Re:AMD May Get Across the CUDA Moat
by incognition on Saturday October 07, 2023 @01:14AM
(#37799169)
Fp64??
[ Parent
| on HN ]
https://en.wikipedia.org/wiki/Double-precision_floating-poin...
NVIDIA fp32 (H100) has 2x more TFLOPS than AMD's fp32 (MI250) and AI doesn't need fp64 precision.
Re:AMD May Get Across the CUDA Moat
by incognition on Saturday October 07, 2023 @10:32PM
(#37807386)
Lol it was meant as I wouldn't be caught dead using fp64