I sent the GPU through two repairs. Is there another problem?
I sent the GPU through two repairs. Is there another problem?
I assembled my system last summer and everything functioned smoothly with the newest BIOS and drivers. A severe thunderstorm hit Nashville in early December, forcing me to shut down my PC and peripherals connected to a power conditioner. After the storm passed, I powered back on but failed to boot. The motherboard displayed DRAM and CPU indicators. I attempted various fixes—reinstalling RAM, PSU, trying different RAM setups—but everything worked after removing the GPU. I then tried booting with an older 1660ti, which resolved the issue. A small chip was visible in the PCIE slot during installation, but it didn’t impact performance. After a first RMA, the same error persisted. I suspected a PCIe slot issue and purchased a mount cable, yet it didn’t help. The second RMA replaced the card, but the problem continued. I’m left without another motherboard or PSU to test. I’ve checked all GPU connections and they light up, though the fans remain silent. What should I do next? Should I request a replacement card, a different one, or return the motherboard for repair?
The GPU isn't the issue because it has already been returned for replacement twice. For clarification, was the power supply still connected during the storm? If yes, then a lightning strike could have passed through the wiring and reached the PC—though this is unlikely as other devices might have stopped working too. Since the GPU lit up but the system didn’t respond, it suggests the power supply unit may be defective. My guess is the problem lies in a tiny chip on the PCIe slot. A picture would help identify what’s damaged.
Refers to the section that makes me unsure. Adding another GPU with the same cables from my power supply—whether it's the 1660ti (8 pins, fewer lanes) or the 7900xtx (24 pins, more lanes)—could still be the motherboard if it fits in the same slot, even on a riser.
Thunderstorms are unpredictable... they can appear far away and move anywhere to cause trouble. To protect your setup, disconnect any copper-connected parts. I usually remove the wall plug since everything is powered from there, and if it's elsewhere, unplug that too. For your rig, try a different power supply first—they tend to fail early. Check the GPU on another system to confirm it's not dead. I've noticed this happen before; the replacement card ended up damaged.
It might be possible to connect the power supply only through certain modular outputs. I wondered if it would work without all components if it were just the power supply, or if a damaged PSU couldn’t manage a larger GPU. My setup with the 1660ti still functions fine.
It shows how PSU performance declines gradually over time, with power delivery dropping noticeably. A sudden surge like a lightning strike might happen, but from a distance it could still shorten lifespan significantly. It's unclear exactly how a PSU reacts during such events if not properly disconnected. Yes, it's possible to run an older GPU but not a newer one. Have you checked the GPU's performance across different setups?