This graphics card can handle heavy video rendering without stopping the whole computer.
This graphics card can handle heavy video rendering without stopping the whole computer.
I recently acquired a used gaming computer from eBay, purchased through a reputable seller with numerous positive feedbacks. It was sold by N&T Electronics, an LLC registered as a legitimate business. I believed this purchase offered excellent value.
The system features an ASRock Z690 CPU with integrated cooling, the latest GPU is an EVGA RTX 3080 X3, and the motherboard supports up to 64GB of RAM in a CL16 configuration. The power supply is an EVGA 1000W 80+ Gold SuperNova GT model. Storage includes two WD Black Sn850X drives at 2TB each, connected via PCIe Gen 4x4 lanes on the motherboard.
During the initial setup, everything worked smoothly. Performance remained strong across various games, and high frame rates were achieved on a 3440x1440 display. However, after upgrading RAM and reconfiguring the motherboard connections due to cable management issues, I noticed unexpected problems. Within minutes of running GPU-intensive applications like F124, artifacts appeared on the screen—small white spots that worsened when video output was disabled.
Investigating further, I discovered the issue might be related to the PCIe slot configuration. Upon moving the graphics card to a different slot, the artifacting ceased. This suggested a possible hardware fault rather than driver-related problems. I also tried booting with a WD Blue SN570 M.2 drive running Windows 11, which resolved the issue entirely. The card’s drivers were up to date, and performance remained stable for everyday tasks.
Despite this, when under heavy load, the system would abruptly shut down, even though temperatures stayed within normal ranges. I’ve tested various RAM setups and power limits, but nothing seems to prevent the video output from cutting out. The only clue was that the PSU should handle both the i9-12900K and the 3080 without issues, yet it failed under stress.
I’m seeking advice on how to resolve this persistent GPU artifacting and prevent future shutdowns. Any guidance would be greatly appreciated.
Running OCCT with full VRAM coverage shows no errors, but that doesn’t confirm stable VRAM performance. It could still indicate potential issues like dying VRAM if stability isn’t maintained under load.
I don't have alternative options for placing my GPU, but OCCT's Power stability tests—which essentially run all parts at full capacity—show consistent results. The GPU power delivery remains stable, delivering around 317-319 watts without any issues. Temperatures are also normal, staying between 77-79°C after six minutes of maximum load. GPU temperatures are within range, and the OCCT VRAM stability test ran smoothly for 30 minutes with no errors or crashes beyond CPU thermal throttling.
Uncertain if this assists, but OCCT stopped after roughly five minutes of Steady 3d Adaptive testing at the "Extreme" preset which uses full GPU power. No artifacts appeared—system cut off and restarted. Report of said crash in Event Viewer: Unable to recover from a kernel exception. The program must exit. Error code: 3 (subcode 7) (pid=17808 tid=1356 occt.exe 64bit). See NVIDIA support page for details.
Yes. Installing Windows 11 in two days for the newest release sounds possible, though I’m not sure. Repasting might work if the issue lies with the GPU core. The EVGA stickers on the screw holes suggest the board hasn’t been fully taken apart yet.