Question PC freezes and display turns off ?

[Blue129]

Ryzen 7 5800x with Strix LC240 AIO
RTX 3070 OC
Strix B550-F
2x16 GB 3600MHz C18 Team Group Night Hawks
1TB Aorus Gen4 M.2 SSD
Seasonic Focus GX 650

Hey

So, I've been struggling with this issue for a very long time. I'm not an expert, so I probably haven't tried everything. It happens only while gaming or during Heaven benchmarking. Other than that I used the PC for Phtoshop and Illustrator that never bothered it, so I assume it has to be GPU related.

As described in the title, my PC freezes starting out by losing the display (no signal) and what comes after is basically random:

• In the best case, it reboots automatically.
• Other times, it keeps running and does nothing. RGB animation is broken and LEDs flash like a christmas tree.
• Or the same occurs, but the animations go on as if everything was ok. Still no display or any response.

Everytime it does this, temps don't go higher than 85C at most, with around 60-70ish being the average for both CPU and GPU. If I leave it in this state, it goes on forever unless I hard reset it. Temps normalize to idle, but the fans run on high speeds as if they still had to cool down the components. I kept monitoring temps and usage in Armoury Crate and iCue during benchmarking. The former showed 98% (marked with red) load on the GPU everytime the freezes occurred. Tested both benchmarks and games in medium to high settings. Setting it lower seemed to extend the operating time. I can't tell for sure tho, since most of the time it seems very random. Sometimes it freezes right after clicking the 'benchmark' button in Heaven, and sometimes it takes 5-30 minutes, or even hours.

PC's been to service before. They said the culprit was the PSU (an RM850x lol). Before that I had an EVGA SuperNova 850GQ with which I had the same issues (again: lol), until one day, it went out with a bang while turning it on. RMAd it and got the Corsair one. Still got the same freezes with that one, and then I took the PC to a technician. For him it worked fine during benchmarking for whole nights after swapping the PSU to his FSP one. He suggested going for a smaller PSU because power consumption never went above 500W. Thus the GX650.

Recently I tried checking the Event viewer and the results seem random to me.
1st: The computer has rebooted from a bugcheck. The bugcheck was: 0x00000116 (0xffffe78eec330460, 0xfffff8015845e0c0, 0xffffffffc000009a, 0x0000000000000004)
2nd: Event ID 14 from source nvlddmkm cannot be found - And this one came up 10+ times.
After that I got a warning: Display driver nvlddmkm stopped responding and has successfully recovered. - Guess I don't have to tell you, the display did not come back.

What I tried:

• Updated BIOS once
• Turn off C-states, but there's no such option in my BIOS
• Limited CPU voltage in BIOS (referring to a yt video)
• Reinstalled Windows (10)
• Reinstalled / Updated drivers
• Swapped PSU
• Command Prompt: sfc / scannow - no results

 

[Wayfall]

What happens when you do a Prime95 burn test?

Does the PSU 'click' off?

 

[Blue129]

What happens when you do a Prime95 burn test?

Does the PSU 'click' off?

For how long do you suggest running it?
I started the test and it's been running for nearly 20 minutes stably.

 

[Blue129]

For how long do you suggest running it?
I started the test and it's been running for nearly 20 minutes stably.

Hi, it's been a while.

I ran the test for hours and the PC did something very weird.
I sat in front of it for quite some time, and when I left for a bit, coming back I saw the PC went to sleep - in the middle of a stress test??? Event viewer refers to the reason as system idle. When I turned it back on the test continued automatically as if nothing had happened. It also didn't report any issues. It went to sleep during Self-test 24K (if that helps). After turning the PC on, it noted right after as 'passed'. There's no way that's part of the test, right?
Other than this, everything went stably without issues.