Question RTX 2080Ti fans running at max speed for 5-10 second bursts at 75+ degrees

[NathanT98]

Hi guys,

I was just wondering if anyone can help me with the following issue please. For some reason when my GPU reaches around 75 degrees or above (when running games etc), the GPU fans will constantly have 5-10 second bursts of max speed, and then go back to their previous speed. This constantly happens every so often on the majority of games I play. It doesn't seem to be specific to a certain game, so I don't think it's a particular game causing the issue here. I've tried manually changing the fan curve in programs such as ASUS GPU Tweak II and MSI afterburner. I've also tried looking in the BIOS to see if there is anything in there I could change, however I can only see fan settings for my CPU in there. I've also tried reading numerous forum posts from people with similar issues, but so far I haven't really found any solutions. Also it might be worth noting that the motherboard I have is not even a year old yet, as I replaced it in December 2020, and even with the previous motherboard, the GPU still had the same issue. Also I've had my GPU for over 2 years now, just in case anyone needs to know.

I have had this problem for quite some time now so I'm really not sure what could be causing it, or how to resolve it. Any suggestions would be much appreciated.

 

[rgd1101]

like your previous thread. full system spec? include brand and model of the psu

 

[NathanT98]

like your previous thread. full system spec? include brand and model of the psu

Sorry I never got back to you on the previous one. Please see full system specs below:

OS: Windows 10 64 bit
CPU: AMD Ryzen 9 5950x
GPU: AORUS RTX 2080 Ti XTREME 11G
RAM: 16GB DDR4
Motherboard: ASUS ROG CROSSHAIR VIII HERO
PSU: Corsair RM850

 

[NathanT98]

Has anyone got any ideas regarding this please? I've exhausted all options trying to fix this problem.

 

[Phaaze88]

Probably letting it get too hot?

Or it's affected by that one fan rpm bug that afflicted a number of Gigabyte 20 series cards. You should check the support webpage for a compatible bios update with that fix patched in.

 

[NathanT98]

Probably letting it get too hot?

Or it's affected by that one fan rpm bug that afflicted a number of Gigabyte 20 series cards. You should check the support webpage for a compatible bios update with that fix patched in.

It usually gets to around 79 degrees max, is that too hot or around about what it should be at full load?

Also do you have a website I can check for that please?

 

[Phaaze88]

It usually gets to around 79 degrees max, is that too hot or around about what it should be at full load?

Gpu core temperature. Good, up until around 82-83C. Critical at 88C. Cooler is always better, because of Gpu Boost 3.0.
Gpu hotspot temperature. Critical at 110C.
Gpu memory temperature. Good up until around 95C. Critical at 105C.

Also do you have a website I can check for that please?

https://www.gigabyte.com/Graphics-Card/GeForce-RTX™-20-Series
You click on the card that's yours, click on the support tab, and then the Bios tab. They should be listed there.
Please note:
You can only update to a VBIOS version of the same series. Visual C++2008 is required for this BIOS flash. Can be checked via Windows Search - Programs.
If your VBIOS version is:
F1, it can only be updated with VBIOS versions F2-F9.
F10, it can only be updated with VBIOS versions F11-F19.
F20, it can only be updated with VBIOS versions F21-F29.

You can see what your card's current vbios is via Gpu-Z > Advanced tab > drop-down box, Nvidia Bios > Message.
For example: GV-N108TGAMING OC-11GD/F1/0543

How to flash vbios: https://www.gigabyte.com/WebPage/40/index.htm

 

[FlintIronStagg]

Get MSI afterburner.
set a fan curve with this method:
- 5% more for every 5 degrees over 45.

45C = 50% fan speed
50C = 55% fan speed.
follow this trend all the way to75C and 80% fan speed

part of the issue with stock fan settings is they try to keep your card whisper quiet until its too late then ramp up to cool it down when it hits critical.

if you are having issues with the temps and fan % after this change, you will need to look into your tower airflow; you simply may not be getting enough.

you also need to assure that you are "applying overclocking at startup" to afterburner.

the clock fan curve is technically recognized as a card overclock. if you are changing the curve yet not setting this it may be why your card isnt respondong.




cheers!

 

[SilverArrows]

YES! I have the same card as you too, Gigabytes RTX2080Ti Gaming OC. Once it hit 75 degree, the fans start to kick to max rpm for like 5 seconds then goes back. It does this all the time when hit 75 degree. I already try setting a custom fan curve, but still the same.

 

[NathanT98]

YES! I have the same card as you too, Gigabytes RTX2080Ti Gaming OC. Once it hit 75 degree, the fans start to kick to max rpm for like 5 seconds then goes back. It does this all the time when hit 75 degree. I already try setting a custom fan curve, but still the same.

Yeah it's extremely frustrating. I have just tried changing my VBios version, however unfortunately it still appears to be having the same issue. Starting to think my only option would be to replace my GPU now, which I didn't really want to do.

 

[NathanT98]

Does anyone else have any possible solutions to this problem? Nothing I've tried so far has solved this unfortunately.

 

[Phaaze88]

Vbios update didn't work, huh?
You've only posted about gpu core... you still never said how high gpu hotspot gets, nor Vram.

Beyond this, you'll need to figure out how to get thermals down so it doesn't do that, or RMA the card if it's still under warranty.

 

[NathanT98]

Vbios update didn't work, huh?
You've only posted about gpu core... you still never said how high gpu hotspot gets, nor Vram.

Beyond this, you'll need to figure out how to get thermals down so it doesn't do that, or RMA the card if it's still under warranty.

When the fans spike at around 78-79 degrees, the hotspot temperature was at around 106 degrees and the VRam temperature was at around 80 degrees.

 

[Phaaze88]

When the fans spike at around 78-79 degrees, the hotspot temperature was at around 106 degrees and the VRam temperature was at around 80 degrees.

Too high. The other 2 are fine.
Looks like the gpu die is due for a repaste.

 

[NathanT98]

Too high. The other 2 are fine.
Looks like the gpu die is due for a repaste.

So it's the GPU temp that is getting too high? Surely the fans shouldn't be spiking like that though? Regardless of what I set in MSI afterburner it makes no difference to the fans once it reaches that temperature. I thought 78-79 wasn't ideal but wasn't a major issue, then anything after 80 is starts to get concerning?

 

[Phaaze88]

Look at the part I highlighted in red - ok, maybe it's your browser or whatever you're seeing this with, but I highlighted this part:
"the hotspot temperature was at around 106 degrees"
The limit is 110C for hotspot. Thermal paste either needs to be reapplied, or the cooler isn't sitting as flush against the die as it used to - usually caused by significant gpu sag.

 

[NathanT98]

Look at the part I highlighted in red - ok, maybe it's your browser or whatever you're seeing this with, but I highlighted this part:
"the hotspot temperature was at around 106 degrees"
The limit is 110C for hotspot. Thermal paste either needs to be reapplied, or the cooler isn't sitting as flush against the die as it used to - usually caused by significant gpu sag.

I see. I'll try checking the thermal paste then and reapplying if needed.

 

[Phaaze88]

Make sure you use something a bit thicker here. Since the gpu is a naked die, it doesn't share the same preferences as a cpu's IHS.
NT-H2(currently using that)
Gelid GC Extreme
Prolimatech PK-3
Kingpin KPx
Cooler Master Mastergel Maker Nano
These are a few pastes that work well here. Refrain from using the widely acclaimed TG Kryonaut - it doesn't do so great on naked die applications.

 

[NathanT98]

Make sure you use something a bit thicker here. Since the gpu is a naked die, it doesn't share the same preferences as a cpu's IHS.
NT-H2(currently using that)
Gelid GC Extreme
Prolimatech PK-3
Kingpin KPx
Cooler Master Mastergel Maker Nano
These are a few pastes that work well here. Refrain from using the widely acclaimed TG Kryonaut - it doesn't do so great on naked die applications.

Went with the NT-H2 anyway as I saw it was good. I'll let you know the results once I get chance to re-paste it. Fingers crossed it fixes the fan issue which I'm guessing it would if the re-paste fixes the high temperatures.

 

[Phaaze88]

Strange that NT-H1 still has such a large presence compared to NT-H2, and the latter's been out a couple years at least.
Maybe Noctua didn't invest much in marketing it

 

[NathanT98]

Strange that NT-H1 still has such a large presence compared to NT-H2, and the latter's been out a couple years at least.
Maybe Noctua didn't invest much in marketing it

Yeah I realized after you said and managed to cancel the NT-H1 order and get NT-H2 instead

 

[Phaaze88]

Great!

 

[NathanT98]

Great!

Thanks for the support. Fingers crossed this sorts my issue once I apply the new paste. Just watched a video and after the guy re-pasted his GPU die he was seeing reductions of 20+ degrees at times.

 

[SilverArrows]

When the fans spike at around 78-79 degrees, the hotspot temperature was at around 106 degrees and the VRam temperature was at around 80 degrees.

Same like mine. My Hotspot goes around 108 degree.

 

[SilverArrows]

Thanks for the support. Fingers crossed this sorts my issue once I apply the new paste. Just watched a video and after the guy re-pasted his GPU die he was seeing reductions of 20+ degrees at times.

Keep us updated man. I'm also planning to repaste and replacing the thermal pad too once my dissertation done. Goodluck!

 

[NathanT98]

Keep us updated man. I'm also planning to repaste and replacing the thermal pad too once my dissertation done. Goodluck!

Yeah I'll let you know if it works. I'm hoping it sorts it as I'm out of options if it doesn't lmao.