[SOLVED] Extremely high fan RPM/speed on Gigabyte RTX 2080

[Stwemar]

Ok, so for a while i have had this RTX 2080(GV-N2080WF3-8GC ), and it always made this unbareable fan noise when under load, but not just any load, on some games it would happen, and on others it won't, from what i have experinced so far, it would be at normal fan speed, and then suddenly it would become a hoover, so i decided to change the thermal paste, after i did that, still happens, update the vbios, still the same issue, so i decided to document this behaviour as best i can, from what i can tell, if the clock speed is 1200mhz or lower, everything is fine, custom fan curves in afterburner works like intended, but as soon as it goes higher(AKA it OCes itself automaticallyi think) the fans go up to ~4500 RPM, no build up, nothing, the temperature is more than fine at 60-70C, and this still occurs, my idle temps look more than fine and normal to me, if you haven't followed me so far, let me show you with some screenshots.

IDLE:

https://imgur.com/y6SIXE0

View: https://i.imgur.com/y6SIXE0.png

Load where the fans don't act crazy:

https://imgur.com/duLz96W

View: https://i.imgur.com/duLz96W.png

Load where i basically can't hear myself talking :

https://imgur.com/MHs4VFG

View: https://i.imgur.com/MHs4VFG.png


Look at the huge difference in RPM, and i am clueless to why this happens, i am guessing that on some games the GPU does not go crazy, becasue the game is not demanding and it's meeting my cap of 144 FPS in nvidia control panel, i don't think its a heat issue, my CPU runs more than fine with a stock AMD wrath cooler in the same chassis.

also if you like at the last screenshot at the fan RPM, you can look at the red cutting off every second or so, which seems to me that it's trying to lower the temp, but the temp seems to be fine at that specific point in mind

do you guys have any idea?

Also here is some info:
BIOS: 90.04.23.00.96
Nvidia Drivers: 516.94

 

[Phaaze88]

Temperature is NOT fine. Your 3rd screenshot shows a hot spot of 107.6C. That's a temperature gap of 34C between gpu core and the hottest sensor on the die - that's really high.
Normal is in the range of 10-20C between those sensors. You're due for a repaste.


The crazy fan rpm is a bios fault with Gigabyte 20 series gpus. I recall there being quite a few, 'Gigabyte RTX 20 gpu, really loud fans' cases here and there.
The solution there is to try and update the card's vbios. Your card is currently on F11. The support page HERE has F12 available.
There's instructions on how to flash further down on the 'How to reflash bios?' button.

 

[Stwemar]

@Phaaze88 i have just repasted today, before the repaste it acted the same, after the repaste it acted the same, also i am confused, you say i should get F12 but the gigabyte website says that bios is for samsung memory version, i have a micron version and i have installed the F2 version, like the screenshot below

https://imgur.com/GSP1SzU

View: https://i.imgur.com/GSP1SzU.png

can you clear this confusion please?

 

[Phaaze88]

I looked up your bios version from TPU's archive:

TechPowerUp

Extensive repository of graphics card BIOS image files. Our database covers submissions categorized by GPU vendor, type, and board partner variant.

www.techpowerup.com

I ctrl+F 90.04.23.00.96 , and got 1 match:
https://www.techpowerup.com/vgabios/207869/gigabyte-rtx2080-8192-181023
Observing Bios Internals, I see:
GV-N2080WF3-8GC/F11/0810
Memory Support(GDDR6): Samsung, Micron, Hynix

Then I click on Find Compatible Bios, and get three results:
https://www.techpowerup.com/vgabios/?did=10DE-1E82-1458-37C1
The first is F12, the 2nd is F11 that your card was running, and the last one is F10. Both the F12 and F10 files list Samsung, Micron, and Hynix memory support.


So, if this is your card: RTX 2080(GV-N2080WF3-8GC )
And the bios it was on was F11, why did you step down to F2?

 

[Stwemar]

@Phaaze88 ok i flashed to F12, restarted and the issue still persists, hotspot is still 107C, do i repaste again, and this time do i go more generous?
, also could this be related to thermal pads on the VRM or chokes or memory?, or is it just the die?

 

[Phaaze88]

Repaste again, but manually spread the paste over the die. You can use an expired ID card or one of those disposable vinyl gloves to do it.

could this be related to thermal pads on the VRM or chokes or memory?, or is it just the die?

Did you or someone else replace the thermal pads at one point? If no, then it's not that, and just the die.
If yes, then it's common for the wrong hardness to be used and screw with cooler mounting pressure.

 

[Stwemar]

@Phaaze88 i did spread the paste on the die, made sure it's all covered, no metal showing, but one of the pads on one of the VRMs slipped a little, like it was wrinkled aswell, don't know if it was like that before, or when i pulled the gpu apart it slipped, but its only of the VRMs, do i remove it completely?, would that be the reason the cooler is not contacting the die correctly?, or the hotspot temp to be precise

 

[Phaaze88]

Can you take a picture of this? You have to use an image hosting site to post pictures here: https://forums.tomshardware.com/threads/image-posting-via-imgur.3773266/

would that be the reason the cooler is not contacting the die correctly?, or the hotspot temp to be precise

Refer back to my previous post with: Did you or someone else replace the thermal pads at one point?

 

[Stwemar]

@Phaaze88 ok so i repasted, and now the hotspot issue is fixed, my fans don't go crazy under extreme load anymore, which is great, here is some pictures i took on my phone
Before:

https://imgur.com/zFGeAlX

View: https://i.imgur.com/zFGeAlX.jpg

After:

https://imgur.com/7aWgIUh

View: https://i.imgur.com/7aWgIUh.jpg

Although the second picture don't show, i spread the thermal paste on the die, and on top of that i put an extra drop on top of it
after i put everything together my hotspot temps are great, they are around the same as my GPU temp.

Thank you for helping me out @Phaaze88

One other thing i am getting this noise at high load:

https://imgur.com/KaIDv71

View: https://i.imgur.com/KaIDv71.mp4
, i doubled checked my screws and that the fans had nothing in their way, on idle there is no sound, i don't know if that's how they supposed to sound normally, since all i heard were the hoover fans xD, the one way i could describe it is it sounds like a sizzling noise, very high pitch

EDIT: It doesn't sound like it's the fans, i made them go very high RPM manually, and the sound isn't there, it's when the GPU is under load, no matter what the RPM is at, it comes up

 

[Phaaze88]

So the pad's just peeled and not ripped. Memory temperature is still good, yes?

Since the fans aren't rubbing against anything, that leaves electrical noise from the card's VRM. Motherboards do it too, though it's easier to deal with by turning off a feature like Spread Spectrum in bios.
Gpu... not as easy to deal with; you may be stuck with it - it is NOT harmful, just a possible nuisance depending on how sensitive your ears are. You can try:
-lowering the gpu's power limit in Afterburner. Like 85-90% - don't want to go too low with it.
-setting fps caps, if the titles played don't already have caps built in.
-RMA the gpu for coil whine if it's still under warranty.

 

[Stwemar]

@Phaaze88 got it ty