I’ll briefly summarize and then go into more detail on what ive tried. Build
TLDR; I’ve separated most of the parts into a separate build and did various tests for all them to be running as they should. Due to this, I believe that the issue may be caused by whatever he is connecting to the PC. Just wanted another perspective on the matter so I know that I covered all bases.
In depth…
Issue first started on Windows 10 and when his games began to crash I immediately assumed it to be the gpu. In response we did DDU, reinstalled drivers and games continued to crash. Ended up resetting Windows 10 entirely but again games continued to crash.
It’s not on the build, but he has a WD Blue 1tb HDD and later got a WD 10tb HDD because of his slow internet, aiming to access larger games that he might not play frequently. I’m aware that you need to cover the first three pins on these shuckable drives and it did work with no issues after that. Though, for the sake of troubleshooting I did remove both of these drives out of the build and again same issue.
When I removed both sata drives I was under the impression it was fixed. I ran heavenly benchmark for an hour, played a few games, and all seemed well. He even had no issues for about a week but then the same issues started to happen again.
GPU
I assumed the GPU was faulty so I ended up testing the card on a separate build I know has zero issues. The GPU ran great. I did the same thing, ran heavenly benchmark, 3dmark, played several games that would usually crash within minutes of him launching but no issues whatsoever.
I reinstalled the GPU into his pc and did the same tests as well as using OCCT. Again of course no issues. The gpu ran with temps on average running around 70c in every game I tried.
CPU
I don’t have any AM5 mb’s to isolate the CPU but ended up stress testing with cinebench and OCCT. The CPU did throttle at 95c like it should and when in games it ended up remaining around 60c average. There were no issues that I could see.
AIO
With how the CPU temps were while running games for several hours I presume the AIO isn’t the issue. Would increase/decrease fans RPM as needed.
RAM
Ran memtest and OCCT but again no errors were presented in either.
PSU
I used again OCCT and didn’t see either the cpu/gpu throttling while doing such. Isolated PSU into another system and ran 3dmark, heavenly benchmark on another build but got same results.
Everything tried…
- DDU
- Reset Windows 10
- Removed all SATA drives
- Isolated & Stress test GPU
- Stress test CPU
- Isolated & Stress test PSU
- Test RAM
With all that in mind I can only conclude that it has to be something USB. From what I know he often plays with a Logitech G920 Wheel and he told me he never unplugs it. I noticed alot of people getting BSOD with this wheel so told him to keep it unplugged when not using it and did what logitech states from there FAQ in regards to BSOD.
Only other peripherals I know of is a HyperX Quadcast, Razer Peripherals, and other logitech peripherals.
I also went through event viewer and from what I can remember I had him uninstall razers manager due to it crashing everyday, fixed DCOM not communicating, and uninstalled microsofts old controller manager that was crashing as well.
Again I dont know where to look or what else to try. Any help would be greatly appreciated and thanks for anytime you spend looking over this.
Also, I’ll get dump files when I can and update here when I have them.
Edit: Finally have the files
Ok, so looking at the BSOD minidumps, the BSOD from 2/13/24 and 1/10/2024 give a bug check event that is typically driver related. My recommendation would be to download prime95 and use the “Blend” test for a couple of hours to try and recreate the BSOD. If you don’t get anything you can try running 3DMark to force it. It’s easier when you can “recreate” the problem on purpose instead of waiting for it to happen. Once you reproduce the BSOD you can try and go through the Event Log and determine which driver was the cause since there will be an Application Error usually present.
If you can’t find anything in the logs, you can always pull out all peripherals and use prime95 for another couple of hours making sure that the issue is gone. If you still get a BSOD, you’ve got other hardware causing the issue. (or you could have a combination of both)
If you can’t reproduce the BSOD after unplugging the peripherals, plug one back in at a time stress testing to determine which peripheral is causing it. But you should be able to find it in the event log. I hope this helps.