How to Fix BSOD MACHINE_CHECK_EXCEPTION (0x0000009C)
What error 0x0000009C means
MACHINE_CHECK_EXCEPTION is one of the oldest and most hardware-centric blue screens in Windows. Modern processors contain a built-in self-diagnostic system called machine-check architecture: when the CPU detects an internal fault — a cache parity error, a bus error, a memory controller fault — it signals it, and if the error is uncorrectable, Windows has no choice but to halt. The blue screen is the CPU telling you, directly, that something in the hardware is broken or unstable.
This is not a Windows bug and reinstalling Windows will not fix it. The usual suspects are heat (a clogged cooler letting the CPU hit thermal limits), an unstable overclock (CPU, memory, or XMP profiles), failing RAM, a dying power supply feeding unstable voltages, or in rarer cases a genuinely failing processor or damaged motherboard. The diagnostic order is the same one repair shops use: cool it down, clock it down, then test components one by one.
Common causes
- Overheating CPU: dust-clogged coolers, failed fans, or dried-out thermal paste push the CPU past safe temperatures under load.
- Unstable overclocking: CPU, memory, or XMP/XPO profiles that were stable when new degrade over time as silicon ages and heat cycles accumulate.
- Failing or incompatible RAM: bad memory modules and unstable memory-controller timings trigger machine-check faults.
- Weak or failing power supply: drooping or noisy voltages make the CPU report faults that aren't really its own.
- Outdated BIOS or damaged motherboard: old microcode can mishandle hardware reports, and bent socket pins or swollen capacitors are physical faults.
How to fix error 0x0000009C — step by step
Work through these fixes in order. The first two resolve the majority of real-world cases.
- Remove all overclocks and reset BIOS to defaults. Restart and enter BIOS/UEFI setup (Del or F2 during boot). Choose "Load Optimized Defaults," which clears any CPU overclock and disables XMP memory profiles. Save and exit. If you use GPU overclocking software like MSI Afterburner or motherboard auto-overclock utilities, reset those to stock settings too.
- Check temperatures and clean out the cooling system. Install HWiNFO and watch CPU temperatures during normal use and under load. Sustained readings above 90°C under load point to a cooling problem. Shut down, unplug, and clean dust from fans, heatsinks, and dust filters with compressed air. If temperatures stay high after cleaning, remove the cooler, clean off the old thermal paste, and apply fresh paste.
- Test the RAM thoroughly. Press Win + R, type
mdsched.exe, and run the Windows Memory Diagnostic. For a deeper check, boot MemTest86 from a USB drive and let it complete at least two full passes. Any error at all means the stick is suspect — faulty memory is one of the most common machine-check sources. If crashes started after a RAM upgrade, test with the old sticks back in. - Update the BIOS and chipset drivers. Visit your motherboard or laptop manufacturer's support page and download the latest BIOS and chipset drivers for your exact model. Install the chipset drivers first, then flash the BIOS following the manufacturer's instructions to the letter. Never interrupt a BIOS update — a failed flash can brick the motherboard. BIOS updates also bring new CPU microcode that fixes how hardware errors are reported and handled.
- Stress-test the CPU and check the power supply. Run Prime95 (Small FFTs) for at least 30 minutes — if workers stop or the system crashes, the CPU itself is unstable, even at stock settings. Also inspect the power supply: a failing unit with drooping rails or worn-out fans causes exactly these faults. If you have a multimeter or a PSU tester, check the 12V rail under load, or borrow a known-good PSU to test.
- Reseat the CPU, cooler, RAM, and check for bent pins. Power off fully and unplug the PC. Remove the cooler, lift the CPU out of the socket, and inspect the socket and CPU pins for anything bent or damaged. Clean off the old thermal paste, set the CPU back in, apply fresh paste, and remount the cooler evenly. Firmly reseat each RAM stick until the clips lock. A loose cooler or poor contact can create the exact heat spikes behind this blue screen.
Did this fix it? Quick checklist
- No more 0x0000009C crashes, even under heavy load
- CPU temperatures stay well below 90°C under stress
- Memory diagnostic passes with zero errors
- Everything runs at stock speeds and voltages
If everything above passes and crashes continue, open Event Viewer, go to Windows Logs, then System, and look for Event ID 18 from the WHEA-Logger source. Those entries often name the exact failing component — a processor core, a cache bank, a memory bank, or the bus. That is the component to replace. When even the component looks fine, suspect the motherboard's voltage regulation or the power supply — unstable power is the ghost that every other test misses.
Frequently asked questions
What does MACHINE_CHECK_EXCEPTION mean?
It means your CPU detected a hardware error through its machine-check architecture and could not recover from it, so Windows halted with a blue screen to prevent data corruption. The fault is reported by the processor itself, not by Windows.
Is 0x0000009C always a hardware problem?
Essentially yes. The error comes from the CPU's own error-detection hardware, so the root cause is physical: a failing processor, overheating, an unstable overclock, bad memory, a dying power supply, or a damaged motherboard. Driver or BIOS updates can only help when the report itself is caused by firmware bugs.
Can 0x0000009C mean my CPU is dying?
It can, but do not assume that first. Overheating and unstable overclocks are far more common causes and are both fixable. Only after cooling, stock clocks, memory tests, and BIOS updates all check out should a failing CPU move to the top of the suspect list.
Can overclocking cause MACHINE_CHECK_EXCEPTION?
Yes, it is one of the most common causes. Even a mild overclock or an XMP memory profile can become unstable as silicon ages, thermal paste dries, and dust accumulates. Returning everything to stock speeds is the first diagnostic step.
How do I find out which hardware caused 0x0000009C?
Open Event Viewer, go to Windows Logs then System, and look for Event ID 18 entries from the WHEA-Logger source. The details often identify whether the fault was reported on a processor core, memory bank, cache, or bus, which points you at the failing component.