High temps and crashes?
High temps and crashes?
Hi Team,
I'm looking for some expert advice on this issue.
Over the past year or so, I've been playing BF V and now BF6, and the game would crash around 10 minutes in. Sometimes even earlier.
I've updated drivers multiple times, and recently I finally upgraded to Windows 11, but it still crashes after a clean install. I thought it might be due to Windows 10, but that didn't seem right.
After more research, I suspect the graphics card is overheating. Someone in the BF6 chat suggested running the game in a borderless window, which resolved the problem. It's unclear why this worked, but it seems like an unusual workaround.
I also checked Fur Mark to monitor temperatures and noticed they're rising.
System Specs:
- CPU: Ryzen 5950X
- GPU: Asus 3060 TI TUF
- CPU Cooler: Artic 360 Liquid Freezer III (1 year old)
- RAM: 128GB
- Case: Corsair Air Flow 4000D (a bit tight, maybe a larger case would help)
- SSD: Samsung 990 Pro 4TB
Fur Mark (GL) shows:
- GPU Usage at 100%
- GPU Temp at 83°C
- GPU Hotspot at 105°C
- Fans running at full speed
- GPU-Z reports TEMP values.
I also ran Timespy, which crashed in about 10 seconds to a blank screen. I had to force a restart. This happened before, so it reinforces the idea that overheating is likely the cause. The Steel Nomad version didn't crash.
83°C feels extremely high for this setup. What should be done next? From what I understand, applying new thermal paste to the GPU seems reasonable. I haven’t tried it yet. Do you think it’s worth doing? I have some leftover Noctua NT-H2 from my CPU cooler—should I use that or is there a better option?
Where exactly is the hotspot located? On the GPU die, or elsewhere?
I’d like to keep this card running for longer since the overall situation is confusing.
I recall that when I bought this card in 2021, it was the only GPU available at the store, so I just picked it up. I planned to upgrade eventually, but five years later I don’t mind waiting. I don’t play much these days.
Any suggestions or advice would be greatly appreciated. Thanks!
My fan and cooler setup is designed to keep things running smoothly.
AIO Rad is positioned at the front with three fans on the front face and two more mounted on the back drawing air in. One of the fans aligns with the memory/cpu area. The other is now directly in front of the GPU. I didn't have this fan before but considered trying it to check if it would help. So far, I haven't noticed any changes.
No, this is a classic symptom of the GPU overheating because of the thermal paste pumping out. You need to repaste the GPU to fix the issue.
It has likely suffered the same thermal paste pump out issue a lot of RTX 30 series have had, and the ASUS TUF cards seem to be among the models that suffer from this issue more often. I had to repaste both my own 30 series TUF cards.
I understand, I see.
Any advice for someone new to this? I've just glanced at some YouTube videos recently.
What paste are you planning to use? I still have some leftover Noctua NT-H2 paste. Is that sufficient?
Cheers
I think the RTX 3060 TI TUF matches the cooler of my 3070, meaning it should require only eight #00 Phillips head screws for removal.
I applied MX-6 on my 3070 since its thicker viscosity helps resist pumpout more effectively, though I suggest using MX-7 for even better resistance and improved cooling. The application method differs from other pastes and shouldn’t be spread with a spatula tool.
You might also consider the zero paste Thermal Grizzly Kryosheet, a graphene sheet TIM, which could eliminate the need to repaste the GPU over time. For the 3060 TI, you’ll need a 25mm x 25mm sheet for the die surface.
Another alternative without paste is PTM 7950, which generally performs better than the kyrosheet but may require some effort to cut to the correct size yourself.