Game crashes due to GPU-related issues during playback
Game crashes due to GPU-related issues during playback
Right after beginning any graphically demanding game, my monitor freezes with GPU-related glitches. Occasionally the sound keeps playing while the cursor remains stationary, and at other times the whole system halts. I've experimented with adding three GPUs to this setup. An R9 280 caused similar problems in games and even artifacts during BIOS updates. The GTX 560 worked fine, while the RX 580 only showed issues during gameplay but not when undervolted. Despite cleaning the entire PC multiple times, all three GPUs were in the same slot, suggesting another factor. I've updated the BIOS and tried several Linux distributions without success. I'm looking for a photo of the artifacts if that would assist. The motherboard model is GA-F2A88XM-D3H with F8 BIOS, and the GPU is an RX 580 running AMD Athlon X4 860K Quad. The OS is 64-bit Arch Linux.
These together suggest a possible overheating issue. The low-heat cards function, while high-heat ones fail, and lowering the clock on a high-heat card can sometimes resolve it. Another potential cause could be a PSU voltage output problem, though heat issues are more frequent and easier to test. A simple test is to flood the case with cold air—remove the side panel, aim a fan at the opening, and run it at full speed. If the system operates normally, the problem is likely with the heat dissipation. Any method that forces airflow onto the graphics card should help identify the cause.
it functions on a standard operating system. this version has been running for about nine years, and its continued operation without any OS-level quality checks is impressive. the last BIOS release was six years ago, so software updated more recent than that should likely avoid compatibility issues.
I experienced the highest temperature I've ever seen my RX 580 reach before it froze, around 48 degrees Celsius. After replacing the thermal paste on my R9 280, the issue persisted both before and after the change. It seems unrelated to other factors—my GPU is quite old, only showing these problems recently, possibly last year. I've been using the same BIOS version with the same hardware for many years prior to that.
It might be useful if the task is simple. Do you have another reliable PSU available for testing? There’s also a memory check option.
Tried it and then faced a total shutdown. The fans ceased operation, the screen went black, and I couldn't power it off. No gaming or heavy tasks were running. I had to unplug the PSU and restart it. Since my new 600-watt unit is brand new, I’m unsure what alternative memory testing would help.
Well then it’s not about heat. That’s why it’s a trial. The alternate memory test is quite old and somewhat makes do with a workaround. It doesn’t need software in the strict sense. It only functions with more than one stick. It’s before the internet era. It depends on the idea that it’s highly improbable for more than one stick in a kit to fail at once (if they do, it’s useless). What you do is insert just one stick of memory into the device and check if it works. If all succeed, it’s likely a different memory port or something unusual is happening. If one or three succeed with two or four, but one fails, that faulty stick indicates a bad memory and you replace the kit. If all fail, it’s not memory at all. It can be time-consuming depending on the issue and requires a lot of effort. App-based memory tests tend to be more reliable and can detect many issues. They’re generally more precise than software, though. It’s usually reserved for situations where other tests have failed and you still can’t pinpoint the problem. Edited August 30, 2022 by Bombastinator