A ~200 LOC Rust crate benchmarked against ten other languages — came out ahead of a monomorphized C++ lib by ~4× at sub-fanout dispatch. Hot loop is six instructions. Can you make it faster?
A ~200 LOC Rust crate benchmarked against ten other languages — came out ahead of a monomorphized C++ lib by ~4× at sub-fanout dispatch. Hot loop is six instructions. Can you make it faster?