Signal degradation in RTX 3090 FE model.
Signal degradation in RTX 3090 FE model.
Hello, thank you for your patience. I've already looked through the forums, but I still couldn't find a clear answer to my issue. I'm trying to figure out whether I should replace my S.I. to get a new card. Notably, my monitor occasionally shuts down completely and restarts—like the driver was resetting. This occurs randomly during gameplay and while using the PC for non-gaming tasks. It usually lasts just a few seconds, though it can happen multiple times in a row. Sometimes it's more frequent on certain days. The bigger concern is that I lose all signal to my monitor at random moments while playing Elite Dangerous. The system stays on, and I can still hear audio, like when chatting on Discord. The only immediate solution is to power off the machine using the physical button on the chassis.
So far, I've tried:
- A fresh installation of Windows 11
- Swapping the Display Port cable
- Running 3DMark and stress tests
- Using Time Spy and Fire Strike Extreme
Results from those tests:
- Time Spy: [link]
- Fire Strike: [link]
- OCCT Stress Tests: All passed
CPU and VRAM passed. Unigene Superposition passed all benchmarks.
But I'm still puzzled because the signal loss seems to happen in Elite Dangerous only. The flickering appears randomly across different uses—gaming and non-gaming. I'm also worried that the 3DMark Extreme tests didn't pass. I managed to detect the issue using HWInfo64, and most GPU sensor readings drop to zero during those moments. You can find my detailed results here:
- https://docs.google.com/spreadsheets/d/1...ue&sd=true
CPU: AMD RYZEN 9 5950X
GPU: RTX 3090 FE
Boards: ASUS ROG Crosshair VIII Hero
PSU: EVGA 850W SuperNOVA P2
Cooling: Custom water loop (CPU, GPU, VRAM with 1x 240 & 1x 360 radiators)
Monitor: ACER Z35P
Please let me know if you need anything else!
Try adjusting the monitor to 60Hz and observe if the problem persists. Alternatively, check another cable—perhaps you can monitor your VRAM temperature with HWinfo64. On your FE 3090, the VRAM temperatures aren’t particularly high; it seems like the GPU isn’t overheating, but the hotspot is above 100°C in some apps. That’s quite worrying.
Thanks for the update! The refresh rate remains at 120Hz on both the monitor and the NVidia panel. I plan to adjust them and run a test to confirm the issue persists. Also, I didn’t mention using a different Display Port cable—it still caused the problem. My hot spot temperature is consistently above 100°C, which is unusual since I believed it was normal with that sensor.
I reviewed my FE data and the maximum hotspot temperature recorded was 79°C. (bone stock) Could indicate a promising trend.
I'm curious whether the issue lies with the card or with how Maingear applied the water block and backplate. I'm considering upgrading to a better water block and backplate, plus EKWB's 3090 FE setup. There might be a diagnostic tool that can help identify the problem. Should I give up my PC temporarily for testing, or do I need to proceed with repairs? If Maingear made mistakes during installation, I could handle the thermal pad replacement myself. But if the card is faulty, I'd definitely request a return for exchange. Would checking CPU temperatures be useful in diagnosing this?
For this problem it seems the spreadsheet is missing, thank you for waiting, sorry for the delayed response—I need you to confirm or I won’t be able to proceed.
Great! I appreciate the update and the follow-up. I reorganized the files for Maingear's review. You can access my logs here: https://drive.google.com/drive/folders/1...sp=sharing
Performance metrics are outstanding, with a great temperature distribution across cores—no problems detected.