I built a relatively budget-friendly PC with a Ryzen 5 5600 running overclocked, a Thermalright Peerless Assassin 120 SE cooler, an ASUS ROG Strix RTX 2080 Ti 11GB, 32GB of DDR4-3600 memory, an MSI B450M motherboard, a 1TB NVMe SSD, and a 650W Bronze-rated power supply. The system is inside a Revenge Terranova case, and I'm planning to replace the power supply with a 750W Gold-rated model.
During about an hour of Counter-Strike 2 at 1080p, the GPU reaches around 75°C on the core with an 89°C hotspot. In a FurMark test, it reached 78°C core temperature, 94°C hotspot, and 92°C memory temperature at roughly 260W and 65% fan speed. My room is also quite warm and has no air conditioning. The GPU is supported with a small wooden block. Are these temperatures safe, and should I adjust the fan curve, improve the case airflow, or undervolt the GPU?
3 Answers
The case airflow is probably the biggest thing to check. Front, bottom, and side fans should bring cool air in, while the rear and top fans should exhaust it. Make sure the bottom fans are actually blowing upward toward the GPU rather than downward. The case’s restricted front and side panels may also be limiting airflow, so moving to a case with a proper mesh front could reduce temperatures noticeably.
Those temperatures are within a normal range for an RTX 2080 Ti. Around 75–78°C on the core and a hotspot in the low-to-mid 90s during gaming or stress testing are not dangerous. I’d only become concerned if the core were consistently approaching the วัน?°C limit, the hotspot were getting close to or above 100°C, or the card were throttling or showing instability.
That’s reassuring. I was worried because benchmark videos showed lower temperatures, but my room is pretty warm right now.
You can try a custom fan curve or undervolt the GPU with a tuning utility. A modest undervolt can reduce power draw and noise while keeping nearly the same performance, but it isn’t necessary based on these readings alone. Before repasting the card, improve the airflow and monitor temperatures under normal gaming loads. GPU thermal paste is usually only worth replacing if the hotspot gets unusually high or the core-to-hotspot difference becomes excessive.
I’ll try adjusting the fan curve first and keep monitoring it. I mostly wanted to make sure I wasn’t damaging the card.
The fans beside the GPU are configured as intakes. I suspected the case airflow was poor, although I originally chose the case because I liked how it looked.