Question Concerned my GPU fans are spinning too much at lower temps. Really loud!

Jul 15, 2022
Hey all! So I recently upgraded my PC's CPU, AIO, Motherboard, and RAM. Everything else listed in the specs below were already part of my previous rig. I've been monitoring performance recently and noticed that my GPU's fans are spinning at nearly 3000 RPM when the GPU is at 60 C under load. I most recently noticed this happening last night while playing Subnautica on High settings (I've seen it a little on other games too).

Average temp under load is roughly 63 C, which is pretty damn good but it sounds like a jet engine taking off. While this is happening, my case fans are still at 1000 RPM. Thing is, I don't recall this happening on my previous setup, which had the same GPU and fans. I've confirmed that my case fans are configured correctly and I've adjusted the fan curves on all of them too. I'm also undervolting my card.

I just want to make sure that I've set up this PC correctly and that this is normal behavior with this card. It just doesn't seem correct that the fans spin up to 3000 RPM. Any help or education would be super helpful. Thanks!

Current specs are:
  • Intel i7-12700K
  • Asus ROG STRIX Z690-E Motherboard
  • RTX 3080 Ti Founders Edition
  • 32 GB of RAM
  • NZXT Kraken z73 360mm AIO
  • Lian Li O11 Dynamic Evo Case
    • Intake
      • 3 Lian Li AL 120mm Unifans on the side of the case
      • 3 Lian Li AL 120mm Unifans on the bottom of the case
    • Exhaust
      • 3 Lian Li AL 120mm Unifans on the top of the case on the radiator
      • 1 Lian Li AL 120mm Unifans on the rear of the case near the I/O


I've been monitoring performance recently and noticed that my GPU's fans are spinning at nearly 3000 RPM when the GPU is at 60 C under load.
Check Task Manager process tab. Something may be using the gpu a fair bit.
Also, while TM is open, keep an ear open for if the fan winds down; some mining malware knows to go into hiding when the user checks TM.

Check gpu hot spot and memory junction. You can use HWINFO or Gpu-Z to see them.


You didn't hear if the fan ramped down at all while you were in TM, did you?

Wow. That memory is running bloody toasty under load... What's it look like when game isn't running?
Gpu core and hot spot are all well and good though.

I think you're going to have to customize your own fan curve. My guess so far is the fans on the 30 series Founders are kinda noisy.
Auto fan curves have almost always been rather poor.


Ok. It's fine there.
Conclusion is that the auto fan curve isn't great. Create your own via afterburner... or increase case fans' speed.

And just the answer this, no I didn't hear any change.

Should I be concerned? I'm open to ideas or suggestions on how to fix.
If it's bouncing up there and back down - not a big deal. If it's sitting up there though, that's a problem if you're intending to keep it more than a few years.
Solutions are:
-Alphacool Eiswolf 2:

-Replace the thermal pads, which can be an endeavor in itself, if others haven't already found pads that work.
Took a little time looking around, and finding that there was success with Gelid Extreme 2mm on the die side and 3mm on the backplate side. Here's one: (works for Ti FE too)
This all also means that you'll have to repaste the die too. If you have some MX-4 or NT-H2 laying around, those should work - some pastes do not work well on bare die because they're too 'runny'.


Depends on:
-your stance on money.
-the gpu AIO is only compatible for that gpu; you won't be able to install it on later models.
-when the AIO breaks, you may or may not be ready to buy a new gpu. If you're not ready, you'll have to put the original cooler back on... meaning you'd have to mess with new pads anyway.
  • Like
Reactions: KyaraM