[SOLVED] Gigabyte RTX 2080 TI Fans Ramping Up and Down

Status
Not open for further replies.
Hello,

Recently my Gigabyte Aorus RTX 2080 Ti Xtreme 11G graphics card has been experiencing an issue where the fans will ramp (while gaming) up to 4000+ RPM as the GPU temperature reaches 85c degrees and then as the GPU drops to 84c degrees, the fans instantly drop to ~2000 RPM. As such, the fans go into this endless cycle of up, down, up, down. With that an increase in extreme noise.

Seen here ( View: https://imgur.com/a/hQUkqio
) in the software, you can see what is happening along with the GPU's current settings for power. The fans are in this up, down cycle while the GPU temperature remains essentially constant and, as expected, so should the fan. I've attempted to clean reinstall the Aorus software in case it was a bug in the Auto fan curve, but the problem remains. Further troubleshooting included setting the fan speed to a Manual setting of 100% and creating a custom fan curve, but each time the GPU ramps the fans to 4000+ RPM the moment it hits 85c degrees. Regardless of the setting. The problem continues on stock settings as well.

The case is properly cooled in a Corsair Crystal 680X case with three LL120 intake fans in the front, two at the top acting as exhaust, one in the rear also acting as exhaust, and finally two Noctua fans positioned at the bottom; again, acting as exhaust. All are 120mm. The rear compartment is cooled by a Noctua slim 92mm fan. The graphics card sits vertically in the case.

My problem was first discovered a few weeks ago, but no major system changes were made then. In attempt to help resolve the issue and rule out the Aorus software having a bug, a full clean reinstall of Windows 10 was performed a week ago. The software has been reinstalled.

Other specifications of the system include a Gigabyte Aorus Gaming 5 motherboard, i5-8600K CPU, 16GB 2666MHz RAM, and a 750W Platinum SeaSonic PSU.

This problem just recently started to happen and I am not sure what caused it. Any assistance or thoughts on what to help here would be appreciated.
 
Solution
No, I did not. Do you believe it's worth a shot to swap them? They aren't easily swappable hence for not doing it tonight.
Yeah, do it when you have time. The reason being:
Horizontal gpu[pretend the bottom fans in the next image are intake]
View: https://imgur.com/VWAjFyi


That same image with those bottom exhausts still there:
View: https://imgur.com/MUtJNa7


Vertical gpu, bottom intake, air stagnates:
View: https://imgur.com/zSlMr5V


Vertical gpu, bottom exhaust:
View: https://imgur.com/fmm5u6u

Phaaze88

Titan
Ambassador
I wouldn't be so sure about that 'properly cooled' part when the gpu is running over 84C... that's the thermal throttle territory, if you didn't know...

It's probably related to why the fans are freaking out, but there have been threads about this model's fans going bananas before.
Does Aorus Engine tell you what bios version is currently being used?

The F3 bios located here contains a fix for the fan rpm issue, but the gpu's vbios needs to be compatible with it: https://www.gigabyte.com/us/Graphics-Card/GV-N208TAORUS-X-11GC/support#support-dl-bios
 
Hey, thanks for the response.

The GPU has been fine in this position in this case for months and I agree it is s hitting a thermal limit, but I am unsure of the root cause at this point. Perhaps I should try mounting the GPU in line with the fans. Thoughts on perhaps reapplying new thermal paste too?

I did notice the BIOS update for this card, however it is quite confusing. The Aorus engine states I am running BIOS F12 which is my motherboards BIOS. Regardless, I was going to flash to F3 anyways, but if you download that BIOS you get two versions with no understanding of which is the correct one. Thoughts there?
 

Phaaze88

Titan
Ambassador
Ok. You already have the F12 bios, so that F3 one on the product support page is useless...
Don't flash that F3, you'll likely end up bricking the gpu, and being left with a paperweight is way worse than what you're dealing with now!

The GPU has been fine in this position in this case for months and I agree it is s hitting a thermal limit, but I am unsure of the root cause at this point.
From what I've found, the case is marketed for high airflow, but does not deliver in the feature it's advertised for, and your hardware has to deal with those higher operating temps as a result.
View: https://www.youtube.com/watch?v=Ik48WmrJrkI



It's just going haywire around the thermal throttle limit, right? Then we need to try and get those gpu temps down.
The case is properly cooled in a Corsair Crystal 680X case with three LL120 intake fans in the front, two at the top acting as exhaust, one in the rear also acting as exhaust, and finally two Noctua fans positioned at the bottom; again, acting as exhaust. All are 120mm. The rear compartment is cooled by a Noctua slim 92mm fan. The graphics card sits vertically in the case.
It would be better to just do away with the vertical gpu setup entirely, but if you don't want to, the alternatives are:
1)The gpu vertical mount.
How close is it mounted to the glass window? Is it the horizontal, or vertical PCIE slots that are being used? The closer the gpu is to said window, the worse it's going to be.

2)"The rear compartment is cooled by a Noctua slim 92mm fan."
Where is this referring to? I don't see the space for this in the specs?

3)The LL120s do not move a ton of air, unless it's the 2200rpm version.
Stronger fans would help with the restriction caused by the glass panels.

4)Vertical gpu is tricky to do with the standard fan coolers.
Horizontal gpu draws cool air through the bottom and then dumps it's waste heat in all horizontal directions; traditional front to back, bottom to top airflow setup.

Vertical gpu draws air from the side and then dump it's waste heat in the compass directions. Heavy negative pressure airflow setup: front to back, front to top, front to bottom.
The chassis' design may not favor, or work against such a setup.
 
Ok. You already have the F12 bios, so that F3 one on the product support page is useless...
Don't flash that F3, you'll likely end up bricking the gpu, and being left with a paperweight is way worse than what you're dealing with now!

Yes, while that is correct to not flash to a mismatch BIOS, I am not convinced that my card is running a BIOS that Gigabyte themselves does not even offer on the website. I actually reached out to their support to inquire.

It's just going haywire around the thermal throttle limit, right? Then we need to try and get those gpu temps down.

Correct, it's just hitting a thermal limit, but it doesn't seem to be decreasing power, just a brief fan ramp. Hence debating opening the card to reapply thermal paste.

There's about two inches between the case's window and the GPU. Yesterday I decided to take a very light ribbon and do some air flow tests. It looks like air flow runs well over the back shield, but poorly onto the block of the card.

I'm going to swap the card to a horizontal position and give it a shot. I'll report back on my findings.
 
What did they say regarding the berserk fans?
I asked about that as well as the BIOS update as I truly believe the software reports incorrectly. In terms of the BIOS update in which one was the correct F3 version or how to even verify, they just told me that I don't need to do the BIOS update and moved into custom fan curves as a troubleshooting measure.

I explained the troubleshooting I've already taken and the response half an hour ago was "In that case it does not appear to be normal, does reseating the card and reloading the driver behaves the same as well?"

This is a fresh install of Windows which I will inform them of. I'm thinking they may think it's software related, nevertheless, I will follow up.

Case airflow test: run your games with the front and side panels off.
If there isn't too much resistance, you won't see a large gap in temps, such as less than 5C, for example.

Not a bad first test before I swap it to horizontal. The front "panel" is just a windowed door, so easy enough to test. :) I'll try that later this evening as well.
 

Phaaze88

Titan
Ambassador
Software, huh?
Perhaps Msi Afterburner is worth a shot? I remember using Aorus Engine for my 1080Ti Gaming OC, but gave up on it shortly.
At the time, all profiles - even custom - except for the silent one were broken. Games kept crashing - I switched to Afterburner - no issues, and didn't look back.
 
You, my friend, I think are on to to something. Of which, by the way, I appreciate your assistance throughout this. I assist on many other forums in the role you are now and will be joining here as well to assist shortly after this is fixed!

I opened the door and loaded up FurMark just to troubleshoot the air flow problem. To my surprise, the card leveled out at 80c degrees which I am used to expecting. The moment I closed the door, the card temperature began to spike.

You can almost tell where in the point I closed the door and reopened it:

View: https://imgur.com/a/iS3t7XD


Seeing that the problem is rather new, I'm going to try to resead the GPU in the vertical slot. Perhaps it's being twisted enough that is not noticeable to me. Either way, now I think we're looking at suffocation rather then any other issue.

To answer your question, I still believe Gigabyte believes it is a software issue. If I am honest, their software really isn't the best...

I know a lot of people use Afterburner, but I've never really experimented with it. Worth a shot in the future too.
 

Phaaze88

Titan
Ambassador
Sorry, stuff came up!

80C is much better!

Ok, you don't need filters on exhausts, so remove those if possible.

Bottom intake is more worthwhile with horizontal gpus.
With vertical gpus: ok, so the gpu is exhausting it's waste heat downwards, fighting against the bottom case intake... the air in the space between those devices stagnates, creating a hotspot.
That definitely won't help gpu temps.
 
Ok, you don't need filters on exhausts, so remove those if possible.

It's on carpet, so I filter them for that purpose. Even if they are blowing out. Unless you've seen improvement, I don't see it as a harm.

----

So I did some testing this evening. I moved the GPU to horizontal positioning and the thing overheated right away. Just as before. Now I don't know what to expect. Seems like the only time it's "moderately" okay is with the case door open. :mad:
 

Phaaze88

Titan
Ambassador
No, I did not. Do you believe it's worth a shot to swap them? They aren't easily swappable hence for not doing it tonight.
Yeah, do it when you have time. The reason being:
Horizontal gpu[pretend the bottom fans in the next image are intake]
View: https://imgur.com/VWAjFyi


That same image with those bottom exhausts still there:
View: https://imgur.com/MUtJNa7


Vertical gpu, bottom intake, air stagnates:
View: https://imgur.com/zSlMr5V


Vertical gpu, bottom exhaust:
View: https://imgur.com/fmm5u6u
 
Solution
Jan 29, 2021
1
1
15
Hi mate, had this problem. Set target temp to 0. power target to 100%
OMG I've been looking for a fix forever and this finally did it...

I bought a new pc case because I thought the airflow in my previous case the "nzxt h500i" was terrible which it is know for and thought that this is the issue,
But upon upgrading to a much better case the graphics card still did the same thing..

This finally solved it after looking for so long I finally find this comment here at the very bottom thank you so much :D
 
  • Like
Reactions: DMRTech
Status
Not open for further replies.