Ok long one so bare with me
Issue - Launching game (tried multiple) results in monitor loosing signal
To restore - only a restart resumes display
Notes -
Computer does not crash (it has only once)

Bugcheck code: 0x3B(0xC0000005, 0xFFFFF8055BADB836, 0xFFFFBB068FA1E820, 0x0)Bugcheck name:SYSTEM_SERVICE_EXCEPTIONDriver or module in which error occurred: nvlddmkm.sys (nvlddmkm+0xDCB836)File path:C:\Windows\System32\DriverStore\FileRepository\nv_dispi.inf_amd64_647b4244e991951b\nvlddmkm.sysDescription: NVIDIA Windows Kernel Mode Driver, Version 512.95

Pressing the power button and waiting will shut down properly as will connecting via teamviewer on my phone and selecting restart.
Sometimes if I restart instead of shut down the BIOS will stop with error code d6 (no console output devices) and once I had an on screen error saying no vbios found.
 
Now I have tried various things.
1. The gpu's other BIOS, no change
2. DDU clean reinstall, no change
3. Re-seating the GPU, no change
3. Different RAM (just cause I had some), no change
4. Updated to newest MB BIOS, no change
 
This issue came about over the place of a couple of days and started intermittantly before now being consistant.
The last test I ran was with Port Royal, this resulted in once I teamviewed in I was able to see my desktop albeit in 640x480.
Accessing device manager showed GPU disconnected (will grab a screenshot after I post this)
 
Specs
Intel 12900KF
G.Skill DDR5@6000
MSI Z690 Carbon (newest BIOS)
EVGA 3080Ti FTW
Thermaltake ToughPower GF1 1200W
GPU & CPU watercooled on separate loops See thread > for more https://forums.evga.com/My-new-labour-of-love-m3508535.aspx
 
Any tips or ideas?
Is this GPU/MB/CPU

What happens when you set your gpu core & mem to max negative values? If that doesn’t help try setting your ram to 4800.

@ Sajin Just tested first lowered the two sliders to min (-502) - Ran port royal, same result
Then restarted via teamviewer and it hung at A2 (IDE Detect) Weird, anyway hit the reset button and back into the bios to disable XMP, rebooted and then lowered the sliders again also, same result.
 
@aka_Steve_b I did update the Vbios a while back and as one of my BIOS checks I reinstalled Precision which was all ticks (Current Vbios 94.02.71.80.86)
 
Thanks for the input ;)
 
I've still got my old PC that the wife now uses for general browsing and some work, running on the iGPU so will have to try and move the GPU into there for testing, oh the joys of watercooling LOL

Glad to hear you’ll be testing the card in another pc. That will definitely get you some answers.

Have you tried testing the card in your other pci-e slot? Have you tried setting the pci-e lane to gen3/gen4 instead of auto? Bsod 0x3b could possibly point to a cpu issue.

Well that was a rollercoaster of trouble, FYI switching to GEN3 for the GPU yeilded the same but it locked up.
 
But here is some more, I restarted from a functional desktop, I was unable to get into the BIOS due to d6 and a1 BIOS codes, various reset buttons got me nowhere until I flicked the psu switch off and back on,  then after the crash I struggled to reboot with constant d6 errors no mater what I did, before eventually gaining boot.
 
The worst part is the PC is 100% stable until I load a game.
 
Ready for the next idea until I get the time to part swap.
 
Also switching the pci slot becomes a little problematic due to cooling loop ;)

Right then
@Cool GTX, love the full array of questions ;)
Monitor is Samsung LC34J791 3440x1440 UW 100Hz
Cable is the suppled cable with monitor I believe (can't remember)
Fast boot and win fast are what ever defaults they are as I don't change them.
 
So anyway today I cracked on with testing the gpu
I removed it and hooked it up to the wifes PC
That was fun /s so after it failing to give me display I reset the BIOS and got it up and running, after sorting steam and running port royal it ran fine, Good not the gpu...
re-assembled my main PC and thought why not reset the BIOS here, even though it was reset when I did the update, let everything as default and booted windows to run the test... It passed (confused dot com)
Rebooted and made the BIOS changes by default, RAM speed, Wi-Fi off, ect.
Loaded windows, ran the test and it passed...
 
So I am back up and running (for now) and will obviously see how it goes. My speculation is that I had a software bug, or tweaked the bios wrong, or even a bad power connection.
 
I wish I knew so I could say for sure.
 
Longer testing to call it working is needed, but atm the strangeness is gone.
 
Another note just for the record, Windows during the problem decided to de-activate due to hardware change and I had to re-enter my code for it to activate again, I don't know if that coincided with the MB BIOS update though LOL
 

Cheers, I was able to game for a little before popping out and all was well.
Perhaps the saved hardware info was corrupt?
 
I don't like not knowing why, even though I'm pleased it is now working.

And it's back :(
 
Was happily gaming, was about to quit when the screen went off, restarted and now am unable to game again....
This is equally perplexing as it is annoying.
So I need to do some more troubleshooting, I think I might start with a BIOS reset and see what the result is.

Dwarfy
 
Also switching the pci slot becomes a little problematic due to cooling loop ;)

Been there, done that.  Pain in the wazoo, etc. 

I think it's best if you purchase a back up set of air coolers for your CPU (CM Hyper 212 EVO) and keeping the air cooler on the GPU, presuming you bought it as air cooled and bought the GPU WB second, should save you lots of headaches during troubleshooting and one of the main reasons why I'm eventually leaving custom H20 to AIO cooling to keep it simple on my next major PC upgrade.

Anyhoo, were your GPU and loop's thermals monitored while gaming?  If so, what were they?

Let us know how the bios reset does.

A little test I did before I do anything else
I tried a couple of less demanding games
Skyrim (locked @60fps (if you know you know ;) ) - Worked fine windowed and full screen
Talsiman Digital Edition - Worked fine
Two Point Hospital - Noped out shortly after saying SEGA
 
Gonna remove the power cable thing now

Dwarfy
New feedback from the removing of the strimer
 
I disconnected the cables and then had to pull them through the bottom to enable them tro reach the gpu.
There are 3 cables from the psu to the gpu
Untangled 1 plugged it in
Untangled the second
Untangled the 3 and suddenly the whole cable is in my hand ???
 
It would seem that it possibly wasn't fully inserted and locked
 
Ran a test and it's fine
 
I am hereby formally handing in my PC building permission slip as it would seem that I am incapable of the simple task of fully inserting psu cables, either that or I'm going back to non-modular LOL
 
I will of course report back in a few days to say it I'm still lup and running, but based on everything thus far and logically this seems to be the fault.
 
Thank you all for help, and my advice if you get any problems... Check all the connections twice ;)
 

Geese. Here I started leaning power issues from the start..

Oh boy.. let's hope all is now fine!!

Though after all that I'd reinstall windows fresh haha.

But yeah. Keep testing it to make sure first.

yes, that would cause issues - if the power cable was loose

those power cables can get tight in the socket before they are fully installed ... you should always check that the little latch has locked them into the socket (on both ends) ... I like to push on the wires, supporting the GPU while installing them
 
 
 
 

It happens to the best of us. 
 
For future reference and hopefully in a quiet environment, allow the cables to 'click' in place lets you know it's been latched on.

Ok it's been a few days, I've gamed on and off over these few days, and I have to say that the issue has not returned.
 
Thank you to everyone for their help