4‑byte padding slashes Go array clear time by 49% on Intel
On Intel x86‑64, aligning a Go slice to an 8‑byte boundary by inserting 4 bytes of padding boosts the throughput of the REP STOSQ clearing loop by roughly 49 %. The same trick yields a modest ~9 % gain on AMD and none on ARM, exposing a CPU‑specific micro‑architectural quirk.