Problems with the 3090 Trinity / mobo/PSU?
Problems with the 3090 Trinity / mobo/PSU?
For a few weeks now, my GPU has been acting up. It began when the PC would shut down unexpectedly while gaming. After that, there was no display and it seemed like the GPU had stopped working. I then connected it to another PCI-e port, breaking both retention clips (I’m still figuring this out). Once I reinserted it into the first slot, it started functioning again—probably because the second slot didn’t fit properly.
At first, when the crash happened, I checked Event Viewer for any GPU-related errors, but found nothing. I tried reinstalling and updating drivers, but the problem persisted. I moved the PC into a larger case to test another PCI-e slot, but it still had issues—2D worked, 3D for short gaming sessions and crashes, temperatures stayed normal. I even replaced the memory and reapplied thermal paste, which helped for about an hour before another crash.
Some odd things have happened:
- Passed stress tests in Heaven/3dmark with the voltage reduced to 1850/875.
- Event Viewer cleared completely after each crash, no trace left.
- System logs disappeared after every failure, no explanation found.
- Task scheduler and other tools didn’t show any related errors.
- The crash log seems to clear itself automatically.
- I can’t boot with XMP profile enabled.
- It appears the motherboard is faulty.
- When I swapped my PSU and GPU connectors, it worked again—suggesting the power supply might be the issue.
I’m planning to try giving the GPU to a friend for testing, as both my motherboard and PSU are still in stock. I’ll also test the motherboard now. What I’ve done so far:
- Moved PC into a bigger case to try another slot (same problem).
- Reseated RAM.
- Swapped GPU rails with CPU rails.
- Repapered or re-applied thermal paste.
What’s next?
- Test the PSU cables.
- Try the GPU on another machine.
- Replace the motherboard.
- Keep PSU cables in order.
Details:
- CPU: 13th Gen Intel Core i5-13600K
- GPU: 3090 Trinity, 24GB RAM (4x8GB)
- Motherboard: Gigabyte Z690 M AORUS ELITE AX DDR4
- PSU: Cooler Master V750 SFX Gold 750W (80 PLUS Gold)
- Task Scheduler and other tools didn’t show any related errors.
- Crash logs are excessively large, sometimes clearing themselves.
- No signs of overheating beyond normal throttling when undervolted.
Use a single stick, test each one separately in the machine. Four dimmers usually point to a problem with one or the board not accepting them.
It has performed perfectly over the last six months, handling games such as Cyberpunk 2077 and God of War at ultra settings without any issues. But now?