When I run dmesg, I got a lot of those lines:
[133920.746977] nvidia-modeset: Freed GPU:5 (GPU-12f9f4a9-0b6a-feef-b2d6-93e0ce5428cd) @ PCI:0000:09:00.0
[133920.901801] nvidia-modeset: Freed GPU:4 (GPU-15387c93-2a5d-b45c-77b0-cb88e629425e) @ PCI:0000:08:00.0
[133921.032925] nvidia-modeset: Freed GPU:3 (GPU-c6762bb7-6a0e-486b-6878-ecfc6eb1f7ef) @ PCI:0000:07:00.0
[133921.142072] nvidia-modeset: Freed GPU:2 (GPU-f256d3d8-c898-c915-3106-76e808cc362e) @ PCI:0000:06:00.0
[133921.236166] nvidia-modeset: Freed GPU:1 (GPU-a3452d3f-ed35-5125-c8af-481622d9adf9) @ PCI:0000:03:00.0
[133922.043551] nvidia-modeset: Freed GPU:0 (GPU-df08cdf1-b3ff-f312-4a54-74f401f5e4be) @ PCI:0000:01:00.0
[133973.039094] nvidia-modeset: Allocated GPU:0 (GPU-df08cdf1-b3ff-f312-4a54-74f401f5e4be) @ PCI:0000:01:00.0
[133973.694116] nvidia-modeset: Allocated GPU:1 (GPU-a3452d3f-ed35-5125-c8af-481622d9adf9) @ PCI:0000:03:00.0
Any idea what causes this issue? The rig doesn't mine.
Is there a way how I can detect a faulty GPU/Riser, without removing the GPUs one by one? I thought maybe the GPU utilization log could do it, but I don't know if the GPU numbering there is the same like the numbering of slots of my motherboard?