Mwangi-Derrick/matrix-interop-demo

Polyglot matrix multiplication benchmark: Zig, Rust, C++ unified via C ABI. 32x speedup via loop reordering. Deep dive into cache locality, SIMD, and why manual optimization can regress performance.

语言：Zig · GitHub 星标：7 · Fork：1

话题：apple-silicon · avx2 · benchmarking · cpp · ffi · high-performance-computing · low-latency · matrix-multiplication

在 GitHub 上查看原仓库