My PC usually runs Fedora Linux, but it also enters a boot loop when I try to start Windows 11. In Fedora, it displays these machine-check errors before restarting:
mce: [Hardware Error]: CPU 2: Machine Check: 0 Bank 5: bea0000000000105
mce: [Hardware Error]: TSC 0 ADDR 1ffffc15c401c MISC d01a000000000000 SYND 4d000000 IPID 500b000000000
mce: [Hardware Error]: PROCESSOR 2:870f10 TIME 1787682678 SOCKET 0 APIC 4 microcode 8701034
The system has also been stuttering and rebooting intermittently for several months. Windows starts loading, then gradually stutters and freezes before rebooting, after which its recovery tool reports that it cannot repair the problem.
Because the issue occurs in two different operating systems, I suspect a hardware fault, possibly involving the CPU. However, I would like to know whether another component—such as the motherboard, memory, power supply, BIOS configuration, or cooling system—could produce the same error. Is there a reliable way to confirm the cause before replacing the processor?
After sitting powered off for a few hours, the PC started normally again. The problem was more frequent during recent hot weather, so overheating may be involved. What temperatures and tests should I check, and would cleaning the system, replacing thermal paste, or installing new fans be appropriate?
3 Answers
Since the machine-check failure happens under both Fedora and Windows, a normal operating-system problem is unlikely. It points toward hardware, but the CPU name in the message does not prove the processor itself is defective. The motherboard, RAM, power delivery, BIOS settings, or cooling can also cause the CPU to report a machine-check error. Check temperatures and stability before buying replacement parts, and return BIOS settings to defaults if anything is overclocked or undervolted.
The fact that it eventually starts after being powered off, and that failures became more common during hot weather, makes heat worth investigating. Once you can boot, record CPU temperatures at idle and during a controlled load test. Also check whether the cooler fan is spinning and, if you use an all-in-one liquid cooler, verify that its pump is operating. Clean dust from the heatsink, fans, and filters, and replace thermal paste if the cooler has been removed or the paste is old. Do not keep stress-testing it if temperatures rise rapidly or it shuts down.
Try a complete shutdown rather than only selecting restart; occasionally a full power-off clears a failed hardware state and allows the next boot to work. That is only a temporary workaround, though. Because the machine has been stuttering and rebooting for months, back up important files immediately and test the RAM, power supply, storage, and cooling as well as the CPU. A working boot after cooling down does not by itself confirm that the processor needs replacement.
A CPU-related error can be triggered by another component, so there usually is not a single message that identifies the failed part with complete certainty. Testing the system with default BIOS settings and checking each major component is the practical way to narrow it down.