• Welcome to TechPowerUp Forums, Guest! Please check out our forum guidelines for info related to our community.
  • The forums have been upgraded with support for dark mode. By default it will follow the setting on your system/browser. You may override it by scrolling to the end of the page and clicking the gears icon.

GPU Overheating Problem

HojoKing

New Member
Joined
Oct 22, 2013
Messages
2 (0.00/day)
I have GPU Overheating issue on my custom built pc. The temp has risen to 97 deg C after playing games for 2 mins. I've had this pc for almost 2 years already and overheating gpu just happened last week. I used HWMonitor to see the temp when playing games. The cpu is around 40 deg C but GPU was around 97 deg C before shutting down. I have googled around for answers but I have found no solution thus far.

-I used 2 cans of compressed air to take out all the dust
-I tried playing the game(D3) with the side panel opened
-I don't think air-flow is the problem
-I did memtest..only 1 pass though
-I installed a few latest nvidia drivers but that didn't stop the overheating
-I used an MSI app to change the fan speed on my vid card(EVGA GTX570) from 45 to 67

Please suggest on what I should try next if you can help. Could it be my power supply? I have a Thermaltake 750w. Will cleaning the heatsink and reapply arctic silver paste solve this problem? I have never done that before so it'll be VERY NEW to me. Can anybody help? Thanks in advance.
 
Replace the TIM on the GPU cooler. 2 year old paste is a bit... meh...
 
It might be a fan problem, but lowering core clock must be the case.
Try to lower voltage and clocks using Asus GPU Tweak software.
I use that for high temps on the gpu
 
It might be a fan problem, but lowering core clock must be the case.
Try to lower voltage and clocks using Asus GPU Tweak software.
I use that for high temps on the gpu

Underclock from stock settings? That's a pretty stupid idea IMO.
 
What kind of cooler is it ?, as some designs get build up that can be a bit more awkward to remove so please post the card you are using..

Replacing the paste ( eg MX-2\4 ) is a good idea but be careful not to over tighten when you put the cooler back on.



If this type of cooler in the link below you normally get a wall of dust which blocks all thee air flow. So get torch\flashlight to make sure it's all removed

EVGA 012-P3-1578-AR GeForce GTX 570 (Fermi) Classi...
 
Underclock from stock settings? That's a pretty stupid idea IMO.

It will save the card at least or try to underclock it so we see if it is the high clocks problem
 
It will save the card at least or try to underclock it so we see if it is the high clocks problem

>Video card is fine for 2 years
>Suddenly, last week, temps soar to 97 degrees
Clocks are not the issue here, they dont spontaneously cause extreme overheating overnight. If clocks and voltage were the issue, it would have been gradual over weeks or months. Unless the OP has been correcting VDroop over the last two years by slowly increasing voltage to maintain stability, clocks and voltage arent to blame.
Likelihood is the fan has failed, or randomly fails at a certain point during loaded use, or the TIM has become a crusty pile of crackers between the die and the copper.
 
Thanks for all the suggestions you guys. Here is my GPU, I went to newegg and checked my order history in 2011 to find the exact model

EVGA 012-P3-1570-AR GeForce GTX 570 (Fermi) 1280MB 320-bit GDDR5 PCI Express 2.0 x16 HDCP Ready SLI Support Video Card
Item #: N82E16814130593

I'm also suspecting that it could be my power supply but I am not sure. I thought I used to have a neon led light coming out of there but I can't confirm. But here is my PSU model.

Thermaltake TR2 RX 750W Bronze W0382RU ATX 12V v2.3 / EPS 12V v2.91 SLI Ready CrossFire Ready 80 PLUS BRONZE Certified Modular ...

I will try to order some TIM and browse youtube on a crash course...
 
Before you do that, check if it's still under warranty.
 
Thanks for all the suggestions you guys. Here is my GPU, I went to newegg and checked my order history in 2011 to find the exact model

EVGA 012-P3-1570-AR GeForce GTX 570 (Fermi) 1280MB 320-bit GDDR5 PCI Express 2.0 x16 HDCP Ready SLI Support Video Card
Item #: N82E16814130593

I'm also suspecting that it could be my power supply but I am not sure. I thought I used to have a neon led light coming out of there but I can't confirm. But here is my PSU model.

Thermaltake TR2 RX 750W Bronze W0382RU ATX 12V v2.3 / EPS 12V v2.91 SLI Ready CrossFire Ready 80 PLUS BRONZE Certified Modular ...

I will try to order some TIM and browse youtube on a crash course...

Get in touch with eVGA aand see what they recommend so you have the best chance to keep your warranty in tact.

Watch it though if they change the card for some thing else as they have a habit in changing things with a lesser warranty although they might give you a better card but still.

Before you do that, check if it's still under warranty.

Well with it being AR and if eVGA have kept the same warranty lettering it should be lifetime warranty
 
wait... diablo 3? why dont you just cap the framerate to 60?

just watch the fan speed & temperatures change in real time while the game is open with gpuz or afterburner to see if something looks suspicious (like low fan speed, normal temps everywhere but high gpu temp only, whatever clues are around to decide what steps to take next)
 
Manually turn the fans to 100%
 
wait... diablo 3? why dont you just cap the framerate to 60?

just watch the fan speed & temperatures change in real time while the game is open with gpuz or afterburner to see if something looks suspicious (like low fan speed, normal temps everywhere but high gpu temp only, whatever clues are around to decide what steps to take next)

Well GPU-Z woth TPU OSD apps then you will not need to minimize the screen all so you get CPU load, cpu temp, cpu clock and voltage for a added bonus.

Gotta love those VRM's on the 6970 lol.
107410_screenshots_2013-10-05_00003.jpg
 
  • Like
Reactions: xvi
>Video card is fine for 2 years
>Suddenly, last week, temps soar to 97 degrees
Clocks are not the issue here, they dont spontaneously cause extreme overheating overnight. If clocks and voltage were the issue, it would have been gradual over weeks or months. Unless the OP has been correcting VDroop over the last two years by slowly increasing voltage to maintain stability, clocks and voltage arent to blame.
Likelihood is the fan has failed, or randomly fails at a certain point during loaded use, or the TIM has become a crusty pile of crackers between the die and the copper.

+1 to this. i had an old 8800GTS 512MB that had random fan stoppages. actually, the fan control became inverted, and it tried to up the speed, which caused the fan to stop. THAT was annoying.
 
little bit off topic

little bit off topic but it somehow connects to OP problem...

some of you guys recommend replacing TIM...

just a quick question, is it ok to use CPU tims on GPU? I have the noctua thermal paste that comes with my NH-D14 cooler. Just dont want to buy another tim. I dont have problem yet but as RCoon mentioned, 2 years tim is like MEH... i might encounter problems sooner cause my gpu's will be almost 2 years next year...

so just wonder if I can use that TIM i have already...

again sorry for the off topic. I dont intend to high jack your thread OP.... :toast:
 
little bit off topic but it somehow connects to OP problem...

some of you guys recommend replacing TIM...

just a quick question, is it ok to use CPU tims on GPU? I have the noctua thermal paste that comes with my NH-D14 cooler. Just dont want to buy another tim. I dont have problem yet but as RCoon mentioned, 2 years tim is like MEH... i might encounter problems sooner cause my gpu's will be almost 2 years next year...

so just wonder if I can use that TIM i have already...

again sorry for the off topic. I dont intend to high jack your thread OP.... :toast:

Yes, you can use that paste. But, you need to be more careful when applying it than you would with a CPU, especially if the paste is conductive. There are a lot more electrical points on the GPU that you have to insure you don't get any paste on, especially conductive paste.
 
Back
Top