Question 2080 Super OEM running very hot ?

KroolYouTube

Honorable
Jul 26, 2016
30
0
10,530
0
Hey guys I recently got an Alienware Aurora R11 that comes with an OEM 2080 Super.
It's been running very hot at 30-50% usage and I don't know why. The case is clean, open case with a large fan blowing cold air into the case.
I know these OEM cards suck at cooling but I don't know why this card even with a high power fan pushing cold air right into the card its pushing 87c with a light load playing older games.

I recently opened the card to clean the dust out of the blower fan and ever since then it's running MUCH hotter than previously on games with the same settings. Playing Smite maxed at 1440p would heat the card to around 75c max, and now it's pushing that 87c at the same exact settings. Did I mess up the paste, should I repaste the processor?

Please help!
 

KroolYouTube

Honorable
Jul 26, 2016
30
0
10,530
0
It's an OEM blower cooling system so you can't actually take it apart, and I don't know if garbage cards like these actually come with thermal pads?
 

Phaaze88

Titan
Ambassador
even with a high power fan pushing cold air right into the card its pushing 87c
A)Cooler not mounted on the PCB correctly.
Regardless of what you do, be it an undervolt, lowering the power limit, and even opening the chassis up to have a big ol' fan blast the inside with air, and the gpu core still runs 85C+...
Yeah, that would be a bad cooler mount - BUT! It could also be the following(B)...

B)Used the 'wrong' paste on the gpu die. The rough, grainy surface of a cpu IHS and the slick surface of a bare die... those surfaces do affect which pastes work well and don't.
The less viscous, or runny pastes, tend to not hold up as well on bare dies and are prone to getting pushed off by the cooler mounting pressure and high thermal cycles... so you then end up with air pockets between the die and cooler cold plate.

C)Wrong paste application. Usually pretty hard to screw this one up, unless there's specific application instructions from the manufacturer.
Using too much paste isn't a bad as using too little - it's just wasteful.

D)Cooler inadequate for the kind of power the gpu pulls.
Really don't think it's this, because you said you were getting 75C before, and that is a good, acceptable range to be around. The card's good up to ~82C.
The default thermal limit is 84C, with 88C being the absolute max before it considers shutting the PC off on you.
 

KroolYouTube

Honorable
Jul 26, 2016
30
0
10,530
0
So I decided to take apart the GPU again and repaste the die with some XTM50 that I used on my CPU. Sadly the thermal compound didn't work that well on my i7-10700F but the GPU was struggling to run timespy on stock clocks, hitting 85c or more within the first 2 minutes. After the repaste, it's now sitting at +75 core +850 memory running timespy hitting Temps around 77c before it stops climbing.
 

ASK THE COMMUNITY