I'm building my first AMD-based PC and am having trouble completing a clean Windows 11 installation. The installer usually reaches about 75% and then the computer restarts. It either tells me to remove the installation media and continue, after which it returns to BIOS with no usable boot option, or it starts the installation over again. The internal SSD sometimes does not appear as a boot device, so I'm not sure whether Windows is failing to create its boot partition.
The system uses an ASRock 850M WiFi motherboard, a Ryzen 5 9600K, 32GB of Corsair Vengeance DDR5 in two 16GB modules, and a Crucial P3 Plus 1TB M.2 SSD. The BIOS display is normal and readable, and the drive was wiped and reformatted before installation.
Windows eventually installed, but the system now crashes with error messages that disappear too quickly to record. The SSD manufacturer's diagnostic reported that the drive is healthy, while Windows Memory Diagnostic reported that hardware problems were detected. Could the RAM be defective, incorrectly configured, or unstable? What should I test next?
3 Answers
Make the crashes easier to diagnose by checking the files in C:WindowsMinidump with WinDbg or another dump-analysis tool. Also reinstall the chipset and graphics drivers from the hardware manufacturers. However, software drivers are less likely to explain a Windows Memory Diagnostic hardware warning, so I’d prioritize testing the RAM, checking that the modules are fully seated, and trying conservative memory settings first.
Before reinstalling again, load BIOS defaults, confirm the motherboard firmware supports the processor, and update the BIOS if necessary. Leave only the target SSD and the Windows installer connected, then boot the installer in UEFI mode. At the disk selection screen, delete every partition on the target drive and let Windows create the required partitions automatically. After the first restart, remove the USB drive or select Windows Boot Manager on the SSD instead of starting the installer again.
The memory diagnostic result is a strong clue that the RAM or its settings may be causing the installation failures and later crashes. Test each DIMM by itself, using the motherboard’s recommended slot, and run several passes of MemTest86 from a bootable USB. If one stick fails while the other passes, replace the bad module. Also try disabling EXPO or any memory overclock and run the RAM at its default JEDEC speed while troubleshooting.
That makes sense. I’ll test the modules individually with EXPO disabled and see whether the errors follow one particular stick.