Question PC freezes while playing games, could it be cause of low quality PSU or a GPU problem?

[riskal]

Hello, last year I finished my build with an used 1080 Ti AORUS XTREME instead of 1070 because I ended up with more money than I expected. I was little bit worried that my PSU (Corsair CV550) won't be enough but after reassurance from some people from this sub I decided to go for it. The day I've got my GPU I've experienced crashes in Overwatch, Cyberpunk 2077 and Unigine Heaven after around 1 minute of playing. I "solved" this problem by setting the GPU power limit to 80% while thinking it's the PSU problem. I started thinking it's the PSU because I've also tried FurMark which ran without any crashes or freezes although it got the GPU to 90-91°C where it started to thermal throttle. This made me think that if it doesn't freeze/crash with high GPU load but crashes/freezes in games with high load on both GPU and CPU it has to be PSU problem.

Fastforward to now, I've tried setting the GPU to 100% again and running Cyberpunk 2077. First few seconds were fine the temp got to 76°C and suddenly the whole PC froze and I had to restart it via power button. I also have some money that I could spend on high quality PSU but I'm not sure if that would fix the problem cause I could be tunnel visioned on the PSU and it could possibly be a GPU problem or something entirely different. Any thoughts about what could be the problem or something I could try to do to indentify it?

If you need any other info that I didn't provide I'll be more than happy to provide it asap.

PC build:
CPU: i7 4770
GPU: 1080 TI AORUS XTREME (I didn't overclocked it. Not sure what previous owner did with it.)
MOBO: Gigabyte GA-H97-D3H
RAM: HyperX 2x8GB 1866MHz
SSD: Samsung EVO 970 500GB
Cooler: Arctic Freezer 34
Case: SilentiumPC Regnum RG6V TG
PSU: Corsair CV550 80+ Bronze (1 PCIe power cable that has two ends 8-pin/6+2-pin)

PS: English isn't my native language so I apologize for any grammatical errors.

Now I've tried running Cyberpunk 2077 and Unigine Heaven. Both crashed but this time it only crashed into desktop and didn't freeze the whole PC, in the screeshot there's my MSI Afterburner settings and two LiveKernelEvents 141 errors.
screenshot

 

[Eximo]

Well, unigine is pretty light on the CPU if I recall. And if it can do furmark without crashing, might not be a power problem. Cyberpunk I imagine would stress the CPU and GPU to the max.

528W on the 12V rail, should be enough for a locked processor and a 1080Ti. Not the best supply out there, but not the worst.

Underclocking the GPU, if that helped, might be worth pursuing. But 80% is a bit much, try lowering the memory frequency a bit and see if that helps. If it doesn't set memory back to normal and lower the clock speed a little. Just see if one or the other is really causing the problem.

Oddly you may need more power, so another option would be to max the power limit on the GPU and see if it behaves any better.

 

[riskal]

Well, unigine is pretty light on the CPU if I recall. And if it can do furmark without crashing, might not be a power problem. Cyberpunk I imagine would stress the CPU and GPU to the max.

528W on the 12V rail, should be enough for a locked processor and a 1080Ti. Not the best supply out there, but not the worst.

Underclocking the GPU, if that helped, might be worth pursuing. But 80% is a bit much, try lowering the memory frequency a bit and see if that helps. If it doesn't set memory back to normal and lower the clock speed a little. Just see if one or the other is really causing the problem.

Oddly you may need more power, so another option would be to max the power limit on the GPU and see if it behaves any better.

I've tried lowering the GPU Memory Clock. First I've tried it without any change, the game crashed and returned the LiveKernelEvent 141, on the next run I've set it to -100MHz and the same happened. But then I've set it to -200MHz and it completely freezed my PC and required power button restart, this time it also returned LiveKernelEvent 1a1 instead of 141. Also it always crashes/freezes when the GPU reaches 76-78°C which can be seen in the photo bellow.
screenshot

 

[helper800]

I would attempt to put all the settings to default for the GPU and then set the power limit to 50%. That would determine if there is an issue with total power delivery from the PSU. If it still crashes while the GPU only needs 50% power than the problem/s is most likely unrelated to the PSU. Another thing I would suggest is to manually put GPU fan speed at 100% for te duration of testing to see if it hitting 76-78 C is the cause of the issue. At 100% fan speed it will be much harder to hit those temps, and if it does not crash after setting the fans to those speeds I would suspect the GPU as the issue. The exact issue this could be with the GPU at that point would be nothing more than speculation from me.

 

[riskal]

I would attempt to put all the settings to default for the GPU and then set the power limit to 50%. That would determine if there is an issue with total power delivery from the PSU. If it still crashes while the GPU only needs 50% power than the problem/s is most likely unrelated to the PSU. Another thing I would suggest is to manually put GPU fan speed at 100% for te duration of testing to see if it hitting 76-78 C is the cause of the issue. At 100% fan speed it will be much harder to hit those temps, and if it does not crash after setting the fans to those speeds I would suspect the GPU as the issue. The exact issue this could be with the GPU at that point would be nothing more than speculation from me.

I used it for around 4 months on 80% mostly for playing Cyberpunk 2077 and Overwatch and I've never had any crashes or freezes while the GPU was on 80%. I think that thermals may be a little bit of an issue here since the GPU stabilizes on 91°C in FurMark but I don't think that the temps cause the crashes since if it can reach 91°C in FurMark, 76°C shouldn't be a problem if I'm not missing something. I should definitely change the thermal paste and pads tho cause I doubt that the previous owner did it.

 

[helper800]

I used it for around 4 months on 80% mostly for playing Cyberpunk 2077 and Overwatch and I've never had any crashes or freezes while the GPU was on 80%. I think that thermals may be a little bit of an issue here since the GPU stabilizes on 91°C in FurMark but I don't think that the temps cause the crashes since if it can reach 91°C in FurMark, 76°C shouldn't be a problem if I'm not missing something. I should definitely change the thermal paste and pads tho cause I doubt that the previous owner did it.

So you have already been running the GPU on a reduced power profile of 80% to avoid crashing? If so I highly suspect the PSU on not delivering necessary power either in amount, cleanly, or properly at all. 91c is hot but Furmark is also ridiculous. Depending on ambient temperature where you live 85c and lower on max load for taxing games is acceptable. Only replace thermal pads and paste for the GPU if you are confident you know what you are doing. Since the heatsink for GPU's rests on the GPU silicon directly be careful with the amount of paste used especially if its conductive.

 

[riskal]

I've tried reinstalling the drivers with the DDU and it made it a little bit better I guess?

After doing so I've started Cyberpunk 2077 and played for much longer the temp got to 84°C (instead crashing instanly on 76-79°C) in that moment I've alt+tab and the PC froze again. I've waited for few minutes and both of my monitors went black for a second and then it unfroze. It showed 3 LiveKernelEvents 141, then Cyberpunk 2077 stopped working and an LiveKernelEvent 117. And in event viewer it showed Event ID 14 nvlddmkm.

After this I've tried Cyberpunk 2077 one more time and the temp got to stable 83-84°C but after I got to really crowded area, the temp went to 90-91°C and after like a minute it crashe again but this time after waiting few minutes it restarted the whole PC and I've got some new errors. First of all the shutdown was BlueScreen Code 133 and in event viewer I got 3 critical errors: 41 Kernel-Power, 10110 DriverFrameworks-UserMode and 10111 DriverFrameworks-UserMode.

EDIT: I've tried running Unigine Heaven again and it crashed on 76°C again with 141 and then Cyberpunk 2077 which also crashed around that temp also with 141.

 

[helper800]

I would try a few things just to make sure its not them even though its a long shot.

1. Open Start.

2. Search for Command Prompt, right-click the top result, and select the Run as administrator option.

3. Type the following command and press Enter, let run until its complete:
SFC /scannow

4. Type the following command and press Enter, let run until its complete:
DISM /Online /Cleanup-Image /CheckHealth

5. Type the following command and press Enter, let run until its complete:
DISM /Online /Cleanup-Image /ScanHealth

6. Type the following command and press Enter, let run until its complete:
DISM /Online /Cleanup-Image /RestoreHealth

7. If step 3 failed but steps 4-6 found corruption and fixed it repeat step 3 again.

This is to make sure there is nothing wrong with your win 10 install. The next thing I would suspect before the graphics card is the PSU not delivering enough power / amperage to the GPU. I would buy a decent Seasonic or Corsair 600+ watt PSU and try that. If that does not solve your issue there may be something wrong with the card itself. At that point you are looking at chasing ghosts trying to fix it. There are plenty of other people much smarter than me that I hope can help you if we cannot figure it out.

 

[riskal]

3.
Windows Resource Protection did not find any integrity violations.
4.
No component store corruption detected.
The operation completed successfully.
5.
No component store corruption detected.
The operation completed successfully.
6.
The restore operation completed successfully.
The operation completed successfully.

I'll try to go to a PC repair/store on monday and ask if it would be possible to lend some better PSU to test it with and update after I'll test it with the PSU.

Thanks for taking your time to help me btw, much appreciated!