Mystery PC crashing
Mystery PC crashing
This article is meant for discussion and imagination only; feedback is always appreciated. In the autumn of 2023 I constructed this setup: processor – Intel Core i9 13900k, memory – 2x16 GB DDR5, RAM type T-FORCE Vulcan, graphics – ASUS TUF GAMING 4090 OC with DDR5 at 5200MHz, motherboard – ASUS TUF GAMING Z790 PLUS, storage – 1TB Samsung SSD 980 PRO, two 2TB Samsung SATA SSDs, wireless – 870 EVO, power supply – CORSAIR RM1000x (2021), case – HYTE Y60. It uses an aftermarket 12VHPWR to a 4x8 ATX cable from CableMod to keep things tidy and avoid the built-in NVIDIA adapter inside the case. Roughly a year into daily use, when I play tough games at high settings, the system occasionally freezes in odd ways. After about 5–10 minutes under stress, the screen turns black and stays dark for roughly ten seconds. Fans spin at full speed, audio keeps playing but sound effects stop or repeat. Keyboard and mouse lag; caps lock still works but ALT+F4 doesn’t stop music. The machine won’t respond until it’s force-shut down and restarted.
I assumed a thermal shutdown happened, thinking the GPU overheated and got stuck in the cramped case. To prevent this again, I took extra steps: used MSI Afterburner to lower GPU temps from 86°C to 75°C, added a Noctua Chromax fan below the GPU hotspot, and cleaned all case filters.
After a 3DMARK stress test, the GPU didn’t hit its baseline frame stability (96% to 92%) but didn’t crash either. So I usually play at lower settings or less demanding games. This week, before playing DOOM: The Dark Ages, I completed all necessary preparations – updated BIOS, CPU microcode, NVIDIA drivers (576.22), and confirmed thermal limits with afterburner.
The surprising issue is the PC crashed right after a 10-hour gaming marathon. Most of the time the GPU was under heavy load (~350W) and the game ran smoothly at 120fps on default ultra settings. It stopped working after about 15 minutes while I checked Discord and emails. My new hypothesis: the power supply (or CableMod cable) is failing and isn’t adapting well to temperature or load changes. I’ve already ordered a replacement PSU (CORSAIR 1500x), which should fix both potential failure points.
Any suggestions or observations? If you missed anything that could be contributing, feel free to mention. Thanks for reading!
I’m not trying to spread bad news, but JayZtwocents shared a video about a comparable problem recently. He mentioned that mysterious crashes like yours usually point to CPU wear and tear. Interestingly, his own machine experienced the same issue shortly after. The same problem, same reason. I’ve included the links below. I hope this is useful.
I was worried that might have played a role, given that the 13.9 million k is included in early batches in 2023 during the microcode incident. I decided not to overwork the CPU then. My main suspicion was the power supply; I changed the cables and PSU to a Corsair 1000x with a Corsair 1500x for reasonable cost. The random crashes in DOOM TDA vanished, and I could safely adjust GPU settings without worrying about power draw exceeding 300W. I understand why you thought about CPU issues at first, but it didn’t make sense that benchmarks were okay while gaming wasn’t. It seems the inconsistent power delivery—either from the cable or the PSU itself—was likely the cause. The reason for the cable was probably to avoid stressing the Nvidia adapter in the box, which could have overheated. Possibly it was also the CPU power delivery that failed. The PC still works well since the PSU change, and I won’t repeat this issue unless it happens again; I’ll stay mindful and treat the processor carefully.
Main causes of these issues include: 1. Faulty GPU power cable or improper connection 2. Outdated chipset drivers 3. Corrupted BIOS settings 4. Defective power supply unit