A few months ago, I had a bit of an issue with my PC, at random, seemingly under intensive GPU load, I would randomly have the application I am using crash, followed shortly be my entire Desktop Environment and OS. This is rather strange, seeing as this is less stable than even Windows tends to be. At a certain point the issue got so bad, I couldn't play VR games anymore, and even lost some footage to the... bug? I though it was a bug. It was not.
After a particularly unproductive day, I decided to investigate the bug further. I noticed that after the application crashes,
but before the DE crashes, I could not open my terminal. Very odd. I also could not open my file browser. VERY odd.
It is at this point I think to check the kernel logs in dmesg. The verdict? Nothing. The logs cut off just before the crash.
This is highly unusual, it's like someone yanked my SSD out mid-session!
Frustrated now, I decided to route the kernel logs to my phone, over SSH, to termux. The result?
Oh dear.
Turns out, my motherboard (or CPU, still not sure which) was dead! Something was wrong with the PCIe bus. This means that my GPU, and my NVMe SSD would both fail simultaneously, leading to the entire system locking up. The fix? Buy new hardware, unfortunately. So, what does one do with a several years old, failed Intel system? Upgrade to AM5 of course! So here I am now on the cheapest AM5 system money can buy 3 months later, and it runs like butter. Very smooth, very happy. I also now have an iGPU, so recording gameplay is even smoother than before.