Rtx3070, xid=79 GPU has fallen off the bus. Reseated / Change PCIE slot/ Change gpu card not work

Hello everyone. I have met “GPU has fallen of the bus” for a long time, and it occurs not very common, e.g., one week(or one month) a time, however, once I met this problem, it may continuously appeared for about 2 or 3 times in the same day and the the problem time interval is about half an hour.

My card is a new GeForce RTX 3070, and I have tried to:

  1. reseat the GPU, not working.
  2. change the PCIE slot, not working.
  3. buy another new 3070 card, and the problem also exists.

By the way, PCIE ASPM of my computer is always disabled. And I have checked the temperature, it won’t exceed 60°C. And the power supply is enough, it’s more than 1000w.

Also, my computer is set intel_idle.max_cstate=0 in grub.

The GPU average usage is about 30%, and I have tried to gpuburn, but it can’t reproduce the problem.

My computer type is Nuvo-8208GC and my nvidia driver is 470.63.01(This computer only support this driver for RTX 30 series card)

My os is ubuntu 20.04 with 5.13.0-41.

Very appreciate for any suggestions!

The attachment is the result of nvidia-bug-report.sh

gpu.logs4 (759.0 KB)

Please see this on how to update your bios so you can use newer drivers:
https://forums.developer.nvidia.com/t/nvidia-driver-problem-rtx3080-12gb-220a/221224/6
If the issue persists, please check/replace your psu.

Hi generix

Thank you very very very much for this advice, I will try it and report the new result later.