Signal drops on RTX 3080 display after 40 minutes of use
Signal drops on RTX 3080 display after 40 minutes of use
Hello World, A fresh setup featuring a SECOND HAND RTX 3080 seems unstable. Whenever I start a game at 3840X1080 resolution, after roughly 40-45 minutes, the display disconnects and I need to restart it to regain control. (Unplugging and plugging the DP cable doesn’t help, and even switching screens while it’s down doesn’t resolve the issue.) So far, nothing has changed: at 1080P it never happens, not on my desktop. The build specs are i7-10700f, 16GB, NVMe 1TB SSD plus HDD, Asus B560M-A, and a 750W PSU. I recorded an hour of gameplay with HWinfo64 and the machine crashed after about 70-80 minutes. Someone might want to check the CSV report—I’d appreciate any feedback on the data.
Gradually moving toward clarity, focusing on the essentials while ignoring extraneous details. The process should run smoothly without interruptions, allowing you to review performance data after losing signal. Using HWinfo64 in the background will provide useful metrics like GPU power usage and temperature ranges, giving you a solid reference for troubleshooting.
Thank you for the details. Yes, you're right about the wattage. The TDP measurement during gameplay in Witcher 3 Enhanced Edition was a steady 162W. Unfortunately, this was the only session without any crashes. I plan to collect more data tonight and update the thread.
The room temperature was around 62 degrees. (There are 9 fans inside the case.)
The die temperature is acceptable, but the HWinfo64 displays memory hotspot readings as well. On my MSI 3080 Suprim, the memory hotspot reached close to 105°C, prompting me to upgrade with a full water block. In my view, the extra expense was worth it to ensure the card remained functional and avoid potential damage over time. If the temperatures are consistently high, it could indicate a safety concern that might cause the card to shut down. Many users have resolved this by replacing the thermal pads with higher-quality components, which are more reliable than the cheaper alternatives commonly used.
Checked the newest GPU firmware updates. Attempted to turn off MultiPlane Overlay MPO.