Wanting to check here as a last resort before potentially RMAing my card. I already went through the official support process and an RMA was approved but I'm still not sure if that's actually necessary or if there's something I can do to fix this problem at home.
 
 
**I should note that this problem occurs independent of any software, even if I completely uninstall Precision X1 or XOC it still happens**
 
 
For a few months now, my FTW3 has been randomly shutting off 2/3 of the fans, and the GPU fan ramps up to 4000+ RPM. It obviously ignores the fan curves that are set when this happens. The weird thing is that RPM is beyond what the fans would run at even if you were to set them at 100% in software like Precision X1. If I manually set fans to run at 100% the highest I ever see them go is like 3300 RPM.
 
If I set the fan speeds manually, the other 2 fans turn back on and run at whatever speed I set them at, but the GPU fan still ramps up past 4k RPM despite the fan percentage showing whatever I set it at. For example, if I set the speed of all 3 fans to 50%, the other 2 fans will run at like 1900 RPM, but the GPU fan will still be at 4100 or something crazy. The noise when it's at this level is insane.
 
As soon as I switch fan control back to automatic where it's supposed to follow the fan curve (I have used the pre-set "aggressive" for 2 years), the other two fans shut off and the GPU fan continues running at 4k+ RPM. Below is what that looks like in the hardware monitor (this screenshot is from Precision X before I switched to X1, but it has the issue using X1 as well)
 

 
The only way I'm able to "resolve" this issue when it starts happening is through some random combination of restarting the PC, fiddling with the fan controls in X1, uninstalling X1 and reinstalling it, etc. I've never nailed down exactly what "fixes" it, it feels like it just sort of stops happening on its own rather than as a result of something I did. And then at some point it just starts back up again, could be later the same day, could be in a few days, could be in a few weeks. But it comes back. When it comes back it's not as a result of any changes I've made within X1. I don't even touch it anymore until the problem starts again.
 
 
So now as I'm posting this, it's been a few weeks since I had the RMA approved, but the problem was gone since the day I made the support claim so I thought maybe I had actually solved it for good this time. But then the problem started back up about 2 hours ago and I spent an hour fiddling with it to get it under control again, so IDK what the deal is.
 
What strikes me as odd is that even with everything functioning as "normal", what I see in X1 doesn't make a lot of sense. It should be noted that what I see in X1 and what I'm going to post in this screenshot is the same as what I see when the problem occurs, the only difference is the fan speeds on the left are showing the fans running at normal speeds as they're supposed to. When the problem occurs it shows fan1 at 4k+ RPM and the other fans at 0. But what stands out as weird to me is the fan speeds on the right. No matter what I do it always shows Fan1 at "18", and the other two at 0. Is this supposed to reflect the actual fan %? I remember it used to when I used precision XOC, those values would change dynamically with the RPM.
 

 
 
So to me all of these oddities combined signal some type of miscommunication in the fan controller of the card. I'm hoping that's firmware/software related rather than a problem with the hardware itself, and that an RMA won't be necessary.
 
I was able to find one thread where a person was having a similar fan control issue with their 1080Ti FTW3 and someone linked to an MCU update page, but that link now just goes to the MCUs for the 2080Ti FTW3. It seems they have replaced the page for the 1080Ti with the 2080Ti which won't help me. Is there anywhere I can find a tool to flash the MCU for my card? Maybe that would help?
 
I'm not sure where else to go with this. I don't want to RMA the card if I can help it, being without a GPU for however long it takes them to send it back is not going to be fun. But at the same time if whatever problem I'm experiencing is indicative of a greater problem that may continue to get worse, doing it sooner rather than later is probably better?