Reuse compression/decompression results in compression benchmarks - #9744
Reuse compression/decompression results in compression benchmarks#9744robert3005 wants to merge 3 commits into
Conversation
Merging this PR will improve performance by 25.56%
|
| Mode | Benchmark | BASE |
HEAD |
Efficiency | |
|---|---|---|---|---|---|
| ⚡ | WallTime | arrow_checked_add_u32_neon[16384] |
20.4 µs | 12.8 µs | +59.98% |
| ⚡ | WallTime | arrow_checked_add_u32_avx2[16384] |
21.3 µs | 17.6 µs | +20.68% |
| ⚡ | WallTime | words_gather_scalar_avx2[65536] |
9.3 µs | 8.2 µs | +13.46% |
| ⚡ | Simulation | allocate_drop_arrow[0] |
456.9 ns | 402.7 ns | +13.45% |
Tip
Curious why performance improved? Comment @codspeedbot explain why performance improved on this PR, or directly use the CodSpeed MCP with your agent.
Comparing rk/compressionbench (d889616) with develop (3170657)
Footnotes
-
206 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. ↩
Polar Signals Profiling ResultsLatest Run
Previous Runs (3)
Powered by Polar Signals Cloud |
Benchmarks: Compression 📖Commits: PR vortex / vortex-file-compressed / ns (0.940x ➖, 4↑ 0↓)
vortex / vortex-file-compressed / bytes (1.000x ➖, 0↑ 0↓)
vortex / vortex-file-compressed / ratio (0.970x ➖, 2↑ 1↓)
vortex / parquet / ns (0.983x ➖, 1↑ 0↓)
vortex / parquet / bytes (1.000x ➖, 0↑ 0↓)
vortex / arrow-ipc / ns (0.937x ➖, 6↑ 0↓)
vortex / arrow-ipc / bytes (1.000x ➖, 0↑ 0↓)
|
Signed-off-by: Robert Kruszewski <github@robertk.io>
c7361f2 to
1fb4aad
Compare
Reusing one Vortex array across iterations let every run after the first find its statistics already cached, so the timed region shrank to just the encoding. Reset the input before each iteration so the measurement matches a fresh conversion. Signed-off-by: Robert Kruszewski <github@robertk.io> Co-Authored-By: Claude Fable 5.1 <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_019CX6MXw5FFYnqQNdV7K4ci
Avoid decompressing and reloading files multiple times in benchmark run