# NVIDIA GeForce RTX 4080 SUPER, 80 SMs, sm_89, iters=20000 reps=3
mode       best TF/s   all runs
f16/f16        210.1 199.2/199.2/210.1
f16/f32        104.8 104.8/104.8/104.8
bf16/f32       104.8 104.8/104.7/104.7
s8/s32         418.9 418.8/418.9/418.9
dp4a           103.9 103.9/103.7/103.8
ffma            49.4 49.4/49.4/49.4
