Unable to determine the device handle for GPU 0000:19:00.0: Unknown Error

jesc3 · August 17, 2021, 12:29am

I have 4 2080Ti GPUs running on a machine with Ubuntu 18.04. Running nvidia-smi, I get

It seems that one of the GPUs is not being recognized. I checked all the hardware connections, and here is the log. Any idea what the issue is?
nvidia-bug-report.log (4.5 MB)

gcunhasergio · August 30, 2021, 1:30am

Have you tried rebooting your system? I was getting the following error and that worked for me:

Unable to determine the device handle for GPU 0000:65:00.0: Unknown Error

Also, it might be useful to share the output of nvidia-debugdump --list on your system.