System Thread Exception Not Handled

Home / ntoskrnl.exe

SYSTEM_THREAD_EXCEPTION_NOT_HANDLED — ntoskrnl.exe

Dumps that say ntoskrnl.exe mean the crash was observed inside the Windows kernel binary — Microsoft notes this can still be caused by a driver underneath. Prove the real module before replacing hardware.

First cut If !analyze clearly names nvlddmkm.sys or etd.sys, leave this page. ntoskrnl-only blame is a “keep digging” signal, not an automatic CPU failure.

Step 1 — Treat ntoskrnl as a clue, not the verdict

Microsoft’s 0x7E resolution text: when WinDbg reports “Probably caused by : ntkrnlmp.exe,” continue with thread and stack analysis. Consumer steps still start with Event Viewer, the named driver (if any), and vendor updates.

Step 2 — Re-read the dump carefully

Follow minidump: run !analyze -v, note IMAGE_NAME, STACK_TEXT, and exception code (see bugcheck codes). Look for third-party .sys frames below ntoskrnl.

WinDbg !analyze output showing Probably caused by ntoskrnl.exe with a red arrow pointing to a third-party .sys on the stack
Scroll past ntoskrnl to any third-party driver frames before buying parts.

Step 3 — Driver and update passes

  1. Enter Safe Mode and roll back recent storage, chipset, antivirus filter, and VPN drivers.
  2. If timing matches patches, use after update.
  3. Uninstall brand-new hardware drivers added the same week.

Step 4 — Memory and system file checks

Branches

Stack shows GPU driver nvlddmkm
Stack shows etd.sys etd.sys
Boot loop boot loop
During clean install during install

FAQ

Does ntoskrnl.exe mean Windows is corrupted?

Sometimes system files are involved — run SFC/DISM — but many dumps still trace to third-party drivers. Confirm with !analyze.

Should I replace the CPU?

Not from the module name alone. Exhaust drivers, updates, memory tests, and OEM diagnostics first.

Official and primary sources

  1. Microsoft Learn — Bug Check 0x7E
  2. Microsoft Learn — Analyze dumps with WinDbg
  3. Microsoft Support — Stop code troubleshooting

Symptom map