General Discussions

[Walluttt]

When i'm gaming in some heavy games, my pc restarts after some hours in. I don't really know why. At first I thought it was linked to OC/UV of my gpu and cpu, but after resetting to default setting it still does it. I did a sfc scan and nothing was detected as well as chkdsk. I don't think it is related to heat spikes because in average my jonc temp for my gpu is around 90 and my cpu 65. I don't know if it comes from my cpu, gpu, ram or psu. Here's my specs:

-Ryzen 5 5600X

-Be quiet! Pure rock 2

-CX 750M

-gigabyte aorus B550 elite V2

-RX 6800

-W10 Pro

-4x8 Corsair Vengeance 3200MHz

[johnnyenglish]

Hi,

I wouldn't discard GPU temps yet, 90ºC seems high for an average.

To put you in perspective, my RoG Strix RX6800 won't even touch 70ºC for a maximum temperature with 2300 core clock.

You can also try this: Do you have DOCP enabled? Disable it and troubleshoot again.

[Walluttt]

I dont think so, anyway i didn't touch any function called this way. Does it have another nam

[BigAl01]

90 C is way too high.  See what you can do to get more cooling air into the case around your GPU and make sure it's not being overclocked.  

Can I assume your PC doesn't restart during web browsing sessions, but only when gaming?

[Walluttt]

I thought it was normal for the junction temps. I didn't optimize the curves with the amd software, would it be a good start ? 

[Walluttt]

and indeed it doesn't crash while browsing or doing anything not gpu heavy

[BigAl01]

Do whatever you can to increase cooling to the video card.  Even undervolting the video card can help.

[Walluttt]

So pc restart can be caused by the gpu overheating ? I will try this then 

[BigAl01]

That's a good question.  I would normally expect the PC to just freeze as the video display stops.  I'm not so sure it would trigger a restart.  Windows can do some strange things when components stop responding though.  A good example is when the RAM is failing and you get a blue screen.

[Walluttt]

A faulty gpu wouldn't cause restart ? Maybe it comes from my psi, it's ten-year old but i don't have anything to test 

[BigAl01]

Your PSU is 10 years old?  So that is also suspect.  It's hard to know when a PSU is failing unless you can monitor the voltages on the rails just prior to the failure.  If it's really 10 years old, I would certainly consider that a candidate for replacement.  Go for a quality brand closer to 1,000 watts if you elect to replace it.  I never mess around with the PSU's I buy.  I just paid over $300 for a 1200 Watt Silverstone unit in the build I'm working on right now.

[johnnyenglish]

You are right, GPU tend to crash and leave a zombie PC. Unless its tied down to a bad PSU not providing enough power to the graphics card, then it will trip and restart.

@Walluttt 

Downclock and Undervolt.
2000, I know its low but its for troubleshooting.
Undervolt to something like 900-1000mV

Disable Zero RPM

[Walluttt]

thanks for the advices, I will try the downclocking and undervolting first to see if there is any improvement then if it isn't done i will buy a psu. Btw what do you mean by "GPU tend to crash and leave a zombie PC" does it prove my gpu is not faulty ?

[johnnyenglish]

Normally when a GPU crashes due to something, it will most of the times not trigger a restart.

And your main issue is a restarting PC. Although, we cannot ignore that GPU temps are high.

The PSU is not the best there is, its a low tier Bronze, you can check if its making coil whine when GPU is under load.


List of things to check:

- Are the pcie power cables well connected? Are they two separate cables from the PSU?
- Are the fans spinning correctly?
- Try disabling XMP/DOCP/AXMP run ram at default
- Downclock GPU
- Check for PSU Coil Whine

[Walluttt]

I can't really tell if i hear the sound coming from the gpu or the psu, but there's coil whine. I plugged it in daisy chain as I don't have any other pcie cable. I can try to disable xmp too

[BigAl01]

Then it's most likely a PSU problem of supplying enough current to your video card.  You should have separate PCIe cables from the PSU to the video card without daisy chaining.

[Walluttt]

As far as I remember there was just one pcie cable for my gpu. But anyway i looked back in my minidump files and found out my pc had some restarts monthes before i bought my new gpu, they were just appearing less often with the rtx 3060

[BigAl01]

I recommend you replace your PSU then.  Keep the old one as a spare for systems that don't require a higher end video card.  You can expect to spend over $100 USD for a quality unit.  I would not buy one below the 750 W level.

[goodplay]

Did you use DDU to remove any leftover nvidia driver/software files?

[Walluttt]

Yes I did use DDU, I did back a driver clean install with the amd software yesterday but it was still the same

[Walluttt]

just changed my psu. It still does it......

[Walluttt]

Ok I may have a clue. Can it be caused by faulty ram ?

[Walluttt]

I updated my bios when suddenly my screen turned black. I decided to remove 3 of my 4 ram sticks and it when on again. I saw on forums that ram could be responsible on making your pc unable to run after a bios update. It really does seems to me that ram is responsible am I wrong ?