Some games are causing my entire Windows 11 system to freeze, sometimes followed by a black screen and an automatic restart. This is more severe than a normal game crash. Event Viewer usually reports Kernel-Power 41, and one incident showed 0xEF CRITICAL_PROCESS_DIED. There are no useful crash dumps. I have a fresh Windows installation with very little third-party software, and I am not using inpoutx64.sys or the suspected KB5121003 update.
The system passes CPU, GPU, memory, and SSD benchmarks, and it remains stable while browsing, watching videos, editing or rendering, and performing other normal tasks. The crashes happen almost exclusively during gaming. Before several crashes, the system drive suddenly reaches 100% active time, but a PowerShell monitoring script did not identify the responsible process.
My hardware is an Intel i7-14700K, Zotac RTX 4080 Super, 32 GB DDR5-6400 memory, a 1 TB Kingston NVMe system drive, MSI Z790 motherboard, Asus ROG 650 W PSU, and a 360 mm liquid cooler. Temperatures appear normal. Could the SSD, storage controller, memory settings, BIOS microcode, or power supply explain why gaming triggers this behavior?
4 Answers
Check the SSD manufacturer's firmware and monitor its temperatures with HWiNFO while gaming. Also try disabling Link State Power Management in the active Windows power plan and move the NVMe drive to another M.2 slot. A failing drive, controller, slot, or power-management state could explain the sudden 100% active time before the crash.
The 650 W power supply is a possible weak point. The GPU can draw around 320 W and the CPU can approach 250 W under heavy loads, before accounting for the motherboard, drives, cooling, fans, and USB devices. Even if games do not max out both components at the same time, that leaves little headroom. Testing with a known-good 850 W or higher PSU would be a useful A/B test.
Because this is a 13th/14th-generation Intel processor, make sure the motherboard BIOS and Intel microcode are fully up to date. Also temporarily disable XMP and run the memory at its standard settings. DDR5-6400 is an overclocked configuration, and instability can appear only under particular gaming workloads even when separate memory tests pass.
Look in C:WindowsMinidump and check whether any dumps were actually created. If none exist, that is useful evidence because the failure may be a hard reset or storage hang rather than a conventional blue-screen crash. After the next incident, review the previous two hours of System events for Disk, StorNVMe, NTFS, volmgr, WHEA, and Kernel-Power entries. Those logs may show whether the NVMe drive or storage controller stopped responding.
The temperatures being normal does not rule out power delivery, firmware, or memory instability. I would test with updated BIOS settings, XMP disabled, and a different PSU before assuming the SSD itself is defective.