Question Constant BSODs and stress test crashes. At my wit's end

[mrsoup1]

I have been having constant BSODS, seemingly out of nowhere, for the last few weeks and I can't seem to diagnose the problem. I have even bought new RAM, as I thought that my old DIMMS might have been the culprit, to no avail. My PC will also completely freeze during OCCT CPU and power stress tests, within 30 seconds. I am not overclocking ANY components (XMP is off too).

Things I've tried:
Resetting CMOS
Updating BIOS
Rolling back GPU drivers and installing newer ones
Buying new RAM
Switching DIMMs to different slots
MemTest (no errors)
Reinstalling windows (10 and 11) on fully formatted drives
Installing windows on a separate drive
System file checker (sfc /scannow)
CHKDSK
Enabling/disabling MCE
Disabling C-States
Installing chipset drivers
Running through onboard graphics

Setup:
ASUS Z690-E
Corsair Dominator DDR5
Intel i9 12900k
ASUS TUF RTX 3080 Ti
SuperFlower 1000w Leadex Platinum 80+
Custom loop cooling (GPU+CPU block, front mounted distro plate, 3x360mm rads)

Temps:
Idle - CPU = 25c, GPU = 25c
Load - CPU = 65c, GPU = 45c

Here are the two latest dump files, if anyone wants to take a crack at it:
https://files.catbox.moe/153at9.dmp
https://files.catbox.moe/4q3ysm.dmp

https://files.catbox.moe/ib7fje.dmp

 

[mrsoup1]

Are you running your rad fans in pull or push?

Since when in push, this will happen (at 3:45 in video):

View: https://www.youtube.com/watch?v=UyC3lZ5WFMk#t=3m45s




Temps, yes, but how about voltages? 🤔
Also, care to share screenshot of HWinfo64 showing all the voltage values? Namely after testing. (E.g launch HWinfo64, start the test, when test ends, screenshot what HWinfo64 voltage sensors show.)

No, my rads are spotless as they are quite literally brand new. I converted my entire build into a custom loop a few weeks ago.

It won't be possible to show my voltage values after the test because it always crashes before any test can complete (usually within a minute).

 

[Aeacus]

It won't be possible to show my voltage values after the test because it always crashes before any test can complete (usually within a minute).

How about HWinfo64 Logging feature? That records and saves the telemetry. Doesn't it save the telemetry until the point of crash? 🤔

 

[mrsoup1]

How about HWinfo64 Logging feature? That records and saves the telemetry. Doesn't it save the telemetry until the point of crash? 🤔

Received a brand new Seasonic Prime TX-1000 today and tested with it; still facing the same problems. All stress tests inevitably cause a shutdown or BSOD. I honestly have no idea what to do anymore. I'm guessing I now have to buy a new motherboard or CPU and isolate those.

 

[mrsoup1]

Which component is it more likely to be?

 

[Aeacus]

Which component is it more likely to be?

Based on similar issue i saw recently, i'd look towards CPU.

My 1st reply in your topic:

It could be a bad CPU. 🤔

Few days ago, we got a person here with similar symptoms and after extensive troubleshoot, it did come down to faulty CPU,
topic: https://forums.tomshardware.com/threads/system-bsods-and-crashes.3787658/

He had i9-13900KF, so a different CPU, still, symptoms are similar. And you don't have sub-par hardware either.

What you could to, is trying some of the troubleshooting steps we did in that topic. E.g using Unigine Superposition to see how GPU fares. Or trying 3D Mark and if it too crashes, like it did for him, it may be time to look towards CPU replacement.