Hard Crashes

leo1983

leo1983

leo1983

Associate
Joined
18 Aug 2017
Posts
41
Hoping I can get a bit of help on diagnosing the cause of these hard crashes I seem to be getting...

Built first PC in February, played Metro Exodus, PC froze up for a second or two then rebooted itself.
Thought maybe it was because I turned ray tracing on etc. so lowered settings and played a while longer. But then crashed again later in the game.

Since then I have had crashes in multiple games, but sometimes it plays well for a few hours without issue.

I always get a 'Default Radeon WattMan settings has been restored due to unexpected system failure' message when it gets back into Windows, which sounds like a GPU issue, but I havent changed from the default settings.

I was getting drivers from Asus for the GPU, but then downloaded Driver Booster and found there was a much newer Radeon Driver which I didn't have, I DL this (and anything else out of date) but still have issues.

I also changed GPU power connection from 2 PCIE connections on same cable to 2 separate cables.

Bit stumped really... any suggestions?


  • Windows 10 Pro
  • AMD Ryzen 7 5800X 8-Core Processor
  • ASUS TUF GAMING B550-PLUS DDR4 ATX Motherboard
  • Corsair RM750x 80+ Gold Fully Modular Power Supply
  • Corsair Vengeance LPX Black 32GB (2x16GB) 3600 MHz AMD Ryzen Tuned DDR4 Memory Dual Kit
  • ASUS Radeon RX 6800 16GB ROG STRIX OC Graphics Car
  • Cooler Master MasterAir MA620P RGB CPU Cooler
  • WD Blue SN550 2TB M.2 PCIe NVMe SSD/Solid State Drive
  • WD Gold 6TB hard Drive SATA 6Gbs
 
I've experienced something similar to this before but it happened only after over clocking my gpu. I suppose you could try to under clock your gpu and see if the issue persists. If it resolves, it could well be your gpu, and if it doesn't, could be something else.

You could also try using DDU to completely wipe and reinstall your drivers just in case something has gone wrong there?
 
Thanks, may try that DDU first then, hopefully its a driver thing. If the GPU is broken its going to be a nightmare trying to get a replacement.
 
You don't happen to have MSI Afterburner installed, do you? I noticed on my machine that even if my card wasn't oc'd, it would still happen. I uninstalled Afterburner and the issue went away
 
Interesting, I’ve started blue screen hard crashes randomly too. Sometimes straight from the desktop. It never used to do it, but did recently install afterburner when messing about with crypto mining and over locking the memory.

Time to uninstall /ddu and see if that works.
 
Ran DDU today, reinstalled drivers using 'AMD Adrenalin' which I guess is there software for keeping things up to date. Just had the first freeze a few hours later. Playing Escape From Tarkov this time, screen just freezes, i need to hold down power until it reboots.
Dont suppose anyone has any other suggestions i can try?
 
leo1983 said:
Ran DDU today, reinstalled drivers using 'AMD Adrenalin' which I guess is there software for keeping things up to date. Just had the first freeze a few hours later. Playing Escape From Tarkov this time, screen just freezes, i need to hold down power until it reboots.
Dont suppose anyone has any other suggestions i can try?
Click to expand...
Do you still crash with XMP disabled on the ram?
 
I'd try underclocking the GPU and seeing if it still crashes and then with the CPU By turning off core performance boost and see if this makes a difference.
 
Ok went into the Radeon Software, Performance, Tuning, GPU Tuning and lowered the Max Frequency % and Voltage % from 100% to 90%... Does that sound ok for underclocking the GPU?
 
leo1983 said:
Ok went into the Radeon Software, Performance, Tuning, GPU Tuning and lowered the Max Frequency % and Voltage % from 100% to 90%... Does that sound ok for underclocking the GPU?
Click to expand...
I'm not that familiar with AMD cards, can you lower the frequency but just leave the voltage at normal? also try the same with the GPU memory, try lower it by 500 but try these tests separately as it should help isolate if the card has an issue. If you don't have any crashes then the card could be the issue.

If you still experience crashes then put it back to stock settings then try turning off core performance boost on the CPU which should lock it at base frequency to check if that's unstable when boosting.

If both are still crashing then I'd run memtest on the ram overnight to check for errors.
 
Last edited:
As others have said drivers first point of call then memory.

Personally I'd go the old school route - start with bare minimum in pc - 1 stick of ram, no USB stuff etc. Even without gpu if possible (use on board gpu). Try the one stick of ram in each slot, repeat with the other. Then start plugging everything back in till you see the issue appear. If it appears from the start you know its issue with either psu, mobo or your os/drivers. Doubt you'd ever have 2 faulty ram sticks.

You shouldn't HAVE to underclock your gpu and the issue I find with doing that is underclocking it can in itself cause instability.
 
Thanks for tips so far.

Not too keen on unplugging everything and slowly adding back unless as a last resort. The crashes only happen when gaming, and some days they don't happen at all, so that could be a long process.

I did run memtest (using Windows Memory Diagnostic) which said no issues were found.

However, some other weird things have been happening...
When I tried opening Windows Memory Diagnostic yesterday... it opened briefly, then while I was reading through the options it closed itself down (this happened several times).
Today, I managed to run the test, and then took a look for the results in Event Viewer. But as of right now, I am trying to open event viewer again - as soon as it opens I get a blue loading spinner on the cursor, then event viewer closes itself. If I try and open again, it does the same thing.
I have noticed this type of thing before on this PC, where something opens then closes straight away on its own.

Maybe something went wrong with the windows install that is causing this and could also be causing the crashes?

Also, just to clarify the types of crashes I am getting, the regular ones I am keen to fix are more of a freeze... the game just freezes and I cant do anything apart from hold the power down until the system reboots.
Thought its worth mentioning that, as what I mentioned in the opening post where PC reboots itself is what happens for Metro Exodus, but that could be an issue with that game rather than this more regular issue I get with various games.
 
After the above, crashed 3 times in a row, culminating in no signal being sent to monitor.
I ended up calling a support guy over who said motherboard is showing an error on the GPU... he moved to the other GPU slot and got the system up and running again.
I did a fresh windows install, played a few games successfully, then I started getting crashes again.

I'm about to return the GPU as faulty... I havent got another one to try in it's place to confirm that, and it's going to be a nightmare sourcing another...
Any last chance saloon doubts it's the GPU before I return it? (and if not, any tips for sourcing another GPU :D)
 
I've had similar issues that turned out to be bad ram. Bad ram can sometimes pass multiple tests, but an overnight Memtest86 as a USB drive will usually detect the errors even if other 2 or passes don't..

Even though sounds a lot like the GPU, bad ram can also give all kinds of errors, maybe run Memetest86 overnight before just to be sure you RMA your GPU.
 
PC777 said:
I've had similar issues that turned out to be bad ram. Bad ram can sometimes pass multiple tests, but an overnight Memtest86 as a USB drive will usually detect the errors even if other 2 or passes don't..

Even though sounds a lot like the GPU, bad ram can also give all kinds of errors, maybe run Memetest86 overnight before just to be sure you RMA your GPU.
Click to expand...

Thanks - I'll give this a shot tonight.
 
Chives said:
Sorry to hijack but I've been having similar issues and my event viewer is full of kernel power errors, 41. There doesn't seem to be any further detail though, are these significant / suggest a particular fault?

Cheers :)
Click to expand...
Is it a 5800X as still in two minds as to RMA the one I have? Disabling the cool and quiet thing it only happens about once every week now.
 
You must log in or register to reply here.
Back
Top Bottom