Windows 11 Help Diagnosing Error Code 117. No full crash, TDR Reset.

[MurayTheFloox]

As the title says. The PC hung, went black, and came back saying the a problem caused hardware to stop working. Was only playing modded minecraft as well. Nothing on the system changed and is only a fresh instal of 24H2 Windows. I went and did an SFC/Scannow and returned no violations. Did a ram test, nothing came back. The only thing I have done differently is had XMP on for my DDR5 Ram to run at its full 6000. My specs are as listed

Corsair RMx Series RM1000x 80 Plus Gold ATX power supply
Intel Core I9-12900K
MSI MPG Z790 ATX DDR5
Corsair Vengeance DDR5 32 x2 16GB sticks 6000 XMP (Changed back to 4200 and disabled XMP now after this.)
MSI RTX 5080 Ventus OC

IF someone could help me, ill provide the files but theres like 7 of them. There isnt a minidump because the PC did not fully crash, the driver did restart itself and it came back.

Temps are fine, everything stays within 50 - 55 degrees C and maybe the CPU will hit 70 sometimes but its liquid cooled. Forgive my ignorance. I have never had this happen before ever. If anything, Ive had full crashed but nothing like this. Only time this happens is when I restart my GPU drivers manually.

 

[ChatGPT]

Thanks — I read your post about the Error Code 117 / driver reset. I don’t see any attachments yet (the files you mentioned didn’t come through) — please attach them when you can and I’ll analyze them. I reviewed your description here: .
Short summary of likely causes

• Error 117 in Reliability Monitor is the “hardware stopped working / device was reset and recovered” message that commonly matches a GPU driver TDR (Timeout Detection & Recovery). Causes range from driver corruption, GPU/driver/firmware bugs, power/cable problems, to memory/BIOS instability (XMP on DDR5 can trigger subtle issues). Forum troubleshooting commonly starts with a clean driver reinstall (DDU) and checking BIOS/memory settings.

What I need to see (please upload these files)

1. DxDiag text: Run Start → dxdiag → “Save All Information…” → attach the resulting dxdiag.txt.
2. MSInfo report: Run msinfo32 → File → Export → save msinfo.nfo (or .txt) → attach.
3. Event Viewer System log around the time of the failure: Event Viewer → Windows Logs → System → find the time of the incident, filter for Errors/Warnings and save those events (or export filtered log as .evtx). Look specifically for entries sourced from Display, dxgkrnl, or nvlddmkm.
4. Reliability Monitor details for the event (type “reliability” in Start → View reliability history → click the red X on the event and “View technical details” → copy/paste or screenshot).
5. Any minidumps: check C:\Windows\Minidump\ (attach any .dmp files) and C:\Windows\MEMORY.DMP if present.
6. Windows Error Reporting folder (if present): C:\ProgramData\Microsoft\Windows\WER\ReportArchive\ (zip relevant folder(s) and attach).
7. Any GPU logs or MSI Afterburner logs if you use Afterburner/OC software.

If you need quick copy/paste instructions for any of the above I’ll paste them — tell me which you’d prefer.
Immediate troubleshooting steps to try now (in order)

1. Keep XMP disabled (you said you already reverted to 4200 — good). Run a test session with XMP off and no other overclocks. If the problem goes away, memory timing is the likely trigger.
2. Clean the GPU driver with Display Driver Uninstaller (DDU) in Safe Mode, then do a clean install of the NVIDIA driver (WHQL). Many TDR cases are resolved by a DDU + clean install. If you want step-by-step for DDU I’ll provide it. (This is standard forum advice.
3. Temporarily increase the TDR timeout (only as a diagnostic step). Run an elevated Command Prompt and paste:
   reg add "HKLM\SYSTEM\CurrentControlSet\Control\GraphicsDrivers" /v TdrDelay /t REG_DWORD /d 8 /f
   Reboot and test (this gives the driver more time before Windows resets it — it’s a diagnostic/workaround, not a fix).
4. Check power/cables: ensure the GPU 8‑pin/12VHPWR cable(s) are fully seated and that you’re using the PSU cable set as recommended (swap to a different PCIe cable from the PSU if available). A failing / loose cable can cause driver resets.
5. Test with a different workload: try a GPU stress app (Unigine/3DMark or even a sustained Minecraft run) and watch for reproducibility; monitor temperatures and the System log time stamps. Don’t run extreme stress tests if you’re worried about the card.
6. Update motherboard BIOS (if not up-to-date) and chipset drivers — Z790 + DDR5 benefits from the latest BIOS for memory compatibility. Then retest with XMP disabled first, and only enable XMP later after testing stability.
7. If the issue still occurs after the above, try a clean Windows boot (disable 3rd‑party overlays, MSI/Afterburner, mods) or try vanilla Minecraft to rule out mod/plugin issues.

What I’ll do when you upload the files

• Correlate the Event Viewer / Reliability timestamp with any driver messages (Display/dxgkrnl/nvlddmkm).
• Check driver versions, BIOS, Windows build from msinfo/dxdiag.
• If minidumps exist I’ll analyze them for the module that caused the reset and tell you if it points to driver, GPU hardware, or a third-party driver.
• Give a prioritized, concrete next-step plan based on the findings (e.g., rollback to specific driver, test hardware, or RMA advice if hardware fault is likely).

A couple of quick questions

• Did this happen only in modded Minecraft or in other games/apps too?
• Do you use MSI Afterburner, any manual GPU OC, or any 3rd-party kernel-level software (Vanguard, etc.?
• What NVIDIA driver version are you currently on, and what is your motherboard BIOS version (msinfo will show this)?

Warnings / notes

• Before major changes (BIOS updates, driver verifier), make a full backup or at least a restore point. Driver Verifier can cause BSODs if it finds bad drivers — only run it if you’re comfortable recovering from boot issues.
• Increasing TdrDelay hides symptoms; it’s for diagnosis only.

If you upload the dxdiag, msinfo, System log and any .dmp/WER files here I’ll start the analysis and give the exact next steps.