# NVIDIA GeForce RTX 4070, 46 SMs, sm_89, iters=20000 reps=5
mode       best TF/s   all runs
f16/f16        132.2 119.1/132.2/132.2/132.2/132.2
f16/f32        132.2 132.2/132.2/132.2/132.2/132.2
bf16/f32       132.0 132.0/131.9/131.9/131.9/131.9
s8/s32         263.7 263.7/263.7/263.7/263.7/263.7
dp4a            65.7 65.7/65.7/65.7/65.7/65.7
ffma            31.7 31.7/31.7/31.7/31.7/31.7
