Running on CUDA 10.X – Python alphafold


It looks like our server’s GPU nodes only support up to CUDA 10.2. With the downgraded versions of tensorflow and other modules/packages, will be output consistent with those produced from the default set-up?


Asked Jul 25 ’21 at 22:21

avatar of skyungyong


1 Answer:


Looking at

it appears that JAX 1.69 supports CUDA 10.1 and 10.2 – you could try those. We don’t use TensorFlow much except for the data pipeline, so I’d expect that the outputs would be fine. We haven’t fully validated this setup, though.

Answered Jul 28 ’21 at 17:42

avatar of tfgg


Read more here: Source link