Torch.cuda.is_available() returns false


I am working on a NVIDIA xavier AGX board, I did my initial setup and installed pytorch using this method.

when I run torch.cuda.is_available(), it returns False.

Can I know the reason for this and how to rectify this issue.

Xavier has:

cuDNN 8.0
CUDA 10.2


Could you use our prebuilt shared in the below command:


This topic was automatically closed 60 days after the last reply. New replies are no longer allowed.