Hello everyone!
I've been using an ASUS TUF RTX 3080 Ti OC (12GB) for the past three years. For at least the last two, I’ve been running it undervolted to improve performance and efficiency. While I never monitored temperatures too closely, I noticed it generally stayed around 70°C under load. Occasionally, I experienced game crashes, which I assumed were related to my undervolt settings. So, recently I decided to revert everything to stock settings and fine-tune a more stable profile. While doing this, I noticed in MSI Afterburner that my temperature limit was being reached, causing thermal throttling, despite the GPU reporting only ~70°C. This happened regardless of whether the temp limit was set to the default 83°C or the maximum 91°C.
To investigate further, I checked HWiNFO and GPU-Z and found that while both GPU and Memory Junction temperatures were around 70°C, the Hot Spot temperature was significantly higher. In F1 24, it spikes to nearly 106°C very quickly, while in Forza Horizon 5, it stabilizes around 100°C. I think these are currently my most demanding games.
Here’s a screenshot of my temps: Imgur Link.
After some research and watching youtube videos, I considered installing a copper plate for the memory modules around the GPU die, but it seems like CoolMyGPU is no longer available, and I’m unsure where to find a proper alternative. Instead, I’m now thinking about replacing only the thermal paste on the GPU die while keeping the original thermal pads, since memory temperatures appear to be fine. I’ve never opened the GPU before, so I’m not sure if the high hot spot temperature is due to poor contact on the die or an issue with the memory modules and their thermal pads.
If anyone has experience with this card or insights into high hot spot temperatures and possible solutions, I’d really appreciate your advice. Thanks in advance!
Relevant system specs:I've been using an ASUS TUF RTX 3080 Ti OC (12GB) for the past three years. For at least the last two, I’ve been running it undervolted to improve performance and efficiency. While I never monitored temperatures too closely, I noticed it generally stayed around 70°C under load. Occasionally, I experienced game crashes, which I assumed were related to my undervolt settings. So, recently I decided to revert everything to stock settings and fine-tune a more stable profile. While doing this, I noticed in MSI Afterburner that my temperature limit was being reached, causing thermal throttling, despite the GPU reporting only ~70°C. This happened regardless of whether the temp limit was set to the default 83°C or the maximum 91°C.
To investigate further, I checked HWiNFO and GPU-Z and found that while both GPU and Memory Junction temperatures were around 70°C, the Hot Spot temperature was significantly higher. In F1 24, it spikes to nearly 106°C very quickly, while in Forza Horizon 5, it stabilizes around 100°C. I think these are currently my most demanding games.
Here’s a screenshot of my temps: Imgur Link.
After some research and watching youtube videos, I considered installing a copper plate for the memory modules around the GPU die, but it seems like CoolMyGPU is no longer available, and I’m unsure where to find a proper alternative. Instead, I’m now thinking about replacing only the thermal paste on the GPU die while keeping the original thermal pads, since memory temperatures appear to be fine. I’ve never opened the GPU before, so I’m not sure if the high hot spot temperature is due to poor contact on the die or an issue with the memory modules and their thermal pads.
If anyone has experience with this card or insights into high hot spot temperatures and possible solutions, I’d really appreciate your advice. Thanks in advance!
- Motherboard: Asus Prime X570-P
- CPU: AMD Ryzen 7 5800X
- CPU Cooler: Noctua NH-D15
- RAM: 32GB DDR4
- Case: Fractal Design Torrent Compact
- Displays: TUF Gaming VG27VH1B (1080p) and Samsung Odyssey Neo G7 (4K)
Facebook
Twitter
Instagram