My PC has developed a very specific freezing problem over the past few weeks. It hard-locks to a black screen, and the keyboard becomes completely unresponsive—Caps Lock no longer toggles its LED—so it appears to be a full system or hardware-level lockup rather than just a display-driver crash.
The problem happens almost exclusively during the first 15–30 minutes after a cold boot, while the system is idle, or right when launching Steam or when Windows finishes uninstalling an update. If the computer gets through that initial period, it becomes completely stable for the rest of the day. I can play demanding games and run long FurMark or 3DMark stress tests without errors once it has warmed up.
The system uses an Intel Core i5-14600KF with Hyper-Threading disabled, an RX 7700 XT, an ASUS PRIME H610M-K D4 motherboard, 32GB of Kingston Fury Beast DDR4-2666 memory, and a Cooler Master Hyper 612 APEX cooler. The motherboard has no VRM heatsinks. The power supply was previously a Chieftec ELP-700S 700W unit.
I have already run SFC and DISM, removed a recent Windows quality update, tried the Ultimate Performance power plan, performed a clean GPU driver installation, updated the BIOS, and reseated and rearranged the GPU and RAM. The graphics card appears stable under heavy load. After a crash, AMD's software resets my custom fan curve and reports an unexpected system malfunction.
Today I replaced the PSU with an EVGA 850 GQ 850W Gold unit and disabled XMP. I also have a Gigabyte B760M Gaming X DDR4 motherboard available to test. Could the cold-start behavior be caused by the H610 board or its VRM/power-state handling, or is there another component or BIOS setting I should investigate?
2 Answers
The H610 board's power delivery is worth taking seriously here, especially since it is running a 14600KF and has no VRM heatsinks. Cold components can behave differently during startup and low-load transitions, and the fact that the machine becomes reliable after warming up points in that direction. Test the new PSU first, then try the B760M board. If the crashes disappear with the better-cooled motherboard, that would be strong evidence that the original board was the problem.
This pattern also fits a low-power-state stability problem more than a GPU failure. Since the card is stable during sustained gaming and stress testing, try running completely at BIOS defaults with XMP disabled. Also check for any CPU undervolting, ASUS power-saving options, or aggressive idle-state settings and turn those off temporarily. The PSU replacement is a useful test, but the H610 board and its ability to handle the 14600KF during idle transitions should be investigated before assuming the graphics card is bad.
Aside from changing XMP, enabling ReBAR, and disabling Hyper-Threading, everything else is still on its default setting. I will test with the default configuration and compare it with the replacement motherboard.
It worked normally for more than a year, but the black screens now happen most often when I launch Steam shortly after a cold boot. Once it survives the first 15 minutes, it runs all day, so I am hoping the motherboard is the cause.