Analyzing the performance of Tensorflow training on M1 Mac Mini and Nvidia V100

Re:Analyzing the performance of Tensorflow training on M1 Ma... by fxtentacle on Thursday January 14, 2021 @06:08AM (#25774961)

Can't use it because most AI frameworks won't run on AMD because they did not implement suitable back-ends (yet).

[ Parent | on HN ]

Re:Analyzing the performance of Tensorflow training on M1 Ma... by breuleux on Thursday January 14, 2021 @11:57AM (#25778766)

There's one for PyTorch, I tested it about a year ago. You have to compile it from scratch and IIRC it translates/compile CUDA to ROCm at runtime which causes noticeable pauses on the first run. There may be other tweaks you have to do too. Once set up it performs decently, though.

[ Parent | on HN ]