System problem detected: Possible dying GPU.
System problem detected: Possible dying GPU.
Hello people,
system specs:
OS: W11(was recently W10)
GPU: RTX 2080S (not overclocked)
CPU: Intel i9 9900K (not overclocked)
RAM: 32GB corsair vengeance RGB RAM
MoBo: GigaByte Z390 AORUS PRO
PSU; Corsair RM850X
Problem:
I've started having an issue with my system where, after around 10 minutes of playing Grand theft auto 5( A medium to heavy game to run) My monitors go black, as in they lose video signal. The fans of my graphics card suddenly spin at 100%. I can still hear audio from my spotify/game playing in the background so it only seems to be a video related issue.
After searching forum for some time, I was convinced it had to be an overheating issue. And so I have re-pasted my GPU with paste & pads from Gelid.
Temperatures have dropped significantly. Where before my GPU temp was 89C and hot spot temp was 110C. After playing a game for 25 minutes, they are now the following:
This was my test run from last night. Believing everything was fine now due to the heat issue being taken care of, I started up GTA again, and surely enough... after about 10-15 minutes my video feed cut out again.
I have found it very complicated to find any hardware related issue myself by doing online research and trying a few things. Looking at the windows event viewer, it does not show any errors or faults. Besides the power button being pressed to power the system off.
I have been logging data with GPU-Z and I cannot find any anomalies. There are currently 4 things that I am suspecting could be the problem.
GPU is dying
PSU is dying
Motherboard is dying
RAM somehow started failing
The reason i am suspecting the first 2 points is because if i limit my GPU power usage, my system does not crash. This is also how i keep my pc running nowadays.
I have also temporarily replaced my GPU with a GTX 1070 from one of my mates and I did not have any crashes running that.
Third point is what I often came across on the interwebs with mixed results.
4th point is because everytime my system has crashed, it has been just over 12GB ram used, but then again it does not happen when I limit my gpu power usage.
Benchmark programs like Prime95 and FurMark2 I can let run for hours without a single issue happening. I am hoping some of you could share your experience/expertise with me. Really am looking forward to learning more about this issue and what steps I can take to rule things out without spending hundreds of euros on a part im not even sure of is the problem.
Hello there,
Do you have a single game in mind or are you checking multiple titles? Are your system drivers and chipset up to date? Please confirm your GPU driver status as well. What BIOS version are you using for the motherboard?
Hello Roland,
I’ve reviewed your points one by one. The issue occurs across all the games I’ve tried—GTA 5, Timberborn, even during Discord sessions. It seems to happen sometimes. Should the Lan/Chipset be updated? I’m still figuring out how to do it properly. My GPU has the latest driver installed; before the problem started, it used a driver from August last year, and updating it was my first attempt but didn’t resolve it. I was using BIOS version F10, which I recently changed to F13. After playing GTA for about ten minutes, the video feed cut out again, resulting in a black screen, maximum fan speed, and a hard reboot via the power button. Thanks for your feedback, we’ll keep working on this.
Following the flash of bios, did you perform a CMOS clear? This can be done by pressing a button on the rear IO panel or by shorting two pins. Edit: Uninstall your GPU driver using DDU and then reinstall it fresh. Check if that resolves the issue. What power plan are you currently using?
I haven't cleared the CMOS yet.
I used DDU to remove the GPU driver and installed a new one, following the Windows safe mode advice.
I discovered an old Reddit discussion about the same problem. The person who posted initially said changing his power cable from the PSU to the GPU fixed it. I’ve done the same, and so far I’ve managed to play GTA for more than an hour without any issues.
How could something so important be caused by just one cable?
I’ll keep updating my thread in the coming days if the problem recurs or not. If it does, I’ll try clearing the CMOS first. Thanks for your support so far, Roland.
EDIT: system crashed again, now clearing CMOS and testing.
Cleared CMOS, turned on XMP and checked. After 10 minutes the video signal disappeared and fans ran at full speed. Unfortunately, no solution has been found yet. It's time for bedtime now—hopefully someone will suggest what to try next day. I won't have time to respond or test until Tuesday, as it's my birthday tomorrow. Have a great day everyone!
Update
System consistently loses video feed, GPU fans increase to full speed while temperatures stay below 60°C.
I noticed a change in my setup just before this problem began—sleep mode was enabled for 10 minutes. Could this be connected to the loss of video feed? It seems my Windows might be forcing sleep mode. I had reset the settings to never enter sleep mode, but it hasn’t fixed the issue.
I’m seeking advice on this matter.
I scanned your system using various tools for malware or viruses, but only a few false positives appeared. I followed a YouTube guide on repairing Windows that suggested additional scans, which uncovered corrupted files and some faulty drivers. I fixed those issues, though it didn’t fully resolve the problem. I’m currently reinstalling Windows.