My ASUS RTX 4080 Super has an unusual overheating problem. As soon as I launch a game, the GPU hotspot jumps to around 105–109°C, right around the throttling temperature. I repasted the card, but the behavior stayed the same, and switching to PTM7950 phase-change material also made no difference. The thermal pads are still the original ones, and the card is out of warranty, so I'm trying to figure out whether this is a mounting, pad thickness, airflow, or hardware issue before replacing the whole card. At idle, the core is about 44°C and the hotspot is roughly 54°C.
3 Answers
It’s worth testing with the side panel removed to rule out case airflow, but airflow usually raises the overall core temperature rather than causing an instant 30°C-plus hotspot gap. Also check whether any overclock, undervolt, or aggressive power setting is applied. If the core stays around the mid-40s while the hotspot shoots past 100°C, cooler contact is still the more likely cause.
The hotspot being 30°C or more above the core strongly suggests poor contact between the GPU die and the cooler rather than ordinary heat buildup. Since it spikes almost immediately when a game starts, check the paste or PTM imprint after removing the cooler. A patchy imprint, especially on one side, can indicate uneven mounting pressure. Thermal pads that are too thick can also hold the heatsink slightly away from the die, even if they’re the original pads.
The factory thermal pads can still be part of the problem if their thickness or compression is preventing the cooler from sitting flat. Graphics cards often use several different pad thicknesses in different locations. Replacing them with incorrectly sized pads can make contact worse; thermal putty is sometimes safer if the exact specifications aren’t known, but the cooler should first be checked for a clear, even imprint.
I switched to PTM7950 to test it, but the result was the same. The pads are still the factory ones, so I’m going to inspect the contact pattern and mounting pressure next.