cancel
Showing results for 
Search instead for 
Did you mean: 

3080Ti TUF OC 165hz monitor randomly loses signal

TQI1127
Level 7
I am having a very weird issue and I want to confirm if this is really hardware issue. My spec: 12900K, ROG Z690-E, 3080Ti TUF OC, Gskill 6000mhz CL36 2x16, two 980 pro 2TB SSD, Corsair 7000D case, EVGA Supernova 1000W PSU. BIOS updated. GPU Vbios updated. GPU not overheating, always 60C around on gpu, 75C on memory under stress test.

Currently I have 3 monitors. One Dell 1440p 165hz HDR, one Samsung 4k 60hz, one LG 4k 60hz.

Ever since I built the pc, I’ve always had this weird issue, from time to time, usually a restart, the two 4k monitors will have the wallpaper scaled smaller and moved to the top left corners. I can correct it if I go to display setting and change the scaling to 100% then change it back to 150%.

Then there was this nvidia 511.79 driver, I used DDU clean installed it as I always do for all driver update. Then I started seeing event viewer nvlddmkm error id 14. Very consistent upon boot 3 errors will pop up. Then all of a sudden I started to randomly lose monitor signal always the 165hz one. Can be in game, can be idle. Sometimes the system freezes with 165hz monitor no signal, two 4k monitors showing image but blurry. Sometimes all 3 monitors lose signal but I can still hear sound. Today I had a teams meeting, I lost all displays but participants can still hear me talking.*

What I had tried: DDU and clean install latest driver, no fix. DDU install previous driver no fix. Reseated gpu worked for some time then next day cold boot issue came back. Fresh windows install total 3 times, no fix. Change all DP cables no fix.

Then this evening system suddenly became stable. I was able to complete timespy stress test 20 loops and pass with 99.5% frame accuracy. Gpu boost clock temperature all good. Then I realize windows somehow reset my 165hz refresh rate to 60hz. I then change it back to 165hz issue came back again, randomly loses signal. I tried a known good 144hz 1440p monitor, same issue. Now I started to feel the card is somehow not handling the high refresh rate. I now set my 165hz monitor to 120hz everything seems stable pass stress test again. And no more weird scaling issue upon restart.*

I just want to see what you guys think, I feel this is hardware issue as I have another pc with 3080, and I used to connect 144hz 1440p, 165hz 1440p and another 4k 60hz on it, it never had any issue.

Sorry for the long post and Thank you for your time. *
3,498 Views
6 REPLIES 6

TQI1127
Level 7
Just want to add I’m on windows 11, and event viewer shows every time the system freeze occurred, it wasÂ*DWM.exe faulting dll dwmcore.dll error. I’ve also seen a WHEA error showing PCIe slot 1 root port error, and the vender ID plus device ID pointed to the PCIe controller. But this only occurred once, after fresh windows install it went away.Â*

Mister_Benson
Level 9
By any chance, do you use a rizer to connect the GPU to the motherboard?

Mister Benson wrote:
By any chance, do you use a rizer to connect the GPU to the motherboard?


sorry for the late response. When I first built the pc yes I had a riser. Then I saw the weird scaling issue I thought it was riser so I removed it but scaling issue never went away. But I believe I fixed the issue I’ve updated above. Thank you for your response!Â*

Silent_Scone
Super Moderator
Hello,

1. Reseat the GPU

2. If using a riser cable plug the card directly into the board and see if the issue persists

3. Try an alternative DP cable and make sure it is correctly certified
13900KS / 8000 CAS36 / ROG APEX Z790 / ROG TUF RTX 4090

Hello, thank you for your reply and sorry for updating so late as I seem to have fixed the issue but wanted to wait a few days before coming here to confirm.Â*

My last attempt seemed to have fixed the issue. So far I’ve had zero PCI express root port warning or errors, and no more weird scaling issues or random black screens. I did reseated gpu but the thing I found was if I install windows with monitors connected to the 3080Ti, with a windows installed nvidia driver, at first everything would be good until a windows 11 .NET framework update kicked in. As soon as the framework update was done and restarted, all of my issues came back. So what I had done was: connect one monitor to igpu, clean install windows, then install all chipset drivers, igpu drivers and other drivers. Then proceeded to update all windows updates including the framework updates. After restart, downloaded the nvidia 511.23 driver. Turn off all internet adapters, turn off my pc. Connect monitors to 3080Ti, go to bios set primary display to PCI, then got into windows and install nvidia driver without internet then restart. After doing all these, it’s been 3 days and my main monitor is running 165hz with HDR, with freesync all. Done almost 10 times the timespy 20 loop stress tests, and doing normal work and play games, no more nvidia driver error, no WHEA errors, and no black screens randomly, the scaling issue upon restart is also gone.

It is a very weird situation. I found this by noticing the multiple monitor scaling issue always occur after the framework update restart. In addition, I have another pc with 10900KF with 3080, after inspecting event viewers, that pc also has a bunch of WHEA warnings about PCIe root port error but windows was able to correct it. On that pc, this warning started on 10/28/2021 and exactly on that day I’ve installed the windows 11 update (KB5006746) and .net framework update (KB5005537) and a preview .net framework update (KB5006363). But my 10th gen pc never had instability issue. The warnings are there but system is stable. However on the 12th gen pc, it caused a lot of graphics related issue as I described above. It seemed as if the framework update when installed with a nvidia driver present, it would do something to the OS… I’m not sure but so far it is stable. If issue comes back I’ll update here. Thank you for your response. Â*

Silent Scone@ROG wrote:
Hello,

1. Reseat the GPU

2. If using a riser cable plug the card directly into the board and see if the issue persists

3. Try an alternative DP cable and make sure it is correctly certified


thank you for your response. I’ve done that several times but I think I’ve fixed the issue and I updated above. Thanks again!Â*