Rust Simd Intrinsics · mnwa bundle Design, implement, review, debug, benchmark, and port performance-critical Rust SIMD code across x86/x86_64 SSE (SSE1), SSE2, AVX and AVX2, and Arm NEON. Use for auto-vectorization, runtime feature dispatch, std::arch intrinsics, fearless_simd, wide, or nightly std::simd; and for byte scanning, compression matchers, image/audio/DSP kernels, reductions, masks, widening, saturation, and vectorized transforms. Do not trigger for ordinary non-hot code where SIMD has not been justified by profiling.
