Gaming causes PC to display a black screen and crash.
Gaming causes PC to display a black screen and crash.
I've experienced this problem for a while now. Over the past year or more, my PC tends to crash after about 20 to 30 minutes of playing demanding games. It shuts down abruptly, becomes unresponsive, and the fans spin at full speed. Usually, I can still hear the game audio and manage chats during calls when this occurs. Using the Windows command to refresh the graphics driver (Win+Ctrl+Shift+B) doesn't resolve the issue.
I've run MSI Afterburner on my second monitor to check temperatures while gaming, and it never spikes excessively—like the 3070 reaching a max of 82-83°C and staying stable. The crashes only happen during gameplay. When I use furmark or prime95 to stress-test the GPU and CPU, it doesn't trigger the crash, which suggests it's not a thermal problem.
I've gone through several driver/bios updates, but everything seems up to date. I haven't tried a full Windows reinstall because my boot drive is a 1TB NVMe SSD with limited space for backups. I've had GPU issues before, but this time there are none. My thought is either the GPU itself is faulty, or the power supply unit is the culprit. From what I found online, similar crashes have been fixed with a new PSU.
I also want to note that BeQuiets' PSU calculator tool says my PSU fits my setup. My current specs include an AMD Ryzen 5900X, Gigabyte B550M AORUS PRO, Gigabyte RTX 3070 Gaming OC, 700W be quiet! PPower 11 CM, 32GB RAM, XMP 3600 (4x8GB), and I don't manually overclock beyond XMP. My monitors are 3440x1440 and 2560x1080, running Windows 10. Any advice would be greatly appreciated.
To energize your GPU, you may need either one or two cables. It's worth testing it out. Have you carefully reconnected all connectors and reinserted your GPU? Are you running at default settings? Did you attempt it without XMP on your RAM and only with 16GB? Consider running a memtest86 to verify your memory. Are your BIOS, Windows 10, and drivers current? Edited August 5, 2023 by leclod
Currently connecting with one cable works for both the 8 and 6 pins. I’ll test another cable to confirm. Yes, the GPU is properly seated and all PC connections are correct. Everything functions as expected, just like when it came out of the box. The only adjustment I made was enabling XMP. I plan to try this myself and share the results.
My Windows, BIOS, and drivers are all current. Problems also appeared on older, updated versions, so I’m confident it’s not a version-specific issue. I’ll check if a memtest or an additional GPU cable resolves it. If my PSU supports a modular GPU cable, that could help.
You've already tested several solutions and haven't found a clear fix. The issue appears to worsen during alt-tabbing or scene transitions, possibly due to temporary load spikes. No data supports this yet, but you're considering replacing the power supply unit as a potential remedy.
I had my PSU returned for replacement, but tests showed it failed under load. The new 750W Bequiet PSU didn’t fix the problems I was experiencing. I’m now running an extended stress test on both CPU and GPU to make sure I cover everything. I switched from Windows 10 to Windows 11 without reinstalling, yet the issues persist. I’ll keep updating until a proper solution is found.