3dmark restarting my pc?

[srt4s256]

So im having an issue that only started to happen yesterday. This is exactly what i did, i did a ddu to go back to driver 532.03 due to the new driver crashing rdr2 constantly(only rdr2). After installing 532.03 rdr2 stopped crashing and it was playable again every other game also is playable. Up until(last night) i decided to try to run 3dmark so i launched it and as its collecting system info, before even getting to the menu of 3dmark pc black screens and restarts.

Thought it was a fluke and was thinking maybe its because im on an older driver? So after it rebooted on its own i decided to try again and same thing. So i gave up there and went to sleep, the next morning i started up my pc and decided to try again and i was able to run 3dmark time spy twice and it was fine it did not reboot.

Fast forward hours later toward the end of the day i saw nvidia released a new driver today that fixed rdr2 crashes, so i downloaded that driver and rdr2 ran good no crashes. Then decided to try 3dmark and it started restarting my system again? Idk wth is going on?

Rtx xc3 hybrid 3080
Ryzen 5600x
Corsair 32gb 3600mhz
Rog strix b450f

[Huperspace]

black screens and restarts = Sudden power down event
https://learn.microsoft.com/en-us/troubleshoot/windows-client/performance/event-id-41-restart
high likely defect PSU.

[UL_Jarnis]

Or an issue with CPU not being fully stable when load rapidly spikes (like during the SystemInfo scan when max CPU turbo clock is being determined)

Granted, problems here usually cause bluescreens, not shutdowns. Of course you may have OS set so there is no bluescreen on failure, but instead just a reboot.

[srt4s256]

Forgot to mention my psu but its a corsair rm850e only a couple months old

[.karizma]

Getting the same problem and never had it happen before, I was running a few timespys around the 2nd no problem but yesterday when I ran it again it was fine the first run then it started rebooting on me. No idea whats causing it all the sudden now.

On a 6700xt/5600x

[srt4s256]

Originally posted by .karizma:

Getting the same problem and never had it happen before, I was running a few timespys around the 2nd no problem but yesterday when I ran it again it was fine the first run then it started rebooting on me. No idea whats causing it all the sudden now.

On a 6700xt/5600x

Glad im not the only one. Im sure my psu is more than enoughvfor this 3080 and 5600x

Is yours rebooting at the same area as mine? Mine would reboot loading before i even got to the menu of 3dmark. Its weird because when i shut my pc off last night i went to sleep and in the morning i proceeded to run 3dmark again and this time i was able to successfully load to the menu and run a test. Then after playing rdr2 for an hour i tried to run another test then pc rebooted at the same place as before. It seems as if i play a game for awhile and try to run 3dmark after it wants to reboot.

What i have done just now was decided to put my bios setting back to stock and reran 3dmark and it ran it fine for now. And now im going to play a game for an hour then try to run 3dmark again to see if it still does it. My cpu was overclocked to 4750 @ 1.29 volts. I even tried to up the cpu voltage to 1.32 thinking that was the problem but it still happened. So now i defaulted my bios

[UL_Jarnis]

If you get reboots while overclocked and everything works stock, the overclock is not fully stable. This is not at all unusual. 3DMark puts a very heavy load (on the CPU in SystemInfo scan and during CPU test and on the GPU during graphics tests) and "but it works in game X" does not mean the same OC is stable in heavy test like 3DMark.

[srt4s256]

Originally posted by UL_Jarnis:

If you get reboots while overclocked and everything works stock, the overclock is not fully stable. This is not at all unusual. 3DMark puts a very heavy load (on the CPU in SystemInfo scan and during CPU test and on the GPU during graphics tests) and "but it works in game X" does not mean the same OC is stable in heavy test like 3DMark.

Thats what i figured but the odd part is it wont reboot in cinebench and my temps are fine 75-76c during cinebench. I would think it would be the other way around no? Usually if my cpu overclock is unstable i would crash in cinebench. But this time im actually crashing just trying to launch 3dmark not actually during benchmark which was weird

[UL_Jarnis]

Load put on the CPU is subtly different and there are scenarios where one works and the other fails. And it can be either one, depending on your CPU. They do not call it silicon lottery for nothing and it is these days pretty hard to really confirm an OC to be stable. And even when you do, you may still end up with an oddball bluescreen out of the blue every few months even when its otherwise stable. All due to current gen CPUs running far closer to the absolute ragged edge out of the box, so it becoming harder to really get any meaningful OC out of them without running into issues.

As an anecdote, I'm still running ancient i9-9900K at home with custom loop. It was "fully stable" OCd to 5200Mhz and temps were quite low. Except it bluescreened maybe once a month, completely randomly. And not even under heavy load, just normal use.

Dropped it to 5100Mhz. Still passed every test I could throw at it, but now it no longer gave random very rare bluescreens... since then worked crash-free for several years.