My PC occasionally shuts off completely while playing Microsoft Flight Simulator 2024, and it has also happened in Uboat. The display goes black, the GPU fans stop immediately, and the system restarts itself after roughly three seconds. There is no blue screen, freezing, or obvious warning, and the computer works normally after rebooting.
The problem is inconsistent: sometimes I can play for days, while other times the system powers off after only about ten minutes. It can happen even with low graphics settings, whereas Cyberpunk 2077, Starfield, and X-Plane 12 can run for hours on Ultra without causing a shutdown. CPU temperatures are usually around 72°C and the GPU core reaches about 80°C.
The original 650W MSI MAG A650BN power supply was replaced with a Corsair RM850e 850W, but the issue continued. A repair shop also ran CPU and GPU stress tests without reproducing the problem. I have cleaned the system and monitored temperatures with MSI Afterburner, but I have not found anything conclusive.
The system is about 15 months old and uses a Ryzen 9 5900X, XFX Speedster MERC 308 RX 6650 XT 8GB, ASUS ROG STRIX B550-F Gaming WiFi II motherboard, 32GB Corsair Vengeance LPX DDR4-3200 CL16 memory, WD Blue SN580 2TB SSD, Corsair iCUE LINK TITAN 240 RX cooler, Corsair RM850e 850W PSU, and Windows 11 Home.
What could cause a complete power loss in only certain games when temperatures, stress tests, and a PSU replacement all seem fine? What should I test next?
6 Answers
Inspect the motherboard reset-switch header. If the case's reset button is connected, disconnect it temporarily and test the machine that way. A faulty switch, loose wire, or bent header pin can sometimes trigger what looks like a sudden power failure.
The GPU core temperature alone does not tell the whole story. Check the GPU hotspot and VRAM temperatures while the game is running, since those can be much higher than the reported core temperature. It would also help to find out exactly which tests the repair shop performed and whether they tested the complete system under realistic gaming loads rather than testing the CPU and GPU separately.
Monitor the AIO pump RPM and coolant or CPU temperatures during a load test. A pump that intermittently stops can cause the motherboard to shut the system down for thermal protection, even if the average CPU temperature looks normal. Also check the cooler's mounting and thermal-paste contact.
Try disabling EXPO or any manual memory overclocking in the BIOS and test with the RAM at its standard settings. Memory instability can appear only in particular games. If that changes nothing, test with CPU C-states disabled as another troubleshooting step, since some Ryzen systems have had crashes during rapid transitions between high load and low-power states.
Look in Windows Event Viewer under Windows Logs and System for entries immediately before the restart. A Kernel-Power 41 entry usually only confirms that the shutdown was unexpected, so also check for WHEA hardware errors, display-driver errors, thermal warnings, or other events around the same timestamp. Those details can help distinguish power delivery, GPU, memory, and motherboard problems.
Check whether the power supply's eco or zero-RPM mode can be disabled and try running with it turned off. It is not the most likely explanation, especially with a new PSU, but similar intermittent reboots have been caused by power-saving fan modes or transient behavior.
Related Questions
Lenovo Thinkpad Stuck In Update Loop Install FilterDriverU2_Reload