Frequent random signal loss, having to force turn off. RTX 3080 at fault?

[pizza_time]

My monitors frequently started turning black with the signal lost pop up. They then continue to stay in that state until I force the pc to turn off and reboot. I bought my Zotac trinity oc rtx 3080 10gb lhr in August 2021. Other specs are an i9 10900k, trident z 2x16gb ddr4 3200mhz, a z490 aorus elite ac, a few hdds and 2 m.2 ssds. I’m running a 850W be quiet! PSU. Nothing pops up windows event viewer, except for the usual errors you get when forcefully turning the pc off. Also there is no pattern when this happens. Just that it happens quite often right after booting up the pc for the day, just a few minutes after each boot then. And then out of nothing it just works perfectly fine without issues for hours upon hours. I have both my gpu and CPU slightly undervolted, it ran perfect like this for over a year. I had a few blue screens here and there back then from assumingly some kind of ram issue or something, this has not occurred in many months though. Is there anything known about this? Is this yet another 30 series flaw? I don’t think I can get a replacement card after this amount of time either. Also I have noticed my gpu clock speeds going from idle to max back and forth somewhat frequently, which might just be due to something in the background needing a bit of gpu power once in a while.

If I’ve missed any useful information please let me know..
This is really annoying as I just have to reboot my pc an unknownly high amount of times every day before i can actually hop into games.

It would already be very helpful to know which part of my pc could cause such an issue. I assume the gpu but i could also imagine the ram or something.

[pizza_time]

*I have done a clean DDU driver installation a few months ago due to my idle clock speeds never going below 1725.

Also I’m on Windows 11. And on my 2nd boot it it the same thing.

[jmarket]

Are you running a liquid cooler or water block or just an air cooler? What are your temps?

[pizza_time]

Everything is air cooled. At this exact moment my pc ran without issues for 4-5 hours again.

GPU sits at a very max of 75°C but I am using a very slightly lower fan curve.
CPU is tough to be cooled with air as it’s a 10 core i9, cooled by my noctua nh u12s and while gaming it’s around 75-85°C depending on the game.

[jmarket]

Let’s try ruling out faulty hardware.

i9 K CPUs have built-in graphics. Have you tried plugging your monitors directly into onboard video and trying a game (yes FPS and graphics will not be good but this is ruling out a faulty video card)

Have you ran stress tests to see where it’s failing? Cinebench is a great stress test.

You have fan curves set. What are the temps inside the chassis?

[pizza_time]

1. No I have not done that yet. I haven’t done excessive testing myself yet also because I don’t have much time with my pc. I planned on trying my old gtx 960 and see if it still occurs but I guess I can also just run the monitors through the cpu, that’s true.
2. The thing is, it mainly happens when there is no game open, only my desktop and maybe youtube. The only game it has ever happened in was f1 22. No other game was involved yet.. Which is also very weird. I just had a game open in the background while watching YouTube, the pc ran fine for 9 hours now. And after I have closed the game but continued watching YouTube I had the signal loss once again after a bit of time. Maybe it is related to the gpu trying to idle?
3. I have rechecked and I actually don’t have custom fan curves on. That’s only for the summer where I’m using higher and there for louder curves. My system has quite bad airflow though, hence why temps are a bit higher than usual but nothing harmful.

[pizza_time]

Addition to 2.

I have done the DDU reinstall a few months ago because my gpu never went to idle speeds but rather always stayed at 1725.
And now the signal loss seems to occur when it wants to idle, but that’s just an assumption. And it also doesn’t explain why it occurs while f1 22 is running. But thst game is a crazy mess itself with everything.

[pizza_time]

Quick note.
I have just read through a similar thread on this forum and one answer was to make sure to not update drivers unless it’s necessary. I have once tried to fix something audio related, but I don’t exactly remember what it was. And for that I’ve used a driver updater and just updated everything there was.. Could thst have to do with anything regarding such an issue? But keep in mind that this update was actually much before this happened for the first time.

[pizza_time]

Another addition:
Now it also happens in other games. And every few hours or so. Which is insanely annoying as I get banned for leaving matches in certain games..
GPU fans went to the max after losing monitor connection twice today. I have changed my main monitor from one display port slot to another. I have completely removed my cpu undervolt. I have updated my nvidia drivers. I have tried to revert everything in device manager to its original driver and then made it automatically update. I actually don’t know what else I could do. The next best thing would be to switch to my old gtx 960 and see if it’s still happening.. But even if it is fixed then, replacing an rtx 3080 is a pretty expensive thing to do…

[jmarket]

Thank you for the update.

If replacing hardware doesn’t fix it, it points to a software issue, most likely Windows glitch

[pizza_time]

Well, replacing other hardware might fix it. Just the gpu isn’t the source. I currently run on only 1 ram stick and haven’t had a black screen since, but i might just be lucky as i sometimes was with 2.

[pizza_time]

Also it cannot be a windows glitch, as others reinstalled it and even formatted their whole pc. If anything it has something to do with the bios or something, but that’s also pretty unlikely i’d say

[pizza_time]

Update:
I now haven’t had this happen anymore since i took out 1 ram stick and put it back in a few hours after it not happening.
Also I have put power management to max. performance and selected my 3080 to be the graphics thing for everything, which led to the mhz staying at stock 1725 permanently. With that temps stayed the same permanently and there was no fluctuation between idle and stock mhz anymore. When in a game it goes up to 2025 and stays there too, temps are much more stable now either. I don’t know if just re-plugging the ram helped or actually having the gpu running at stock permanently. Either way it may take a bit more electricity now and stay a tiny bit warmer, but that’s fine if it doesn’t happen anymore. Will update on this next weekend as I don’t have my pc for the week. Possibly it’s fixed now, but maybe it’s not and i’m just lucky.

[pizza_time]

This is now fixed. When it was fixed i did 2 things, replugged one of my ram sticks and what stated above with power management settings and gpu selection.

My idle speed is the stock 1725mhz and my load speed is 2025mhz (or whatever else i’d have set it to).
If anybody else has this, you might wanna try those nvidia control panel settings.