The 9950X3D continues to fail repeatedly without any bootloader access.
The 9950X3D continues to fail repeatedly without any bootloader access.
Hello, I recently assembled a new PC and after a couple of days everything functioned properly. However, soon after, it began to malfunction repeatedly—crashing or failing to boot into Windows. The signs were clear: monitors went dark, fans spun wildly, and the CPU indicator lit up, yet no debugging information appeared. Initially, this occurred after leaving the machine on a locked screen for hours; later it worsened, making it rare to even launch Windows.
Could you help identify what’s causing this issue? Details I have: processor model 9950X3D with GPU 9070XT, NXT PSU C1200, Gigabyte X870E AORUS Pro storage, Samsung 990pro RAM, and a Flare X5 64GB (2x32GB). Firmware and drivers are Windows 11 Pro, BIOS version FA3a, with the latest updates via Gigabyte Control Center.
I’ve tested everything: stress tests on CPU and GPU, long gaming sessions without problems, and even reinstalled Windows after a crash during Stellaris. I flashed the BIOS again, reset CMOS, tried Expo on/off, and ran memory checks (MEMTSTREAM, Prime95). All tests showed normal results except for occasional crashes during stress tests.
Memtest didn’t reveal issues until the first pass at 1/4, but temperatures stayed within safe limits. Prime95 reached 6 cores before failing, with stable temps afterward. I reset the CPU and verified no bent pins, but a full GPU stress test still crashed on CPU testing.
In BIOS mode it works fine, but getting into Windows remains a lucky event. The buttons (Del, F12) don’t respond. It rarely exceeds 80-90% CPU load during stress tests.
I’ve tried disabling Steam and Discord via Task Manager to ease booting, but that didn’t consistently help. Windows Energy settings didn’t make a difference.
I’m unsure what else to try next—any advice would be greatly appreciated. I also shared this on Reddit yesterday, hoping more people could assist.
Yes, the entire system may stop working during the RAM tests in MemTest86.
I understand. I assisted a friend with his failing laptop that was showing various BSODs and sudden crashes. It lasted about 3 to 4 minutes before it stopped working. The issue turned out to be a faulty RAM stick, which I replaced and it functioned properly afterward. Because the machine shut down quickly, we couldn't test multiple RAM modules at once, so we identified the problem by checking each one individually.
Hi, thank you for your responses. I tested my ram sticks by first disabling expo again (cmos reset) and testing one stick at a time. The system nearly got through a new windows install onto the second and nearly empty nvme ssd (only games were copied onto there). This was done to rule out driver/windows issues. Im pretty sure it crashed right at the last step where its jumping from the Installation to the normal Desktop. As when i restarted the pc i got into Desktop. The second stick also caused the system to crash. to rule out just a dead motherboard ram slot i also performed the test on another slot, but with the same issue. sadly i did not try to lower the clockspeed further than non-expo speeds. since i was all* out of things i could do myself i gave the system to a computer Shop where i used to go to have my old systems build. It was my last resort and it still hurts a bit but for only 25€ i cant complain, as long as they dont mess anything up. Right now im just waiting to hear from them again (they guess it will be the coming monday). And i really hope it was just a lose cable so i dont have to wait for rma. I waited 2 months abroad to build this ;( *the only thing i didnt check to be honest were the cables headed into the mb for cpu power and on the psu side. I noticed it waaay to late. But im sure they all clicked fully in and i didnt move anything as soon as the system worked. ill try to keep you posted for updates and appreciate any further input
Hey, here’s a quick update on what’s going on. The PC shop is having trouble locating the problem, and because it can take hours or even a whole day to resolve, it’s really hard to track down completely. On another test bench, my CPU ran fine—even with all my components, just swapping the mainboard fixed it. My mainboard with a different CPU (9800) also worked. So both the CPU and mainboard seem to function independently, but when combined again, the issue resurfaces. I’ve replaced every part except the mainboard, tried various BIOS versions, and even tested an external Windows drive. All that points to something else being the root cause. The latest BIOS (FA5b) didn’t help either. I ordered a replacement mainboard from Amazon to test it while I return the faulty one for repair. Unfortunately, when I picked up my system yesterday, it had visible scratches, dents, and dirt everywhere—including on the CPU bracket. It broke me the same day. After years of saving and planning, this really threw me for a loop. I’m going back today to ask the shop staff about what’s happening and whether I’ll be compensated. Even if I am, it still hurts knowing I invested so much time and money. I’ll keep checking in and see if the system stabilizes over a few days. Just wanted to share this with anyone else who might run into the same problem. Here’s a photo of my CPU cooler—GPU is fine, but the case is covered in scratches, dents, and dust, even the CPU bracket looks messy. The NVME heatsink I received was full of dirt, smudges, and hair. I lost it when he handed it over.
I get how you're feeling, buddy. Things are tough right now with the old system, but things won't stay this way forever. You'll figure it out and end up with a functional PC. Don't forget the warranty on new components—make sure to check it out soon and use it if necessary.