Measuring the True Cost of Div (32‑bit vs 64‑bit with Rust and Inline Asm)
Modern CPUs are fast—but some instructions still hide surprising costs. One of the most misunderstood is DIV. Is 32‑bit division faster than 64‑bit? Does instruction width matter anymore on x86‑64?
To answer this properly, we need more than wall‑clock timers. We need cycle counters, instruction retirement statistics, serialization barriers, and tight control over CPU affinity.
