# NVIDIA GeForce RTX 4090, 128 SMs, sm_89, iters=20000 reps=5
mode       best TF/s   all runs
f16/f16        320.4 315.2/314.2/317.1/318.2/320.4
f16/f32        320.4 318.3/307.9/320.4/317.8/307.1
bf16/f32       320.2 320.2/317.6/317.6/306.6/320.2
s8/s32         640.3 640.1/640.1/640.3/619.1/629.9
dp4a           159.1 133.2/159.1/159.1/159.1/159.0
ffma            76.3 76.3/76.2/68.3/76.1/76.1
