1080 Ti keeps falling off the bus

Hello,

somehow, since NVidia announced they don’t support the 1080 Ti anymore, my 1080 Ti keeps falling off the bus sporadically. There were around 3 occasions, where this happened multiple times per day, then stopped and everything worked again.

I am now in the 4th occasion of this and cannot use my GPU at all anymore, starting any game (even 2D games) makes the GPU fall off the bus after a few minutes tops. I even had multiple crashes just browsing the internet or watching YouTube videos.

I’ve also tried solving this with NVidia support, but after nothing worked, I was sent here to report my issue.

My specs are as follows:
OS: Garuda Linux (arch)
CPU: AMD Ryzen 9 5900X 12-Core Processor
RAM: 2x 16 GiB DDR4 at 2133 MT/s (G.Skill Tridentz RGB)
Motherboard: MAG X570 TOMAHAWK WIFI (MS-7C84)
Power supply: be quiet! Dark Power Pro 750W

I updated my BIOS back in february, where the 3rd occasion occurred, I also changed some BIOS settings (concerning PCIe) like the support suggested, tried changing the resolution and Hz, nothing has worked thus far to restore function.

I’ve got some logs I’m gonna put in the attachments here, but unfortunately I have not found a way to run nvidia-bug-report.sh after the GPU has fallen off the bus, the system seems to not react at all anymore. I even created a hotkey to run the script as sudo, enter my password and let the script run, I practiced running this multiple times while the GPU was still on the bus to be able to do it blindly the next time it falls off the bus. I also tried this multiple times and unfortunately not once did I get a file back from the script. Changing to another terminal via ctrl+alt+F1-F6 is completely impossible as well.

I still put 2 of those reports in the attachments in case they might help.

I’m completely at my wits end here, I even opened my computer multiple times to try replugging the GPU and its power connectors, but that didn’t work as well - especially since everything was plugged in nicely to begin with.

If you’ve got any ideas what I could still try, or if you still need any more information which might help solving this, please do respond, I’m sincerely desperate.

Off the bus_2026-07-20.zip (6.7 MB)

Off the bus_2026-07-19.zip (6.1 MB)

Off the bus_2026-07-18.zip (13.1 MB)

In the words of an NV eng:

Since you’ve eliminated BIOS/firmware with a reasonable degree of confidence, I’d look into the thermals and the PSU: it’s possible that you need to replace the thermal paste or some fan or that your PSU is aging (although with BeQuiet! it’s less likely).

I just recorded my CPU, RAM and GPU while loading into Cities: Skylines 2 to check for potential thermal reasons and I think that might be it. I uploaded the video here and to me it looks like it might actually be the temps. I recorded until the GPU fell off the bus again (which corrupted the video, but .mkv made it still watchable) and it looks like a) the GPU fan does nothing at all (and I don’t know why) and b) the GPU gets up to 95°C shortly before crashing.

Looking at other specs, it looks to me like it’s rather not the PSU, since the highest wattage I saw was 99W and that was somewhere in the middle of the video, but that’s of course no guarantee.

So my question now is: How likely do you think it is that the temps are the problem? Also: How do I get my GPU fan to do something again?

journal log of the crash in the attachments ofc

crash-cs2.log (214.0 KB)

I’m 95% sure that 95 degrees and the idle fan is the reason ;-]

The fan is either physically broken or some driver update screwed the firmware on your model: try maybe some older driver version to see if it makes it move, if not, replace it.

Also, don’t try to reproduce the issue unless really really necessary to avoid physical damage to the GPU chip and the VRAM.

I just tried downgrading my driver to 510, since I’m 100% positive that with this driver version my fans worked, but my distro (Garuda) seems to not really be compatible with downgrading the GPU driver unfortunately - but since setting the fan speed in sudo nvidia-settings does nothing as well, I’m pretty positive the physical fan is broken.

Now I know how to fix this though, so thank you very much for your help! :)

Edit: In a last ditch effort I opened my PC while running, touched the fan that wasn’t spinning and through my healing hands or something the fan started spinning again - I just tested with Cities: Skylines 2: The temps stay where they’re supposed to be now, the fan adjusts to the temps like it’s supposed to and the GPU stays on the BUS.

Of course this problem will reoccur, but at least I can now play some games until I got everything I need to fix the problem. :D

Hey, you know, I have arthritis is some joints… ;-)