I've been dealing with recurring shutdowns and blue screens on my PC. A repair shop has already replaced the power supply and SSD, and I recently performed a completely fresh Windows installation. The system was stable for about a month, but it has now crashed while sitting idle with an ATTEMPTED_EXECUTE_OF_NOEXECUTE_MEMORY bugcheck (0x000000FC).
Windows, BIOS, and drivers should be up to date, and I've only installed games through Steam. The system includes an Intel Core i9-14900KF, MSI PRO Z790-P WIFI motherboard, MSI RTX 5070 Ti, 32GB of DDR5-6400 memory, a 2TB PCIe 4.0 NVMe SSD, and a 360mm liquid cooler.
I'm trying to determine whether this is most likely a driver problem, unstable memory settings, or a failing processor. I have warranty coverage, but I'd prefer to gather better evidence before replacing more parts. A crash dump is available for analysis.
3 Answers
I’d stop replacing parts randomly and focus on CPU and RAM stability first. A 14900KF running with DDR5-6400 makes XMP an important variable, so reset the BIOS to stock settings, disable XMP, and make sure the latest Intel microcode is installed. Then test the memory with MemTest86 and run OCCT while watching for errors.
Bugcheck 0xFC can be caused by a bad driver, but corrupted memory can also make Windows try to execute invalid data. Since the SSD and Windows installation have already been replaced, the operating system itself is less suspicious. Check C:WindowsMinidump with WinDbg as well; the dump may identify a driver or provide evidence of hardware instability. If crashes continue at stock settings with clean memory tests, the 14900KF becomes a stronger suspect.
The minidump is the most useful next step. One crash is difficult to interpret, so collect several files from C:WindowsMinidump if possible and inspect them with WinDbg. If Windows is not creating dumps, configure it to save small memory dumps and check that the system drive has a page file enabled. A dump can distinguish a specific driver crash from a more general memory-corruption pattern.
You should also investigate the known 13th- and 14th-generation Intel stability and degradation problems. Updated BIOS versions and Intel microcode reduce the risk, but they cannot repair a processor that was already damaged. The fact that games run normally does not completely rule out a CPU problem, especially since the failures happen randomly while idle. If the machine still crashes with the BIOS at safe defaults and the memory passes testing, using the CPU warranty would be reasonable.
That possibility has been worrying me because the BSODs began months ago. My BIOS has been updated through the relevant fixes, and the crashes are so random that I haven’t been able to reproduce them reliably. I’d rather have stronger evidence before requesting a processor replacement, but the warranty is available if the stock-setting tests point in that direction.
I did try disabling XMP in the past, but that was before replacing the SSD and reinstalling Windows, so I’m willing to test it again. Should I run OCCT with XMP disabled first? I have warranty coverage through the repair shop, but I’d like to narrow this down before using it.