[SOLVED] GPU overheating

Feb 10, 2020
12
1
15
So I just built a new PC and my GPU is physically getting scolding hot when I try to run any graphic intensive game for more than 10 minutes and has crashed my PC. I feel like the system should have plenty of fans to properly cool. Could GPU be defective? Am I not properly cooling? I have 3 Intake fans in the front, 1 exhaust fan in the back, and 2 exhaust fans on top. Any ideas??? (See below for parts list)

CPU - Ryzen 7 2700x
CPU cooler - Wraith prism (stock cooler)
GPU - MSI RTX 2070 Super Gaming X
Motherboard - ASUS B450-F Gaming
Case - cooler master TD500

Full list: https://pcpartpicker.com/user/IllWhiteBread/saved/g3f9NG
 
Last edited:
Solution
Create a custom fan curve via MSI Afterburner or a similar program were the fans should be at 100% by 80C.

Also enable vsync and cap your FPS, this will help your GPU from getting hotter then it needs to. What is your monitor's refresh rate?
Graphics cards do get very hot under heavy load but specifically what are your card's and CPU temperatures under heavy load?
This will help to establish any problems.

Thanks

Andy
Graphics cards do get very hot under heavy load but specifically what are your card's and CPU temperatures under heavy load?
This will help to establish any problems.

Thanks

Andy
GPU will reach 85+ °C. I do have it set not to reach past 80°C.
I have not seen the CPU go past low 60's until the GPU really starts producing heat, then it jumps to 70's
I am not sure how high it has gotten when it crashed.
 
Create a custom fan curve via MSI Afterburner or a similar program were the fans should be at 100% by 80C.

Also enable vsync and cap your FPS, this will help your GPU from getting hotter then it needs to. What is your monitor's refresh rate?
I do believe the fans are set to run 100% on afterburner but I will have to confirm that when I get home later today.
The refresh rate of my monitor is 144Hz.
 
Do you run your games above 144hz? If so cap them and see what your gaming temps are.
Everything I run is between 60Hz adn 144Hz.

So far I have run everything high - ultra graphics, with uncapped fps(after your recommendation I will defiantly cap now).
For some game examples:
PUBG, Hunt, and COD will heat past 80°C within 10 minutes and crash after two games, so maybe about an hour.
I managed to play GTA5 for about 4 hours with 70-80°C until the last 20 minutes it started heating past 80 and the exhaust air stared getting warm.
 
Ok so that's good.
So when I get home tonight I will enable some fps caps, check and make sure GPU fans are set to run as they should, and I will monitor temps and see what happens.

Say these don't change and the back-plate of the GPU is still getting as hot, and core temps are still rising, what problem would you think I have?
Bad airflow? (I'm not that experienced but it doesn't seem terrible) Maybe a defective card?
 
Well I'd make sure to look into the case when it's running and make sure the GPU fans are spinning. While there are 3 fans in the front maybe it still has poor intake. Also monitor fan speeds to make sure they are spinning faster with increasing temps. On that case there doesn't appear to have much of intake into the case which may be part of the issue. If the front was mesh then it would work a lot better then with it covered the way they made it.
 
Well I'd make sure to look into the case when it's running and make sure the GPU fans are spinning. While there are 3 fans in the front maybe it still has poor intake. Also monitor fan speeds to make sure they are spinning faster with increasing temps. On that case there doesn't appear to have much of intake into the case which may be part of the issue. If the front was mesh then it would work a lot better then with it covered the way they made it.
I have already confirmed the GPU fans are spinning. I will just need to confirm they are gaining speed as temps rise.
With the case issue, I have already come up with a solution. I have temporarily removed the acrylic cover and left it open. I plan on grabbing a mesh sheet and replacing it with that. After I did remove that cover, I did not notice a drastic temperature difference, even with the increased airflow after it was removed.
 
broo I did this. Maybe you should try it. I was getting 83 degrees while I was playing anygame. my solution to this problem was set the temp limit at 73 using MSI afterburner. The thing is the GPU goes to its highest clock because the NVIDIA boost and you can't control it. for example, my rtx 2060 has a clock around 1750MHz but always goes to 1950MHz automatically. Setting the temp limits around 75 limit the gpu power and the clock frequencies. now Im getting 75 degrees and a clock around 1850 that it's still higher than factory base clock.
 
  • Like
Reactions: Illwhitebread
broo I did this. Maybe you should try it. I was getting 83 degrees while I was playing anygame. my solution to this problem was set the temp limit at 73 using MSI afterburner. The thing is the GPU goes to its highest clock because the NVIDIA boost and you can't control it. for example, my rtx 2060 has a clock around 1750MHz but always goes to 1950MHz automatically. Setting the temp limits around 75 limit the gpu power and the clock frequencies. now Im getting 75 degrees and a clock around 1850 that it's still higher than factory base clock.
I'll give it a shot. Thanks for the tip!
 
Hmm, usually with the side off it should drop the temps at least a few degrees Celsius. Anyways for the GPU fan speeds the newest version of HWmonitor works well if it doesn't report it correctly in MSI Afterburner (it's been awhile since I've used Afterburner).
Great news! did everything you recommended and it's doing great with temps! I appreciate all the help!
 
  • Like
Reactions: WildCard999