Rightek Rescue · Guides · WHEA Uncorrectable Error
WHEA_UNCORRECTABLE_ERROR? This one is the hardware talking
Most blue screens are Windows reporting that its own software went wrong. This one is different. WHEA stands for Windows Hardware Error Architecture, and this stop code means a physical component reported an error it could not correct, and Windows passed the message on. Reinstalling Windows will not fix it, because Windows was never the problem.
What you're seeing
- A blue screen reading WHEA_UNCORRECTABLE_ERROR, stop code
0x00000124 - Crashes that feel unpredictable — sometimes at idle, but classically under load: gaming, rendering, compiling, or a large file copy
- Often no pattern in the software involved, which is itself the pattern
- Frequently after adding memory, changing a CPU cooler, or enabling an overclock or XMP profile
Why it happens
Something in the hardware detected a fault severe enough that it could not be silently corrected. The usual sources, roughly in order of likelihood:
- An unstable overclock — including XMP or DOCP memory profiles. These are factory overclocks. They are usually fine, and they are also the single most common cause of this code. Enabling XMP counts as overclocking even though the BIOS presents it as a one-click preset.
- Faulty RAM. Memory that fails only under specific access patterns or once warm produces exactly this kind of unpredictable crash.
- Heat. A CPU running into thermal limits, usually from dried thermal paste, a seized fan, or dust. Crashes that only happen under load point here.
- Power delivery. An ageing or overloaded power supply produces momentary voltage dips that the CPU reports as uncorrectable errors.
- The CPU or motherboard itself. Least common, and worth suspecting only after the others are ruled out.
Drivers can occasionally trigger it — a driver instructing hardware to do something invalid — but this is the minority case, and chasing drivers first usually wastes days.
Why we are not giving you repair commands here
On most guides this section lists the commands to run. Not this one. There is no
sfc or DISM equivalent for a failing memory module, and running system
repairs against a hardware fault produces a clean bill of health that tells you nothing. The
useful work here is narrowing down which component is at fault, and all of it is safe.
What you can do yourself
1. Undo every overclock, including XMP
Enter the BIOS/UEFI (usually Del or F2 at power-on) and load optimised or default settings. That disables CPU and memory overclocks together, including XMP and DOCP profiles. Run the machine at stock for a few days.
If the crashes stop, you have your answer: the system is not stable at those settings, even if it was for a year beforehand — components drift as they age. Either leave it at stock or tune back up carefully.
2. Test the memory
Windows has a built-in tester: search for Windows Memory Diagnostic and let it run. It is quick and catches obvious faults, but it misses subtle ones, so a clean pass is not proof. For a real answer, run MemTest86 from a USB stick for several full passes, ideally overnight.
If you have more than one stick, the cheapest test costs nothing: remove all but one, run the machine, and rotate. A fault that follows one stick is that stick.
3. Check temperatures
Watch CPU temperature under load with any monitoring utility. Sustained temperatures at the manufacturer's limit mean cooling, not computing, is your problem — usually dust in the heatsink, a fan that has stopped, or thermal paste that has dried out after a few years.
4. Consider the power supply
Hardest to test without a spare, and worth suspecting if the machine is more than five years old, the crashes come under heavy load, or a power-hungry graphics card was added without upgrading the supply.
5. Read what the crash actually said
Each crash writes a dump file to C:\Windows\Minidump. Those files record which
component reported the error, which is considerably better than guessing. Reading them requires
the right tooling, but the information is sitting there.
What this is not
It is worth being blunt, because the internet is full of the opposite advice. Reinstalling
Windows will not fix WHEA_UNCORRECTABLE_ERROR. Neither will sfc /scannow, a registry
cleaner, or a fresh set of drivers. If a reinstall appears to help, what has usually changed is the
load pattern, and the crashes return. The one thing worth doing on the software side is applying
BIOS and chipset updates from your manufacturer, because those genuinely do change how the hardware
is driven.
⚡ The one-click way: let Rightek Rescue do it
Rescue reads the crash dumps and the hardware error records to tell you which component reported the fault, rather than leaving you to guess between memory, heat and power. When the answer is hardware it says so plainly instead of running repairs that cannot help. On any working PC: download Rescue, put it on a USB (one click makes it bootable), boot the affected PC from it, and choose Fix My PC. The diagnosis is free and it tells you what it found before it changes anything.
FAQ
Can software fix WHEA_UNCORRECTABLE_ERROR?
Generally no. The stop code means a physical component reported an uncorrectable fault, so system-file repairs and reinstalls do not address the cause. The exceptions are BIOS and chipset updates, which change how the hardware is driven, and reverting an unstable overclock — which is a settings change rather than a repair.
I am not overclocking. Why do you keep mentioning it?
Because XMP and DOCP count. Those one-click memory profiles in the BIOS run your RAM above its guaranteed specification, which is an overclock even though the interface presents it as a preset. Many people who say they are not overclocking have XMP enabled.
Are my files at risk?
Your files are not directly at risk from this stop code — it is a processing fault, not a storage one. The indirect risk is that repeated hard crashes can corrupt whatever was mid-write when the machine stopped, so keep your backups current while you are diagnosing it.
It only crashes when I game. Does that narrow it down?
Yes, usefully. Load-dependent crashes point at heat, power delivery, or an overclock that is only unstable when everything is working hard. Crashes at idle are more suggestive of memory. Neither is conclusive, but it tells you which end to start.
Should I just replace the whole PC?
Not before you know which part is at fault. The most common causes here are a BIOS setting, a single memory stick, or a heatsink full of dust — none of which need a new computer. Find the component first.