I’m using 2 Tesla M60，linux18.04+cuda10.1 and my driver version is 418.56,
when i type nvidia-smi it’s works fine,
show me a list of status.But when i try to use both of them to trin a model.
it crashed and reported me the error
Unable to determine the device handle for GPU 0000:B1:00.0: Unknown Error
i
'm quiet new to these thing ,somebody help
Please run nvidia-bug-report.sh as root and attach the resulting .gz file to your post. Hovering the mouse over an existing post of yours will reveal a paperclip icon.
[url]https://devtalk.nvidia.com/default/topic/1043347/announcements/attaching-files-to-forum-topics-posts/[/url]
Hi,sorry i didn’t reply.
But problem was solved.
It’s caused by over-heating,I added two mini fans for each M60.
Not like before,the temperature went over 90°,it didn’t go more than 65°.