# NVIDIA GeForce RTX 4090, 128 SMs, sm_89, iters=20000 reps=5
mode       best TF/s   all runs
f16/f16        332.8 288.6/310.6/319.8/332.8/324.3
f16/f32        162.7 161.8/162.7/159.9/161.3/161.6
bf16/f32       162.9 162.9/162.7/161.2/160.1/160.3
s8/s32         661.6 636.2/637.1/661.6/661.5/625.7
dp4a           164.6 144.7/164.6/164.6/151.4/164.5
ffma            79.3 79.3/79.3/72.0/79.3/74.7
