Improve take on Chunked(FSL) and FSL(Chunked) array alongside implementation of take using PiecewiseSequence - #8881
Open
robert3005 wants to merge 8 commits into
Open
Improve take on Chunked(FSL) and FSL(Chunked) array alongside implementation of take using PiecewiseSequence#8881robert3005 wants to merge 8 commits into
robert3005 wants to merge 8 commits into
CodSpeed HQ / CodSpeed Performance Analysis
succeeded
Aug 4, 2026 in 0s
Performance Gate Passed
⚡ 13 improved benchmarks
✅ 1829 untouched benchmarks
⏩ 44 skipped benchmarks1
Performance Changes
| Mode | Benchmark | BASE |
HEAD |
Efficiency | |
|---|---|---|---|---|---|
| ⚡ | Simulation | take_chunked_fsl_sorted[32, 64] |
1,293.4 µs | 461.9 µs | ×2.8 |
| ⚡ | Simulation | take_chunked_fsl_random[32, 64] |
1,303.5 µs | 477.9 µs | ×2.7 |
| ⚡ | Simulation | take_fsl_chunked_elements_sorted[32, 64] |
1,056.2 µs | 422.5 µs | ×2.5 |
| ⚡ | Simulation | take_fsl_chunked_elements_random[32, 64] |
1,031.3 µs | 435.7 µs | ×2.4 |
| ⚡ | Simulation | take_chunked_fsl_random[16, 64] |
899.4 µs | 439.2 µs | ×2 |
| ⚡ | Simulation | take_chunked_fsl_sorted[16, 64] |
888.5 µs | 436.4 µs | ×2 |
| ⚡ | Simulation | fsl_large |
209.2 µs | 105.4 µs | +98.43% |
| ⚡ | Simulation | take_fsl_chunked_elements_random[16, 64] |
698.5 µs | 398.2 µs | +75.41% |
| ⚡ | Simulation | take_chunked_fsl_random[8, 64] |
718.9 µs | 412.1 µs | +74.43% |
| ⚡ | Simulation | take_chunked_fsl_sorted[8, 64] |
697.7 µs | 412.3 µs | +69.22% |
| ⚡ | Simulation | take_fsl_chunked_elements_sorted[16, 64] |
648.7 µs | 393.3 µs | +64.94% |
| ⚡ | Simulation | take_fsl_chunked_elements_random[8, 64] |
480.3 µs | 374.4 µs | +28.28% |
| ⚡ | Simulation | take_fsl_chunked_elements_sorted[8, 64] |
456.3 µs | 378.7 µs | +20.5% |
Tip
Curious why this is faster? Comment @codspeedbot explain why this is faster on this PR, or directly use the CodSpeed MCP with your agent.
Comparing rk/chunkedfsltake (f5e3f57) with develop (f9d2abb)
Footnotes
-
44 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. ↩
Loading