Until yesterday I had a small, homemade ethereum mining rig with 5 GPUs, 4 of which were "sticking out" of my computer (4 GPUs were connected via risers). Everything was going smoothly and I decided to fully devote my hardware to crypto mining. I bought a 6th GPU and intended to connect it to the M2 port via an adapter. There were two 750W PSUs connected to the rig.
I reassembled my rig on a homemade stand. I managed to turn on the computer with 1 GPU installed. However, when I connected all the other GPUs and turned the computer on, the CPU started burning giving a significant amount of smoke and a strong, sharp smell followed. Now, the computer won't start even with 1 GPU. The motherboard quickly blinks with the BIOS error light and the power goes down.
There is no visible damage on the CPU.
The differences from the setup from yesterday are as follows:
- I changed the thermal paste on most of the GPUs
- The machine is now installed on a homemade stand.
- I plugged in the M2 to PCIE adapter.
My question is: how can I diagnose what went wrong and prevent this from happening again?
