# NVIDIA GeForce RTX 4090, 128 SMs, sm_89, iters=20000 reps=5
mode       best TF/s   all runs
f16/f16        320.5 305.1/300.9/306.8/320.5/317.9
f16/f32        320.4 320.4/320.4/317.9/320.4/306.7
bf16/f32       318.8 316.7/318.8/305.2/316.3/318.8
s8/s32         637.4 637.2/637.3/628.9/637.4/627.4
dp4a           158.5 158.5/134.8/158.5/158.5/158.5
ffma            76.3 72.7/76.3/76.3/76.3/57.0
