I'm experiencing freezing issues with my laptop's GTX 1650 Mobile GPU (in an Acer Nitro 5 AN515-43) during heavy usage, although it works well at moderate loads (50-80%). Stress tests like FurMark crash instantly, and games such as F1 2022 fail to load or crash shortly after starting. The GPU temperatures are normal, ranging from 40-60°C, and the integrated GPU operates without problems. I've already reinstalled drivers and confirmed the charger works fine.
To mitigate the issues so far, I've limited the GPU clock with nvidia-smi, set it to High Performance mode when plugged in, and maximized the fan speed. This makes FurMark stable and allows some games to run, but the freezes return if I attempt to increase clock speeds or reduce cooling.
I also need to mention that when I repasted, I used thermal paste instead of thermal pads for the VRAM and VRM. Now I'm wondering if I should replace the thermal paste with proper thermal pads to improve stability or allow for slightly higher clock speeds, or if this is likely a sign of hardware degradation that I'm just managing by underclocking. I'd appreciate any advice or similar experiences!
2 Answers
Definitely consider using thermal pads instead of paste for the VRAM and VRM. Those components often have gaps that need to be filled for better thermal contact. Your GPU should run fine if everything is set up correctly. Just stick to the approved methods for any clock limiting or cooling adjustments.
I doubt it's a hardware degradation issue; it seems more like a problem with the thermal contact. If you're using paste instead of thermal pads, you might be creating a gap, which can lead to instability especially under heavy load.
So should I just buy thermal pads for the VRAM?
I noticed thermal paste on the heatsink when I opened it, so I think it's making contact. Still, I guess I'll order some thermal pads just to be safe.