when stress is applied, graphics cards display dark screens while cooling speeds increase significantly
when stress is applied, graphics cards display dark screens while cooling speeds increase significantly
I've been facing this problem for four days now. Many others have reported it, but nobody has shared any fixes. As mentioned in the title, the issue appears suddenly—no updates or changes to my PC have triggered it. Once I start playing a game, everything functions properly until about 20 minutes in, when the screen goes black and the fan speeds reach maximum. I can still hear background sounds and speak on Discord. Temperatures were normal beforehand (voltage temps, etc.). This problem started without warning, not after any updates or adjustments. My setup is: MSI Z590 plus (MS-7d11), Intel i9-11900K, Gainward RTX 2080. I've tried everything: new power supply, dust cleaning, checking cables, reseating RAM/GPU/CPU, reinstalling Windows, rolling back NVIDIA drivers and installing an older one, scrolling through forums endlessly—still no solutions. I even disassembled the PC multiple times to check components and reattach everything. Since I can't find any crash dumps, I recorded the screen hoping to capture the error message, but the video got corrupted because of a crash. I really hope someone has a solution for this.
It seems there might be a thermal problem. Components could be reaching their temperature limits. It doesn't have to be something logged in Windows. The VRM and PSU have built-in overheating safeguards, but over time these can degrade naturally (like thermal pads on VRMs). Since you've already changed the PSU, try reapplying or adding more padding to the GPU—just need to determine the original thickness of the pads.
Yes, you can look up the specific pads required. Thermal paste might be enough if your system supports it, but checking the manufacturer's recommendations is best.
The VRAM and VRM connections don’t require paste, and the heatsink needs pads. You can locate them on Amazon; scissors will work for cutting them into strips. It’s likely you’ll find similar parts online—perhaps even searching on Bing Chat could help. The thickness usually ranges from 1mm to 2mm, often around 1mm. You’ll get a clearer understanding once you disassemble and check their condition after prolonged use.
Ensure all pads are measured accurately since some cards may feature 3 or 4 varying sizes ranging from 0.5 to 2 mm thick. Using the correct thickness is crucial; otherwise, contact issues may arise or uneven pressure could damage the GPU chip during stack remounting, leading to insufficient heat transfer in certain areas. In extreme cases, this can cause circuit board fatigue. Handle the GPU chip gently during cleaning, as tiny resistors are present around it—any failure might require sending the card to a repair shop for replacement. Apply high-quality cooling paste when reassembling to maintain optimal performance.