dwm.exe crashes during playing or doing intense load.

Question

dwm.exe crashes during playing or doing intense load.

Daskywarsboi 0

So, ive had this laptop (Asus tuf gaming a16) for quite a while now. It was awesome for the first few months, then came the dreaded dwm.exe crashes. I dont know when it began but once while I was playing a non intensive game, dwm.exe suddently crashed, making my screen turned black for a few seconds, crashing the game and my entire screen was just black / gray except the taskbar. Its been only getting worse and worse. Ive seen many people with the newer nvidia RTX gpus having problems with these crashes without finding fixes, ive tried everything and I mean EVERYTHING everyone else tried. Also with that im having weird issues where restart manager keeps randomly failing to restart many apps showing in the event viewer and I used to have around 1 - 10 hardware errors a day on reliability moniter from my AMD iGPU for no reason. The crashes occur specifically when doing a tiktok live while playing BeamNG.Drive or sharing my screen on discord and changing my car a few times. Crashes can go from a 2 second black screen coming back with no issues, 3 second black screen crashing the game, or 3 second black screen crashing the game and showing the entire desktop as black except the taskbar working with no apps able to be visible, or a 5 second black screen and comes back as just plain black with a mouse cursor and nothing is visible except the alt + tab menu. From recent chats with my friend, you can clearly see the random errors given in the event viewer

Armourycrate userSessionHelper.exe cannot be restarted, AMDADLXServ.exe cannot be restarted, Discord.exe cannot be restarted, steamwebhelpler.exe cannot be restarted, AmdRadeonSoftware cannot be restarted, AMDRSSrcExt.exe cannot be restarted, NVIDIA Overlay.exe cannot be restarted (5 times), EpicGamesLauncher.exe cannot be restarted, Epic Web Helper cannot be restarted, BeamNG.drive.x64.exe cannot be restarted, AcPowerNotification.exe cannot be restarted, all of them say Application SID does not match Conductor SID. Then information "Ending session 3 started 2026-09-05T20:57:58.41304500Z."

Then application error, vaulting application name dwm.exe, faulting module name nvwqf2umx.dll, vaulting module path: C:\Windows\System32\DriverStore\FileRepisitory\nvami.imf_amd64_734224edc3748001\nvwqf2umx.dll, then information "Event name: APPCRASH, Problem signature: P1: dwm.exe, P2: 10.0.26100.8972, P3: df4b2407, P4: nvwqf2umx.dll, P5: 32.0.16.1656" Then warning from Dwminit "The Desktop Window Manager process has exited. (Process exit code: 0x000000ff, Restart count: 1, Primary display device ID: Nvidia Gefore RTX 5050 Laptop GPU)" 1000222425

then an application error "Faulting application name: NVIDIA Overlay.exe, Faulting Module Name: libcef.dll.

All these restart manager things actually happen multiple times at random times and dont require a crash for it to happen. And the amd igpu was fixed by updating its driver. Before going live on tiktok live studio, playing BeamNG.Drive switches the display to run on the nvidia gpu. I can notice before starting the live, it would sometimes flicker, the contrast of overlays would be too low and like the dark black would just be a light gray and the text barely visible, sometimes inside the studio trying to do something I could see BeamNG in tbe background suddently having a thick contrast and turning like dark for a second or going light for a second and can sometimes happen on the entire screen, and after gpu crashes during the stream, the entire screen keeps flickering black for a half a second at random. I really do not know how to fix this or whats going on, so if someone can help it would be greatly appreciated.

Specs: AMD Ryzen 7 260, RTX 5050 Laptop GPU, 16GB RAM DDR5 1000222419

1000222432

1000222416

0 comments

1 answer

Your answer

Answer 1

The immediate action is to remove the NVIDIA driver completely in Safe Mode and install the ASUS-approved RTX 5050 laptop driver offline, because the recorded dwm.exe fault is inside the NVIDIA graphics module and the accompanying VIDEO_MEMORY_MANAGEMENT_INTERNAL (0x10E) stop code is usually caused by an improperly behaving video driver.^1^

1. Clean-install the NVIDIA driver

This must be a complete removal—not another normal update over the existing package.

Download the graphics driver for the exact ASUS TUF Gaming A16 model/submodel from ASUS Support. Keep the installer locally.
Disconnect the laptop from the internet so Windows Update cannot install another display package during removal.
Start Windows in Safe Mode.
Use Display Driver Uninstaller (DDU) to remove the NVIDIA display driver, then restart. DDU is third-party software; obtain it from its publisher and review downloads carefully.
While still offline, install the downloaded ASUS NVIDIA package and restart again.
If NVIDIA App or GeForce Experience is installed, disable automatic driver updates temporarily.

A closely matching NVIDIA laptop-driver case recommends this offline DDU procedure and specifically prefers the laptop manufacturer’s driver.^2^

If the ASUS package still crashes, repeat the clean removal and test one older ASUS-approved version, if ASUS offers one. A similar Windows 11/NVIDIA case was resolved by moving from the current driver to a slightly older stable version after DDU removal.^3^

Do not change TDR registry timeout values. That can delay detection without repairing the driver or hardware condition.

2. Test without overlays and capture software

Your NVIDIA Overlay crash involving libcef.dll, plus reproduction while using Discord screen sharing or TikTok Live Studio, makes an overlay/capture conflict a separate testable possibility.

Before reproducing the load:

Disable or uninstall NVIDIA Overlay/Instant Replay.
Disable the Discord, Steam, and Xbox Game Bar overlays.
Close Armoury Crate monitoring/OSD components and any GPU tuning or frame-monitoring software.
Run BeamNG.drive by itself first.
Then add Discord sharing.
Finally add TikTok Live Studio.

If BeamNG is stable alone and fails only after a specific capture or overlay component is enabled, update or leave that component disabled.

For a broader isolation test, perform a clean boot:

Search for msconfig and open System Configuration.
Select Services → Hide all Microsoft services → Disable all.
Open Startup → Task Manager, disable enabled startup applications, and restart.
Test the same workload.
If stable, re-enable half the items at a time until the conflicting service or application is identified.^4^

What the brief black screens mean

Windows Timeout Detection and Recovery (TDR) detects a GPU operation that does not finish within the permitted time and resets the graphics stack. The visible result can be a short flicker or black screen; applications that do not recover their graphics device can remain black or crash afterward.^5^

During a black screen, press Windows + Ctrl + Shift + B once. This requests a graphics-driver reset and may restore the desktop, but it is only a recovery shortcut, not the underlying repair.^6^

Restart Manager messages

“Application SID does not match Conductor SID” events are from Restart Manager sessions used by installers and updates to close and restart applications holding files in use.^7^ They do not, by themselves, identify the cause of the DWM/GPU crash. Concentrate on the dwm.exe faulting module, 0x10E stop code, and NVIDIA Overlay failure.

When to seek ASUS hardware service

Escalate to ASUS warranty/service if crashes continue after:

an offline clean installation of an ASUS-approved NVIDIA driver;
overlays and capture applications are removed from the test;
the system is returned to stock settings with no overclock or undervolt; and
ASUS hardware diagnostics still report errors.

Microsoft lists inadequate cooling or power, unstable component settings, memory problems, and defective parts among possible causes when display timeouts persist after driver remediation.^8^ Provide ASUS with the 0x10E photo, Reliability Monitor history, Event Viewer export, and any files under C:\Windows\Minidump or live-kernel reports. Repeated failures under a clean OEM driver and clean-boot workload are the threshold for investigating the GPU, VRAM, motherboard, cooling, or power delivery rather than reinstalling applications again.

References