The second NVMe version led to DRAM problems.
The second NVMe version led to DRAM problems.
I recently managed to assemble a powerful system (details below). I began with a 2TB NVMe drive, but due to affordable pricing, I upgraded to an additional SSD. After purchasing the ADATA M.2 2880 Gen4 2TB unit, everything functioned perfectly without any problems. But once I carefully removed the 4090 to reach the M2_2 slot, the system failed to boot, and the motherboard displayed a persistent DRAM light. Could you help identify what might have caused this? Although I've worked with computers before, I haven't modified one in nearly a decade, aside from a few graphics card updates. Checking the RAM confirmed it was intact, but I had to disassemble almost everything except the CPU at least four or five times, testing each DIMM slot and more. After removing the old SSD (which had been sitting during troubleshooting), the system booted into BIOS normally. Any suggestions on possible causes? Thank you for your advice! CPU: AMD 7800X3D Motherboard: ASUS TUF Gaming X670E-PLUS WiFi RAM: G.Skill Trident Z5 RGB 2x32GB DDR5 6400 GPU: ZOTAC Gaming 4090 AMP Extreme AIRO SSD: WD_BLACK 2TB SN770 PSU: EVGA SuperNOVA 1000 G5 Cooling: DeepCool LT720 AIO
The debug indicators aren't actual error signals. They display which section of the POST process the board is executing, helping you pinpoint where things may stall. This doesn’t confirm RAM failure—it suggests the board is spending time testing RAM during POST. If multiple sticks show similar issues, the problem likely isn’t with the RAM itself. It could stem from the board or CPU (though I’m not certain). A faulty M.2 slot might also be involved, as NVMe drives use PCIe that often connects directly to the CPU. On AMD systems, having many lanes means a problematic slot could affect the CPU. If the CPU becomes unstable, it may freeze unexpectedly during POST, which can trigger DRAM lights. This scenario is mostly theoretical; I haven’t encountered bad NVMe or M.2 slots causing DRAM debug lights before. Since you have an integrated graphics card, it’s best to run tests without it to save time.
It seems the SSD was likely the issue. I managed to check it from outside and encountered some problems. After replacing the same type of SSD (Samsung 980 Pro 2TB) in the same M.2 slot, performance improved significantly.
The same problem occurred with a crucial T500 2tb on an Aorus Master Z390. The system wouldn’t boot with C4 code on the MB and displayed a red light for DRAM. I tested multiple T500 2tb units in all M2 slots (M2A, M2M, M2P) without any SATA or devices connected, even the GPU. I also tried Lexar NM710 and Intel SSD6, which worked across all three slots. The only successful run was with just two RAM bars, as the system detected only one while the crucial was installed. Running with 8GB out of 32GB allowed me to use the Crucial drive. In my opinion, the Crucial T500 isn’t compatible with the Aorus Master Z390. It seems the POST is trying to check the DRAM on the M2 drive, but the other two M2 NVMEs don’t have DRAM, which might explain why it works fine otherwise. I’ll need to test another DRAM SSD M2 drive to confirm. For a Gen 3 system like this, I can’t reach the full speed (max 3500mb/s). I haven’t seen similar issues on forums, so I’m giving up on troubleshooting.