[bench] sparse bitpacked filter and take

[robert3005]

Adds sparse bitpacked filter and take benchmarks for u8, u16, u32, and u64, providing a baseline for the extraction optimization in #9723.

[codspeed]

Merging this PR will degrade performance by 3.28%

⚠️ 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.

⚡ 1 improved benchmark
❌ 1 regressed benchmark
✅ 2200 untouched benchmarks
🆕 72 new benchmarks
⏩ 218 skipped benchmarks¹

Warning

Please fix the performance issues or acknowledge them on CodSpeed.

Performance Changes

	Mode	Benchmark	BASE	HEAD	Efficiency
❌	Simulation	decompress[u64, (4000, 1024)]	71.1 µs	86.4 µs	-17.76%
⚡	WallTime	scalar_subtract_neon	13.4 µs	11.8 µs	+13.74%
🆕	WallTime	filter_neon[32]	N/A	1.8 µs	N/A
🆕	WallTime	filter_neon[48]	N/A	2.1 µs	N/A
🆕	WallTime	filter_neon[8]	N/A	1.3 µs	N/A
🆕	WallTime	filter_neon[64]	N/A	3.2 µs	N/A
🆕	WallTime	filter_neon[8]	N/A	1.3 µs	N/A
🆕	WallTime	filter_neon[80]	N/A	3.4 µs	N/A
🆕	WallTime	filter_neon[160]	N/A	5.7 µs	N/A
🆕	WallTime	filter_neon[192]	N/A	6.1 µs	N/A
🆕	WallTime	filter_neon[8]	N/A	1.3 µs	N/A
🆕	WallTime	filter_neon[16]	N/A	1.3 µs	N/A
🆕	WallTime	filter_neon[24]	N/A	1.5 µs	N/A
🆕	WallTime	filter_neon[8]	N/A	1.2 µs	N/A
🆕	WallTime	threshold_neon[32]	N/A	2.9 µs	N/A
🆕	WallTime	threshold_neon[48]	N/A	3.2 µs	N/A
🆕	WallTime	threshold_neon[8]	N/A	2.2 µs	N/A
🆕	WallTime	threshold_neon[64]	N/A	4.7 µs	N/A
🆕	WallTime	threshold_neon[8]	N/A	3.2 µs	N/A
🆕	WallTime	threshold_neon[80]	N/A	5.1 µs	N/A
...	...	...	...	...	...

ℹ️ Only the first 20 benchmarks are displayed. Go to the app to view all benchmarks.

Tip

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

---

_{Comparing rk/fastlanes-sparse-benchmarks (f17daff) with develop (d5e2856)}

Footnotes

1. 218 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. ↩