feature: Add fastlanes bit unpacking cuda kernels

[robert3005]

I'm happy to big bang the whole thing in one pr or we can merge this as an intermediary step that generates the cuda kernels

Signed-off-by: Robert Kruszewski github@robertk.io

[joseph-isaacs]

shall we have this in -cuda for now?

[robert3005]

We should not, we should move things from vortex-cuda, there's too many things there

[codspeed-hq]

Merging this PR will degrade performance by 82.84%

⚠️ 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.

⚡ 7 improved benchmarks
❌ 11 regressed benchmarks
✅ 1143 untouched benchmarks
⏩ 1323 skipped benchmarks¹

⚠️ Please fix the performance issues or acknowledge them on CodSpeed.

Performance Changes

	Mode	Benchmark	BASE	HEAD	Efficiency
❌	Simulation	canonical_into_non_nullable[(10000, 1, 0.0)]	25.7 µs	36.2 µs	-29.12%
❌	Simulation	canonical_into_non_nullable[(10000, 10, 0.01)]	222.6 µs	306.1 µs	-27.28%
❌	Simulation	canonical_into_non_nullable[(10000, 1, 0.01)]	32.2 µs	41.1 µs	-21.61%
❌	Simulation	canonical_into_non_nullable[(10000, 10, 0.0)]	195.5 µs	279 µs	-29.91%
❌	Simulation	canonical_into_non_nullable[(10000, 1, 0.1)]	48 µs	57 µs	-15.76%
❌	Simulation	canonical_into_non_nullable[(10000, 10, 0.1)]	382.1 µs	471.6 µs	-18.98%
⚡	Simulation	canonical_into_nullable[(10000, 100, 0.0)]	5 ms	4.4 ms	+14.03%
⚡	Simulation	into_canonical_non_nullable[(10000, 1, 0.01)]	46.3 µs	39.1 µs	+18.27%
❌	Simulation	into_canonical_non_nullable[(10000, 10, 0.01)]	229.2 µs	309.2 µs	-25.87%
❌	Simulation	into_canonical_non_nullable[(10000, 10, 0.0)]	201.6 µs	282.4 µs	-28.59%
❌	Simulation	into_canonical_non_nullable[(10000, 10, 0.1)]	385.1 µs	471.5 µs	-18.34%
⚡	Simulation	into_canonical_non_nullable[(10000, 1, 0.0)]	40.4 µs	33.1 µs	+22.14%
⚡	Simulation	into_canonical_non_nullable[(10000, 1, 0.1)]	62.8 µs	55.2 µs	+13.74%
⚡	Simulation	into_canonical_nullable[(10000, 10, 0.0)]	540.8 µs	458.2 µs	+18.02%
⚡	Simulation	into_canonical_nullable[(10000, 100, 0.0)]	5.1 ms	4.3 ms	+16.84%
❌	Simulation	into_canonical_nullable[(10000, 10, 0.1)]	632 µs	718.9 µs	-12.09%
⚡	Simulation	into_canonical_nullable[(10000, 100, 0.1)]	6.9 ms	6.1 ms	+13.74%
❌	WallTime	u8_FoR[10M]	6.9 µs	40.5 µs	-82.84%

---

_{Comparing rk/cudabitpacking (960ce72) with develop (68130ce)}

Footnotes

1. 1323 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. ↩

[joseph-isaacs]

Can we have a test running a kernel?