Hacker News new | ask | show | jobs
by eth-mld 1126 days ago
https://github.com/NVIDIA/TransformerEngine - mixed precision FP8 is already here and provides similar accuracy to FP16/BF16
1 comments
villgax 1125 days ago
Yeah and for some reason limited to only hardware support of H100. Even the cost of doing it in software is outweighed by the speed & storage gains from it
link