# NVIDIA GeForce RTX 4090, 128 SMs, sm_89, iters=20000 reps=5
mode       best TF/s   all runs
f16/f16        312.7 305.9/304.4/311.9/296.5/312.7
f16/f32        327.6 316.3/319.7/316.0/321.5/327.6
bf16/f32       323.0 312.3/304.8/319.8/308.8/323.0
s8/s32         645.2 634.0/629.1/634.2/632.6/645.2
dp4a           164.6 164.6/152.8/158.2/152.2/154.5
ffma            79.3 79.2/79.3/79.3/79.3/79.3
