Subject: Repeated BSODs 0x133 & 0x113 with subtype 19 on Win11 24H2 + RTX 5070 Ti Laptop — nvlddmkm.sys fault
Body:
Hi, I’m encountering repeated Blue Screen crashes on a Windows 11 (24H2) laptop with an NVIDIA RTX 5070 Ti (laptop version).
Below is a summary of my crash history and what I’ve analyzed.
Crash summary
# Inserted Dump / Time BugCheck & Subtype Key Module / Stack Activity at Crash Hypothesis
| 1 | Early dump | 0x133 (DPC_WATCHDOG) | nvlddmkm in ISR / DPC | Idle / normal use | Driver / interrupt path stuck too long |
| 2 | Mid | 0x133 → then 0x113 | nvlddmkm / graphics path | After long uptime | Timing / power / driver-state mismatch |
| 3 | Mid | 0x113 | dxgkrnl failure | Moderate usage / idle | Graphics subsystem recovery / TDR failure |
| 4 | Mid | 0x113 | dxgkrnl | Wake / runtime | Driver recovery failure on wake |
| 5 | Mid | 0x113 subtype = 19 (UNEXPECTED_DEFERRED_DESTRUCTION) | nvlddmkm + dxgkrnl PnP paths | Extended runtime / possible wake / switching | Driver cleaning up resources after removal incorrectly |
| 6 | Latest | 0x113 subtype = 19 | Same pattern | Light usage: WeChat (video), Chrome, Excel, OneNote | Device removal / context loss during operation, driver mis-cleanup |
Additional log clues
In GPU / Chrome logs, there’s a message:
Renderer11.cpp … The D3D11 device was removed, HRESULT: 0x887A0005
(i.e. the GPU / driver lost its rendering context mid-operation)
Also “context lost via ARB/EXT_robustness” in OpenGL paths
GPU process crashes and restarts under Chrome
Interpretation / suspected bug path
The driver often attempts deferred destruction / cleanup of GPU resources, contexts, or buffers (especially after device removal / wake / PnP transitions).
The subtype 19 (“UNEXPECTED_DEFERRED_DESTRUCTION”) strongly suggests that these cleanup routines are being invoked when they shouldn’t be (or without proper synchronization) and dxgkrnl deems it a violation and crashes.
The “device removed / context lost” logs indicate that the GPU / driver is dropping the rendering path mid-operation, perhaps due to power / wake / bus reinitialization.
The progression from 0x133 (DPC / ISR delays) to more frequent 0x113 crashes suggests the driver’s complexity in handling transitions is unstable under newer Windows / power modes.
What I’ve tried / will try
Clean driver uninstall via DDU + reinstall stable / older versions
Disable HAGS / disable hardware acceleration in apps
Increase TDR delay in registry
Turn off fast startup / mixed sleep / PCIe link power saving / ASPM
Update BIOS, chipset, AMD and iGPU drivers
Use Driver Verifier to catch driver faults
Would love help identifying:
Which exact driver version (NVIDIA) handles deferred destruction properly on Win11 24H2 for RTX 50xx laptops
Whether this is a known bug in nvlddmkm / dxgkrnl PnP cleanup paths
Any patches or hotfixes from NVIDIA or Microsoft to address subtype 19 or device removal crashes
