[SOLVED] 2nd GPU and Monitor Randomly not being recognized.

[Vezken]

Since yesterday I've been having an issue where when I boot up my computer from a powered off state, it will not always recognize that I have a 2nd GPU and monitor installed. When it boots only my main monitor will display and the device manager will only show one of my 2 GPUs . Also the nVidia control panel will be missing from the right click options on the display and geForce Experience will not be shown down in the taskbar. I'm honestly not sure if this is a software or hardware problem, or a BIOS issue.

The most recent change I made to my computer was I switched from running a single Asus PG348Q 34" ultrawide to now running a 27" Asus PG27UQ along side a Dell P2314H in a dual monitor configuration. I made the change on Monday and didn't have any issues until Wednesday and turned off the PC several times after changing the setup.

Yesterday when I woke up and powered on my computer I noticed the issue for the first time. I played with it for a bit and tried a few trouble shooting steps and eventually got both GPUs to display, but after powering off and on again the issue reappeared. I've seen the issue resolve itself both when I'm in Windows or when I'm in the BIOS. What I mean by this is I notice the lights on the NVLink SLI bridge will all of a sudden turn on, so its like its just all of a sudden has power or something.

Things I've tried so far:
1.) Clean uninstall of GPU drivers using DDU in safemode and then reinstalling the latest drivers with and without geforce.
2.) Updated Primary Display in BIOS to be PCIE as well as updating link speed to be GEN3
3.) purchased and installed a new NvLink bridge.
4.) wiggled parts, ensured both GPUs were seated well.

My computer specs are as follows:
i9-9900K overclocked to 5GHz
32GB G.Skill trident DDR4 RAM
2 x 2080ti nVidia reference cards
Asus Maximus XI Formula Mobo
Asus Thor 1200W psu

My computer also has a custom hard line water cooling loop I did, so simply moving a GPU to the other slot and trying that is a huge hassle involving draining the loop, cutting new tubes to fit, refilling the loop, then testing.

 

[Vezken]

After writing all that I started thinking, maybe this is a power issue? The fact that the issue has resolved itself both when I was troubleshooting within windows and when I was in the BIOS makes me lean that way. Once again by resolve itself I mean I can see the RGB on the NvLink bridge (old and new one I purchased) turn on randomly after some time. Why would it come on after the computer has been running for a bit and not right away though? should I plug the video cards power cables in a different spot on my powersupply and see if that resolves the issue or might it be something else more obvious to someone out there?

 

[Vezken]

Troubleshooting out loud here at this point, but I think it might be a power issue. Two things I've tried since my last post, both of which seemed to work:

1.) unplugged the PG27UG display port cable before booting.
2.) Turned of Fast start in windows power settings.

I turned off my computer for about 30 minutes after turning off fast boot and then booted again to see if the problem would come back and so far no. Yesterday when trying to fix this I turned off my computer for like 5 minutes and it was back instantly , so this is promising, but when I turn my computer off after a nights sleep will be the true test.

 

[Vezken]

Well turned back on my computer this morning and back to square one... The 2nd GPU isn't recognized, nVidia Control panel is missing, and 2nd display isn't showing up. What could be causing all this. I can tell it isn't going to work instantly because when I boot the NvLink SLI bridge doesn't light up with everything else, so it's like it doesn't have power or something.

 

[Vezken]

Another day of troubleshooting and another day where I "fixed" it (for now anyway). This morning I reached that point where you say, "screw it, time to format and reinstall everything and see when the problem starts". I started by attempting to update my BIOS which was already current, but figured I would re-flash and see what happens. Upon attempting to boot into my BIOS I couldn't. Each time I tried my keyboard wouldn't start working until I got to the windows login screen. I've had issues with keyboards in the past on this machine and usually just unplugging them and plugging them back in fixes it. that didn't work this time though. My keyboard kept being unusable until I the windows login. I plugged in a spare keyboard to see if that would work and had the same issue. I plugged in the spare keyboard into a different USB port and it worked and once again magically my NvLink bridge for my GPU's lit up. I tested again by plugging back in the original keyboard and it too worked from the different USB ports.

So now I'm thinking maybe I got some bad USB ports or potentially some sort of short happening with the two my keyboard was plugged into.

As this issue keeps coming back after a cold boot especially long ones such as after sleeping, I tested keeping my keyboard in different USB ports and turning off my computer for longer periods of time. I tested the following timespans:

30 Minutes, psu turned off and power cable removed = no issues
1 hour = no issues
2 hours = no issues

I want to say this is solved and it was something to do with some USB ports on my computer or potentially the keyboard itself, but I won't know for sure until I leave this off overnight again and check. I bought a different keyboard that doesn't need 2 ports and am trying that at the same time. I will report back with my progress just in case my rambling helps anyone in the future with a similar issue.

 

[Vezken]

Left the computer off overnight again and this morning everything worked! I'll be honest, not entirely sure what the problem ended up being, but here are the things I can't rule out as contributing to the fix . If anyone has a similar issue in the future, here is where I would start (in order of most likely):

1.) Confirm your USB ports aren't the problem. Unplug everything and plug your mouse and keyboard into different ports and see if the issue persists. Continue to boot by plugging in your mouse and keyboard into different slots. I noticed that the top two ports on the back of my computer weren't working as well as the others. It took longer for items to be recognized if anything was plugged into them.
2.) Uninstall and reinstall your displays one at a time and change which display your computer uses to boot.
3.) update your BIOS and remove any overclocks.

One of those there things did it for me. I'm assuming I had something going on with my USB ports.