VIDEO_TDR_FAILURE: Causes, diagnosis and real solutions

When your computer crashes in the middle of a task and the dreaded blue screen appears with VIDEO_TDR_FAILURE, it's normal to think that everything has suddenly broken; in reality, the system has tried to recover the GPU after a crash and failed. This stop code It is closely linked to the graphics driver and the timeout, detection, and recovery (TDR) mechanism., so understanding how it works is half the solution.

In the following lines you will find a complete guide that combines the official technical explanation with field procedures and tricks that work in real-life scenarios. The goal is for you to be able to diagnose, correct, and prevent this error Whether you are an advanced WinDbg user or prefer simple steps from Windows.

What is VIDEO_TDR_FAILURE (0x00000116) and why does it appear?

The VIDEO_TDR_FAILURE error corresponds to the bugcheck 0x00000116 and indicates that Windows attempted to reset the video driver and GPU after a timeout and was unsuccessful. The Windows graphics subsystem monitors the GPU with TDR: If the GPU becomes stuck and unresponsive, it attempts to restart the driver to recover the desktop.

This TDR mechanism has a default timer of about 2 seconds: if the graphics task does not finish in time or the driver restart is not completed, the error check is triggered.

When the recovery process is successful, you will see the message “The display driver stopped responding and has recovered.” When it fails to recover, the system stops execution and displays the blue screen with VIDEO_TDR_FAILURE.

Bugcheck parameters and how to interpret them

In addition to the 0x116 code, the blue screen and memory dumps include four parameters that help debug the exact source of the VIDEO_TDR_FAILURE error. Knowing what each parameter means guides the diagnosis and speeds up the solution.:

Parameter	Description
1	Pointer to the internal TDR recovery context (TDR_RECOVERY_CONTEXT), when present.
2	Address within the responsible driver module (used to identify the driver involved).
3	NTSTATUS code of the last failed operation, if available.
4	Context-dependent internal data, available in certain scenarios.

In real traces, a .sys file of the graphics driver is usually involved: for example, nvlddmkm.sys (NVIDIA), igdkmd64.sys (Intel) or atikmdag.sys/atikmpag.sys (AMD/ATI). That clue is key: it points to the provider and driver path that is causing the crash..

Why it happens: typical causes in the real world

The most common pattern is that the GPU remains busy for too long under a heavy load (gaming, CAD, decoding, or rendering) and doesn't respond in a timely manner. From there, The most common triggers are:

• Outdated or faulty drivers:
• Conflicts between integrated and dedicated GPU.
• Overheating

Beyond software, hardware also plays a role when it comes to the VIDEO_TDR_FAILURE error: aggressively timed memory modules, overclocked motherboards, insufficient power supply/power, degraded VRAM/GPU, or poor cooling in compact laptops. If the equipment is running very low on temperature or power, the TDR jumps earlier and more frequently..

There are also environmental factors: too many background processes competing for resources, Recent updates installed incorrectly or graphics packages (DirectX, OpenGL, Vulkan) out of date. A system with high background load and misaligned graphics components is a perfect candidate for this bugcheck.

Diagnostics with WinDbg and logs: how to identify the culprit

If you handle memory dumps (minidumps), WinDbg will allow you to confirm the responsible driver and the immediate cause of the VIDEO_TDR_FAILURE error. The header command is !analyze -v, which generates the bugcheck summary, the module involved and the latest NTSTATUS.

kd> !analyze -v
VIDEO_TDR_FAILURE (116)
Attempt to reset the display driver and recover from timeout failed.
Arguments:
Arg1: <puntero TDR_RECOVERY_CONTEXT>
Arg2: <dirección en el módulo responsable>
Arg3: <código NTSTATUS último>
Arg4: <datos de contexto>

Starting with parameter 2, you can list the module with lmvm and check the timestamp, path, and size. If nvlddmkm.sys or equivalent appears, you already have your prime suspect..

kd> lmvm nvlddmkm
start end module name
... nvlddmkm.sys
Image path: \SystemRoot\system32\DRIVERS\nvlddmkm.sys
Timestamp: <fecha>

With parameter 1, when present, it is possible to examine the TDR_RECOVERY_CONTEXT and see the reason for the timeout promoted to adapter restart. This helps differentiate a simple pipeline hang from a repeated reset failure..

kd> dt dxgkrnl!_TDR_RECOVERY_CONTEXT <puntero>
+0x010 TimeoutReason : TdrEngineTimeoutPromotedToAdapterReset
...

Finally, the stack trace with k/kb/kp shows the code path culminating in KeBugCheckEx, with dxgkrnl and dxgmms1 functions typical of the TDR flow. A “clean” stack towards TdrResetFromTimeout confirms that it is a genuine TDR, not a general kernel issue.

kd> k
# Child-SP RetAddr  Call Site
00 ... nt!KeBugCheckEx
01 ... dxgkrnl!TdrBugcheckOnTimeout
02 ... dxgkrnl!ADAPTER_RENDER::Reset
03 ... dxgkrnl!DXGADAPTER::Reset
04 ... dxgkrnl!TdrResetFromTimeout
05 ... dxgmms1!VidSchiRecoverFromTDR
...

Solution steps: from basic to advanced

Before opening the registry editor or running debuggers, focus on the metrics with the best effort/impact ratio. Start with drivers, then move on to cleaning/reinstalling, and if that's still the case, tackle advanced tweaks and hardware verification..

1) Correctly update the graphics driver and chipset

Outdated or buggy graphics drivers are the most common cause of TDRs. Go to Device Manager, expand Display Adapters, right-click your GPU, and choose "Update Driver." Better yet: download the recommended version for your model from the official NVIDIA/Intel/AMD website..

If you use NVIDIA with professional software (CAD, editing, 3D), try the Studio drivers instead of the Game Ready drivers. On laptops with Intel iGPU + NVIDIA dGPU, also install the latest Intel package and chipset from the computer manufacturer.

2) Clean Driver Reinstallation (DDU)

When "over-the-counter" updates aren't enough, it's a good idea to clean up remnants of the previous driver. The Display Driver Uninstaller (DDU) tool removes driver components, services, and registry entries. It is recommended to run it in Safe Mode and then install the newly downloaded driver..

Typical sequence: safe mode, DDU “Clean and Reboot”, start Windows normally, and then install the driver downloaded from the manufacturer’s website. This process avoids version conflicts and corrupted files that the standard installer does not always resolve..

3) Try a previous stable version

Occasionally, the latest version introduces a regression. In that case, use the "Previous Driver" (if available) or manually install a previous stable version. With NVIDIA GPUs, many TDR issues in CAD disappear when rolling back to a previous Studio driver..

For manual installation, use “Browse my computer for driver software” and “Let me pick from a list of available drivers,” selecting the previous build for your GPU. Reboot when finished to validate the change.

4) Repair system files with SFC

If the problem occurred after an update or a failed installation, run System File Checker. Open CMD as administrator and run sfc /scannow. Windows will check protected binaries and replace damaged ones automatically..

sfc /scannow

When finished, reboot and try your usual graphics load again to see if the TDR has disappeared. If SFC repairs files, it is a good sign that there was system corruption influencing the graphics stack..

5) Windows Startup Repair

If TDR appears repeatedly after logging in or when opening the same apps, run Startup Repair by going to Settings > Update & security > Recovery > Advanced startup > Troubleshoot > Advanced options > Startup Repair. Windows will attempt to fix services and startup components that are not loading correctly..

After the repair, test the equipment with the same usage pattern that triggered the failure. If it stops playing, it was a partial boot issue or corrupted dependencies.

6) Safe mode and driver reinstallation

Safe Mode loads the minimal drivers and allows for uninstallation/reinstallation without interruption. Select "Safe Boot" in msconfig or access it from Advanced Startup, and once there, uninstall the adapter in Device Manager. Then install the fresh driver downloaded from the manufacturer.

On computers with two GPUs, uninstall and reinstall them separately: first the integrated one (Intel), then the dedicated one (NVIDIA/AMD). This prevents Windows from re-injecting an inappropriate generic in between..

7) Specific driver file repair (AMD/NVIDIA)

On AMD/ATI GPUs, there are cases where regenerating atikmdag.sys or atikmpag.sys resolves the issue. Rename the C:\Windows\System32\drivers file by adding .old and expanding it from the manufacturer's package. The expand command creates the new .sys file which you then copy to the drivers folder..

expand.exe atikmdag.sy_ atikmdag.sys
expand -r atikmdag.sy_ atikmdag.sys

For NVIDIA, the idea is analogous to nvlddmkm.sys: after uninstalling, rename nvlddmkm.sys to .old, expand NVLDDMKM.Sy_ and copy the generated .sys to System32\DRIVERS. This replacement removes corrupt files that survive surface installs.

Expand.exe NVLDDMKM.Sy_ NVLDDMKM.Sys

8) Temporarily disable the Intel iGPU if there are conflicts

If you are using a dedicated GPU as your primary (e.g., NVIDIA), disable Intel HD/UHD in Device Manager to rule out problematic switching. On some laptops, autoswitching triggers TDRs when moving apps between iGPU/dGPU.

Even more refined: in the NVIDIA Control Panel, assign the specific dGPU to demanding applications (acad.exe, games), and leave the rest on automatic. This way you minimize adapter changes when opening/closing windows and avoid graphical context resets..

9) Check hardware and thermals

Check temperatures with a reliable tool: if the GPU exceeds thresholds and throttles, TDRs will appear. Clean the dust, renew thermal paste if necessary, and ensure good ventilation. On laptops, a cooling pad can make all the difference during long sessions..

Validates RAM (Windows Memory Diagnostic Tool) and system stability without overclocking. Aggressive memory settings or voltage limits cause instability under graphics load. CPU/RAM/VRM stability is just as important as the GPU itself..

10) Advanced TDR Settings (with caution)

For heavy workloads that legitimately take a bit longer, raising TdrDelay may help. Path: HKLM\System\CurrentControlSet\Control\GraphicsDrivers. Create/set the 32-bit DWORD TdrDelay to 8–10 and reboot. This doesn't fix bad drivers, but it gives more leeway before declaring a timeout..

Only use it if you've already updated/cleaned drivers and checked thermals; if the driver is the problem, extending the time only postpones the BSOD. If after increasing TdrDelay the crashes continue, go back to 2 and fix the root.

Quick checklist before considering the problem solved

To be sure that you have resolved the VIDEO_TDR_FAILURE issue, you need to perform the following checks:

• Verify that repeated TDR events no longer appear in the Event Viewer during a long session with your usual load. If you spend 30–60 minutes without TDR in real-world conditions, the fix is ​​robust..
• Run a moderate stress test (graphical benchmark or complex scene in your app) by monitoring temperatures. A stable graphics card maintains sustained clocks without throttling or triggering errors..
• Make a copy of the known-good driver (and note the version) so you can quickly roll back if a future update reintroduces the problem. Prevention saves hours when it's time to update again..

Although the VIDEO_TDR_FAILURE error is scary, it can usually be fixed by applying order and method: correct driver, clean installation, thermals under control and, if necessary, fine-tuning the TDR. With this combination, creative tasks, games, and CAD work again without crashes..