Random driver timeouts occur during gameplay on 7900xtx.
Random driver timeouts occur during gameplay on 7900xtx.
Title: 5800x3d, 32GB RAM @3200, 7900XTX, TUF X570+, Corsair PSU (in use for 6 years). Used this card for almost 8 months. Recently experienced freezes or crashes only. Occasionally during games, the screen would freeze briefly with sound, then go black without sound. Sometimes I could close the game manager to end it, or if it didn’t work, I needed a full PC restart. When no restart was required, I saw an AMD driver timeout message, and in some games like Horizon Forbidden West and Valorant, there were similar warnings saying “Your video driver is too old, please update.” I tried removing Nividia and AMD drivers using DDU in safe mode, then reinstalled the latest AMD drivers. After switching my 2080ti to 7900XTX 8 months ago and again today, I ran a memtest and it passed. The BIOS was updated to the newest version. Temperatures in games stayed around 70°C CPU and 50°C GPU, which were stable and within limits. I’m running out of ideas and would really appreciate any advice. Thanks ahead.
Another user is facing the same problem with their 7900xt. Their fix was avoiding dx12 mode in whatever game caused it, which happens consistently. After switching to dx11, there have been fewer occurrences. They also mention having a 7800xt that experiences similar issues occasionally, but not too often enough to change settings.
Did you ever fix this problem? I’m facing the same issue with my 7900xtx and planning to revert to Nvidia since it keeps happening during Hunt Showdown for DX11. I’ve reinstalled Windows, which should prevent it from being a glitchy setup.
If anyone else encounters this problem, here are some possible causes. My GPU ran hot on the Hotspot and GCD even though temperatures were normal and it was watercooled. I noticed this in the hardware monitor. Raising the pump speed on my cooling loop fixed it. For others, they might just boost fan speeds. Once I resolved it, crashes stopped occurring. My GPU never exceeded 49.2°C during long gaming sessions, while GCD reached 100.
No, I didn't install hwmonitor or inspect the gcd and hotspot temperatures on the card.
Reduce clock speed by about 200Mhz intentionally to test stability. It seems automatic overclocking isn’t working—maybe it’s already near its limit.
I just put everything on stock. Perhaps adrenaline is really stepping up.