Question Blue Screens & Prime95 fails. Now stable - - - what’s going on ?

[0bscryn]

Hi,
I don’t usually ask for help, but today I’m completely stumped.

I have an Intel system with an i9-13900K and 32GB of DDR5 RAM running at 7200MHz.
Out of nowhere, the machine started having issues. It began with one blue screen crash, and then it wouldn’t even boot into Windows because it kept blue-screening.

All the error codes pointed to a RAM issue, though it could also be a CPU problem—but I doubt it.
A motherboard issue seemed like an 8% chance at most. I kept all possibilities in mind, though. I checked the BIOS settings, and everything seemed fine, but I still reset them to default just to be safe.

Next, I decided to try a fresh Windows install. I booted from my USB drive, got past the language settings, hit "Next," and—bam—a forced reboot. Weird. It kept happening. I thought, "It can’t be my USB stick, right?"
But just to be thorough, I reinstalled Windows 11 on the USB drive. Same problem. Okay, this is getting annoying. I don’t have any spare DDR5 RAM lying around to test with, so that’s a bummer.

I double-checked that the BIOS was reset and even cleared the CMOS battery for good measure. The BIOS was indeed reset, everything looked good, and it was already on the latest version. I researched whether it could be a BIOS version issue, but that led nowhere. Frustrated, I booted from the USB again. After 25+ tries, it finally let me proceed. I reinstalled Windows 11 and actually made it to the desktop. I was thrilled!
To test things out, I ran Prime95 in Blend mode. Not even a minute in, three workers reported a "FATAL ERROR." That’s insane—I’ve never seen that happen so fast.

At this point, I was certain it was the RAM. Just to be sure, I ran Prime95 again, but this time only stressing the CPU. It ran fine for about 6 hours. For some reason, I decided to rerun Blend mode. A minute passed—no failures. Huh? I let it keep going. I had to step out, and when I came back 26 hours later, it was still running with no failures. What the heck? How is that possible? It makes no sense. I also ran MemTest, and it passed with flying colors.

So, I started fixing up the system—removed the bloatware, installed the latest drivers for my NVIDIA card and motherboard. Not a single blue screen since. I’m sitting here trying to figure this out, and this plot twist just doesn’t add up. It’s like I was chasing a problem that didn’t exist, and now I feel like an idiot chasing a ghost. I genuinely don’t understand what happened.

 

[Aeacus]

Windows works in mysterious ways. 🥷

Trial-and-error method eventually, may find the culprit. But it's like looking a small needle in a BIG haystack. And even then, when all resources are exhausted and still nothing, the remaining 1% is always reserved for the ghost in the machine.

I've seen my fare share of Windows anomalies, including hardware anomalies.
Since no explanation can be given as of why it happened, i've classified such events under "ghost in the machine" term. Aka unexplained phenomena.

Spoiler: Optional read: One of my very own ghosts in the machine (click here to read)

I've had unexplained phenomena with my build. And to this date, i, nor anyone else, have been able to get to the bottom of it.

When i 1st bought the parts of my Skylake build (full specs with pics in my sig), i had MSI GTX 1060 Gaming X 3G GPU and for few months, all of it was powered by Seasonic S12II-520 80+ Bronze (best group-regulated PSU ever made).
MSI had dedicated software for the GPU: MSI Gaming App, from where i was able to control the GPU LED and also fan profiles. One of the options was "cooling boost" (or similar named), where the idea was, that software puts the GPU fans at 100% for additional cooling for ~5 seconds, after which, fans wind down to previous RPMs.

But for me, GPU fans remained 100% after selecting that option from MSI Gaming App.
So, when i restarted the system, it caused the GPU fans to ramp up almost to the max and stay at ~80%. Build worked fine otherwise.
3rd restart and now, GPU fans ramped up and stayed at ~60%.
4th restart and GPU fans spin even slower at fixed speeds, ~40% or so. And so forth.
Took me 6 restarts until GPU fans stopped completely and GPU fans worked like it was supposed to (if GPU below 60C, fans RPM is 0).

And it wasn't isolated incident. Every time i did the "cooling boost" from MSI Gaming App, GPU fans were fixed at certain RPM, which dropped gradually with each restart of the PC. I was able to reliably reproduce the issue every single time.

Few months onwards, bought new PC case and PSU: Seasonic PRIME 650 80+Titanium [SSR-650TD]. After system swap and powered by new PSU, the "cooling boost" did work exactly as it was said on the tin - GPU fans ramped up to 100% and after 5 seconds, back to normal. No issues what-so-ever.

I can't even fathom how it was possible, but it was. I've classified it under "ghost in the machine"/"unexplained phenomena" label.

I have an Intel system with an i9-13900K and 32GB of DDR5 RAM running at 7200MHz.

Full system specs, including PSU make and model (or part number) is? Also, how old the PSU is, and was the PSU bought new or used/refurbished?