ji/small-u8-table-take-benchmark (d49393c) during the generation of this report, so develop (ef43456) was used instead as the comparison base. There might be some changes unrelated to this pull request in this report. ↩

View more details on CodSpeed HQ