Turkey (Turkish)
Germany (German)
Worldwide (English)
Microsoft defines 0x124 as a fatal hardware error reported through WHEA.
Halkali merkez neighborhood fatih st ozgur apt no 46 , Kucukcekmece , Istanbul , 34303 , TR
Diagnose 0x124 WHEA_UNCORRECTABLE_ERROR across CPU, RAM, PCIe, thermals, power, BIOS and WHEA event records.
Microsoft defines 0x124 as a fatal hardware error reported through WHEA.
Microsoft specifically mentions heat, defective hardware, memory, processor and overclocking.
The WHEA error record can identify machine-check, PCIe, generic hardware or other error-source classes.
WHEA is the Windows Hardware Error Architecture. With 0x124, Windows reports a fatal hardware error record from the platform. The source can be a CPU machine check, PCIe, platform or another hardware error class.
Reinstalling the operating system does not remove real power, thermal, memory, CPU or PCIe instability. Identify the failure pattern and WHEA source first.
Microsoft explicitly lists overclocking as a possible cause. During diagnosis temporarily disable XMP/EXPO, manual memory timings, CPU PBO/Curve Optimizer, undervolting and GPU overclocking. The goal is to reduce variables.
If using Load Optimized Defaults, first record VMD/RAID, Secure Boot and boot-order settings. Factory defaults are not always the correct storage/boot profile for an existing installation.
WHEA-Logger events can describe the error source. Text such as Cache Hierarchy Error, Bus/Interconnect Error, Memory Error or PCI Express and processor/APIC data can guide diagnosis. A single Event ID does not universally prove one physical part is bad.
Get-WinEvent -FilterHashtable @{LogName='System'; ProviderName='Microsoft-Windows-WHEA-Logger'} -MaxEvents 40 | Format-List TimeCreated,Id,LevelDisplayName,MessageIf the crash occurs only during CPU rendering, AVX or simultaneous CPU+GPU load, thermals, PSU/VRM and tuning become stronger suspects. Idle-only failures can shift attention toward voltage tuning, C-states, firmware and platform behavior.
Repeating the same workload at stock settings is valuable. If the error disappears with tuning disabled, hardware may not be physically defective; stability margin may simply be insufficient.
A clean Windows Memory Diagnostic result does not eliminate all intermittent memory instability. Test with XMP/EXPO disabled, isolate DIMMs/slots and consider firmware memory-training behavior.
On ECC servers, inspect iDRAC/iLO/IPMI/BMC memory events in addition to Windows logs. Rising corrected-ECC counts or repeated errors on one DIMM slot can provide stronger evidence.
mdsched.exeIf WHEA points toward PCI Express, do not blame only the GPU. NVMe, NIC, HBA, riser, slot, power and BIOS link settings share the same interconnect. Recent PCIe additions and slot changes are useful clues.
BMCs such as iDRAC, iLO and IPMI can record CPU Machine Check, ECC, PCIe AER, fan, PSU and thermal events before Windows crashes. Do not rely solely on Event Viewer when diagnosing 0x124 on servers.
| Pattern | First suspects |
|---|---|
| Only under heavy load | Thermals, PSU/VRM, tuning |
| Starts with XMP/EXPO | RAM/IMC stability |
| WHEA PCIe source | GPU/NVMe/NIC/slot/riser |
Do not use destructive boot, BitLocker, firmware, partition or storage commands without backups and a rollback plan. Never copy example drive letters without verification.
No. Depending on the WHEA source, CPU, RAM, PCIe, power, thermals or another platform component can be responsible.
Not if the root cause is hardware instability. Diagnose at stock settings and use hardware records first.
For boot, storage, RDP or update problems on a Windows VPS or physical server, send the exact error code, Windows version and console screenshot to technical support.