perf: new simd kernels for improved performance

[d-v-b]

No description provided.

[codspeed-hq]

Merging this PR will improve performance by 60.94%

⚡ 3 improved benchmarks
✅ 42 untouched benchmarks
🆕 6 new benchmarks
⏩ 21 skipped benchmarks¹

Performance Changes

	Benchmark	BASE	HEAD	Efficiency
⚡	f64_to_f32/nearest_even[1000000]	10.7 ms	7.4 ms	+45.05%
⚡	f64_to_f32/nearest_even[1024]	11.5 µs	7.7 µs	+49.08%
⚡	f64_to_f32/nearest_even[64000]	688.3 µs	427.7 µs	+60.94%
🆕	f64_to_f32/towards_zero[1024]	N/A	20.2 µs	N/A
🆕	f64_to_i32/no_oor[64000]	N/A	497.9 µs	N/A
🆕	f64_to_i32/no_oor[1024]	N/A	9.2 µs	N/A
🆕	f64_to_f32/towards_zero[1000000]	N/A	18.8 ms	N/A
🆕	f64_to_i32/no_oor[1000000]	N/A	7.8 ms	N/A
🆕	f64_to_f32/towards_zero[64000]	N/A	1.2 ms	N/A

---

_{Comparing d-v-b:perf/beat-numpy (276b0da) with main (48edf3a)}

Footnotes

1. 21 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. ↩