Question Blue screens and Prime95 stop working. Now it's stable - but what's happening?
Question Blue screens and Prime95 stop working. Now it's stable - but what's happening?
Hi,
I usually don’t seek assistance, but today I’m really puzzled.
My setup is an Intel system with an i9-13900K and 32GB of DDR5 RAM operating at 7200MHz. Suddenly, the machine began experiencing problems. It started with a blue screen crash, and then it failed to boot into Windows because it kept displaying a blue-screen error.
All the error messages suggested a RAM fault, although a CPU issue seemed unlikely. A motherboard problem was considered possible at most 8% of the time, but I remained open-minded. I verified the BIOS settings; everything appeared normal, yet I reset them to defaults just to be safe.
Then I opted for a fresh Windows installation. I booted from my USB drive, bypassed language options, selected "Next," and—suddenly—a forced reboot occurred. Odd. It repeated. I wondered, “Could it really be the USB stick?”
To confirm, I reinstalled Windows 11 on the same USB drive. The issue persisted. This was becoming frustrating. I didn’t have any spare DDR5 RAM to test, which made things worse.
I double-checked that the BIOS had been reset and even cleared the CMOS battery for extra caution. The BIOS was indeed reset, everything seemed correct, and it was running the latest version. I looked into possible BIOS version problems, but found nothing. Still, I booted from USB again. After more than 25 attempts, it finally allowed me to continue. I reinstalled Windows 11 and successfully reached the desktop. I was relieved!
To test further, I ran Prime95 in Blend mode. Within seconds, three workers reported a “FATAL ERROR.” That’s unreal—I’ve never seen such a rapid failure.
By this point, I was convinced it was the RAM. To verify, I reran Prime95, this time stressing only the CPU. It performed well for about six hours. For some reason, I decided to try Blend mode again. A minute passed—no errors. Huh? I let it run a bit longer. I returned after 26 hours and everything was still functioning without issues. What’s going on? This doesn’t make sense.
I also ran MemTest, which completed successfully.
So, I began making repairs—removed unnecessary software, installed the latest drivers for my NVIDIA card and motherboard. No more blue screens. I’m stuck trying to understand what happened, and this situation feels like a complete misunderstanding.
Windows operates in puzzling ways. 🥷
Through trial and error, we might eventually identify the cause. Yet it's akin to searching for a tiny thread inside an enormous haystack. Even then, once all resources are spent and nothing remains, the remaining 1% is always attributed to the ghost in the machine.
I've experienced my share of Windows oddities, including hardware quirks.
Since no clear explanation exists for why it occurred, I've labeled such occurrences under the "ghost in the machine" category—also known as unexplained events.
Spoiler:
Optional read: One of my own mysterious occurrences in the machine (click here to read)
I've encountered unexplained issues with my system. To date, neither I nor anyone else has managed to pinpoint the reason.
When I first acquired the components for my Skylake build (complete specs and photos attached), I had an MSI GTX 1060 Gaming X 3G GPU, powered by Seasonic S12II-520 80+ Bronze—considered one of the best group-regulated PSUs ever produced.
MSI provided dedicated software for the GPU: MSI Gaming App, which allowed me to control the GPU LED and fan settings. Among the options, there was a "cooling boost" feature (or similar), suggesting the software would increase fan speed briefly (~5 seconds) for extra cooling before returning to normal.
However, for me, the fans stayed at 100% after selecting that option.
Restarting the system caused the fans to surge almost to maximum and then settle around 80%. The build functioned normally otherwise.
A third restart led to fans increasing and stabilizing near 60%. The fourth restart resulted in slower fan speeds, hovering around 40%.
It took me six restarts until the fans ceased completely and operated as intended (fans stopped below 60°C, RPM dropping to zero).
This wasn't an isolated case. Each time I used the "cooling boost" feature, the fans consistently maintained a steady speed, which gradually decreased with each restart. I could reliably reproduce the problem every time.
Several months later, I upgraded to a new PC case and PSU: Seasonic PRIME 650 80+Titanium [SSR-650TD]. After swapping the system and using the new PSU, the "cooling boost" function worked as advertised—fans reached 100% and returned to normal after about five seconds. No problems occurred.
I still can't understand how this was possible, but it did happen. I've categorized it under "ghost in the machine"/"unexplained phenomena."
Complete system specifications, including PSU details (make, model, or part number), are? Also, what is the age of the PSU, and was it new or used/refurbished?