The top PCIe port isn't functioning properly—could it be due to a faulty CPU?
The top PCIe port isn't functioning properly—could it be due to a faulty CPU?
Previous configuration: I had a Ryzen 2600 with Asus Rog Strix B450-F Vega 64, upgraded to a 5600X, updated the BIOS, and installed it. Continued experiencing VGA error LED on the motherboard. Switched to the bottom slot but it didn’t work either. Believed the GPU was failing. Tested with R9 390 in both slots; only worked in the bottom. Reinstalled Vega 64, which functioned in the bottom slot this time. Assumed the motherboard was faulty—possibly outdated or defective. Upgraded to a new 5600X and it worked fine. Now using an MSI Tomahawk B550, updated BIOS for Ryzen 5000. The Vega 64 still doesn’t work in the top slot but does in the bottom. On both old and new boards, GPU logo lights up and fan spins, though the GPU load display (GPUTach) remains inactive. Everything else operates normally; XMP settings loaded well, no further problems. My conclusion is a faulty CPU—no bent pins, looks good physically, and I handled transfers carefully. Any advice would be appreciated. Is this eligible for an RMA? Appreciate any insights!
You've tested two different motherboards with the same issues; the CPU is likely the culprit. B450 boards tend to split PCIe lanes when they have multiple slots, while 550 boards sometimes behave differently based on the manufacturer's design. It seems the board might be using lane pins 8-16 for the second slot when a GPU is installed there, while the first slot uses a different configuration. The B450 likely supports both slots with 1-16 pins and lets the CPU manage lane allocation. At this stage, RMA appears to be the most viable solution since further testing options are limited.
Sure, I’ll proceed with the return. Appreciate you sharing the details—just wanted to double-check everything before the CPU was taken out for a while.
I recognize this discussion is quite outdated, yet I think it could still be useful for someone. Not everyone has spare components ready for such fixes, but I learned a lot. My CPU was definitely failing, which prevented the primary PCI-E x16 port from working and triggered error 43 in Device Manager, turning off the GPU. I dug deep into troubleshooting steps. I manually disabled or enabled the GPU, swapped out two old GPUs, reinstalled Windows 10, tried Windows 11 with new drivers, updated drivers, used DDU, etc. I checked power connections, changed power supplies and cables, swapped RAM around the flashboard and GPU BIOS, even tried different motherboards. It lasted about 30 minutes before it failed again. This made me wonder if the CPU controller itself might be the problem. I only experienced occasional freezes, not major issues. The system had been running about three years before these problems began. After swapping the CPU and GPU, everything worked fine in the main slot. So I can't rule out a faulty CPU here. Wishing anyone facing this trouble good luck, and please don’t blame me if it happens to you.
I know this is also a very late reply but may help people looking for help with the issue. I ended up RMA'ing my cpu and everything was solved with the new processor. Looks like there was some defect that prevented PCIe lanes from working. AMD was great with the RMA process and luckily I had a backup CPU for the wait time.
I'm having the same issue after checking multiple threads. The PC keeps showing error 45 when trying to connect the PCIe slots, even though I've swapped cards and tried different ones. The latest power supply also triggers the same error. It seems like the motherboard might be the problem. I'm considering a new motherboard now—hopefully it fixes it!