Frustrated 980ti temps

[mojothehut]

Hey all.
So lately my EVGA 980ti Classified has been running stupidly hot.
Nothing about my case/ambient temps have changed.

In almost every game, Im regularly hitting 82-85c
It used to be 73-77c. I've taken the card out and carefully cleaned and inspected it.
No visual signs of heatsink separation or anything.

Even cranking my fan speed up to 85% hasn't helped much.

So now I'm considering just buying a new card. I really wanted to wait
for Volta, but now it sounds like that won't be out until next spring.

Are there any 1070s/1080s that are exceptionally cool on air? Just kind of
tired of the inferno my 980ti is causing.

 

[nurturedhate]

If the fans are working and it's clean it's more than likely the TIM causing the issue. I'd replace the TIM and check the temps. As for a replacement the only decent upgrade would be a 1080ti. 1070's are a sidegrade and overpriced and I wouldn't call a 1080 a worthwhile upgrade at $500+ for 10% to 30% (depending on your 980ti oc) worthwhile.

 

[mojothehut]

Well I took it apart, cleaned the paste off and applied some new. Ran some tests, only noticing about 1c drop on average. =\
Wish Vega would have forced Nvidia to release Volta sooner.

 

[EXCellR8]

i have a pair of 980 classified's and ones a kingpin... never need over 60% fan. are you overvolting a bit and can you confirm that BOTH fans on the card are working? do you use EVGA Precision X by any chance?

 

[maddie]

Could the heatpipes be damaged. My reasoning. A rare case admittedly, but not impossible.

If power hasn't increased hugely, then same GPU power consumption.
If new TIm hasn't worked then approx same heat from GPU is not being efficiently dissipated.
If fans are working as before then heatsink is not working as before.

Only reason left must be heatpipes not dissipating the generated heat over total heatsink surface area.

 

[mojothehut]

Both fans are indeed working. I haven't messed with the voltage, nor have I overclocked it
aside from the factory OC of the model.

So if the heatpipes are damaged somehow, I guess take advantage of EVGA's awesome warranty?

 

[Headfoot]

Could it be that you were running easier games that limited at 60 FPS (frame limiter, vsync) so the gpu was not fully pegged until recent harder to run games came out? Like your average utilization used to be less than 100% and now it is 100%.

If you go back and test a game you used to play when you had better temperatures, does it also exhibit increased temps?

 

[96Firebird]

Have you checked your case fans for dust build-up and to make sure they are working?

 

[mojothehut]

I usually run over 60fps, using a gsync @ 1440p. My case and fans are very clean.

I do notice when I set a frame cap (example 60fps) that my temps drop a little. So apparently its only when im pushing 75-120fps that my temps push 85c

 

[EXCellR8]

yes, the card will work harder as it tries to maintain the highest possible framerate. that's also a lot of pixels for it to render so I would cap the FR and perhaps disable AA (since you don't really need it with higher resolutions and set the fan @ 70%. it sounds like you've done all of the standard troubleshooting steps so if you're unhappy with the card and can't take advantage of the warranty I'd just get something else.

 

[Headfoot]

It sounds like the card is fine and you're just pushing it harder than you had before

 

[maddie]

Why would anyone assume that he's pushing it harder?

I see no CPU change or new game portfolio or any indication of more stressful settings. It appears to be a sudden increase in temps for the same workloads. TIM is replaced, case & ambients are unchanged, plus even with high fan speed (85%), temps remain higher. That is certainly not normal, especially when is was recently running cooler.

The problem is not the absolute temps so much, but the fact there was a ~8C rise, seemingly for no apparent reason.

 

[Headfoot]

Why would anyone assume that he's pushing it harder?

I see no CPU change or new game portfolio or any indication of more stressful settings. It appears to be a sudden increase in temps for the same workloads. TIM is replaced, case & ambients are unchanged, plus even with high fan speed (85%), temps remain higher. That is certainly not normal, especially when is was recently running cooler.

The problem is not the absolute temps so much, but the fact there was a ~8C rise, seemingly for no apparent reason.

"I usually run over 60fps, using a gsync @ 1440p. My case and fans are very clean.

I do notice when I set a frame cap (example 60fps) that my temps drop a little. So apparently its only when im pushing 75-120fps that my temps push 85c"

 

[Carfax83]

This card uses the ACX 2.0 right? That cooling solution is notably inferior to some of its contemporaries, but I don't know if it would hit in the 80c range without additional overclocking. Reviews showed it hitting mid to upper 70s at stock clocks.

So are you overclocking the card beyond the factory settings?

 

[mojothehut]

This card uses the ACX 2.0 right? That cooling solution is notably inferior to some of its contemporaries, but I don't know if it would hit in the 80c range without additional overclocking. Reviews showed it hitting mid to upper 70s at stock clocks.

So are you overclocking the card beyond the factory settings?

Not overclocking beyond the factory settings

 

[Rifter]

It could also be the temp sensor itself is starting to go bad.

 

[Snarf Snarf]

Have you tried frame capping at 120 fps and lowering the power limit? I had the same thing happen to me with Overwatch, I was getting temps over 80C when it never used to happen before with the in game frame limiter on. I took the power limit down to ~85% and now I'm still pulling the same 120fps as before but now with max temps around 65C.

 

[Face2Face]

I'd call or email EVGA support. Those guys are awesome and could help you better troubleshoot.

 

[Smoblikat]

What TIM did you use?

What program(s) are you using to monitor temps? I have a reference cooler 980Ti, and my temps dont get above 77 or 78 under furmark loads. My computer room is very cool though, CPU idles under 20C in the mornings and evenings.

 

[mojothehut]

What TIM did you use?

What program(s) are you using to monitor temps? I have a reference cooler 980Ti, and my temps dont get above 77 or 78 under furmark loads. My computer room is very cool though, CPU idles under 20C in the mornings and evenings.

I used Arctic Silver 5. I'm using GPUz to monitor mostly, sometimes Precision X. Both show the same temp.
My room isn't as cool, usually around 22-23C.
World of Warcraft, Fallout 4, GTA5 push the temp into the 82-85c range. Doom, Prey and Civ6 generally don't push above 79.

 

[Smoblikat]

I used Arctic Silver 5. I'm using GPUz to monitor mostly, sometimes Precision X. Both show the same temp.
My room isn't as cool, usually around 22-23C.
World of Warcraft, Fallout 4, GTA5 push the temp into the 82-85c range. Doom, Prey and Civ6 generally don't push above 79.

I dont think its the culprit here, but I think I remember reading that you shouldnt use AS5 on graphics cards because of the high metal content. I think ceramic paste is preferred, but maybe times have changed. Anyway have you tried setting a custom fan curve in MSI afterburner? Also make sure your voltage is normal and not spiking extremely high due to a stupid factory OC thing, you can use afterburner to change it if necessary.

 

[Headfoot]

Either this is normal behavior or there is a problem with the unit, either way RMA is probably your path forward.

 

[Guru]

Wasn't Maxwell operating temps, especially for the 980ti absolutely fine up till 102C?

85C does indeed seem a bit too much for a custom card, they all had better cooling than the reference design, but in summer hot days, more demanding games possibly, a bit of wear and tear on the fans and you could see that.

You also said you cleaned the thermal paste and applied a new one, which thermal paste did you apply, because in 99.9% of the cases applying a high quality paste is much better, it's like night and day difference, so applying some run of the mill paste might actually make things worse or have things the same.

Also the way you apply the thermal paste is very important, it needs to properly cover at least 99% of the GPU, the 1% percent being the very outward sides not being fully covered in like less than a millimeter.

Also you need to properly scrub the previous thermal paste and clean the surface with alcohol and cotton microfiber dry cloth.

 

[Tup3x]

I dont think its the culprit here, but I think I remember reading that you shouldnt use AS5 on graphics cards because of the high metal content. I think ceramic paste is preferred, but maybe times have changed. Anyway have you tried setting a custom fan curve in MSI afterburner? Also make sure your voltage is normal and not spiking extremely high due to a stupid factory OC thing, you can use afterburner to change it if necessary.

I wouldn't use AS5 on anything. It's subpar these days, messy and can be dangerous if you accidentally spill it all over the place.




My GTX 1070 run at <65c while fans stay <1400 RPM (overclocked but stock voltage). I did do the EVGA thermal mod and used Gelid GC-Extreme thermal paste.

 

[thejunglegod]

I had the same problems about a couple months back. With my Zotac 980Ti Amp Extreme edition reaching temps of more than 87°C and finally resulting in a pc restart. It used to do this irrespective of any game I used to play. For eg, it used to reach 85°C while playing Age of Mythology. I tried everything as well. Re-applied a new TIM (Noctua-the one that came with my NH-D15), checked if the fans were working, set custom fan profiles on MSI Afterburner etc.

What worked for me was realizing that my front 240mm intake fan on my CM 690III cabinet had stopped working which was my only intake fan from the front. The hot air vented out by the fans would probably stay in the graphic card area, since there was no way to exhaust this hot air.
I have kept my cabinet open since then and voila! No more high temps. It does tend to touch 70°C in Dishonored 2 but I think that is quite normal.

TLDR, my cabinet airflow was very very bad and opening the side wall of my cabinet helped in venting out the excess hot air which helped in stabilizing temps.
I have since installed a proper 240mm front intake fan, one top intake and one top exhaust, adding to the default back exhaust fan and the bottom intake fan.