How to Fix DXGI_ERROR_DEVICE_HUNG & GetDeviceRemovedReason Game Crash in Windows 11 (2026 Guide)

You are in the middle of a high-intensity session in Black Myth: Wukong, Call of Duty, Cyberpunk 2077, or an Unreal Engine 5 title. Suddenly, the frame freezes for two seconds, your audio loops, and the game crashes straight to the desktop with a dialog box stating: DirectX function "GetDeviceRemovedReason" failed with DXGI_ERROR_DEVICE_HUNG ("The GPU will not respond to more commands"). Even on top-tier modern graphics cards such as the RTX 40-series and RTX 50-series, this persistent error continues to disrupt PC gaming on Windows 11 24H2. Following our deep-dive diagnostic for EXCEPTION_ACCESS_VIOLATION (0xC0000005) crashes, this guide systematically eliminates the root hardware timeouts and graphics subsystem desynchronizations causing device hung failures.

Why DXGI_ERROR_DEVICE_HUNG Occurs in Windows 11

The DirectX Graphics Infrastructure (DXGI) acts as the bridge between graphics applications and the underlying operating system kernel. When a title crashes with DXGI_ERROR_DEVICE_HUNG, the failure originates from one of four architectural points:

• TDR (Timeout Detection and Recovery) Threshold Exceeded: By default, Windows allocates exactly 2,000 milliseconds (2 seconds) for a GPU command buffer to complete execution. If a complex shader calculation or Ray Tracing BVH traversal takes 2.01 seconds, the Windows kernel assumes hardware failure and forcibly resets the display driver.
• Microsecond Transient Voltage Dropouts: Modern GPUs feature aggressive boost algorithms that shift clock speeds by hundreds of megahertz within milliseconds. When transitioning from high load to low load, temporary voltage sags cause the GPU core to lose synchronization with the memory controller.
• Corrupted DirectX Shader Pipelines: Game updates and driver installations frequently leave fragmented compiled binary blobs in the local DirectX shader cache, causing immediate execution exceptions when shaders are recalled.
• PCIe Bus Power State De-sync: Aggressive PCIe Link State Power Management (ASPM) can drop the motherboard expansion slot into low-power states during loading screens, causing the GPU to disconnect momentarily from the system bus.

Method 1: Expand Windows TDR Registry Timeout Limits

The most reliable fix for compute-heavy modern engines is increasing the TDR delay threshold from the restrictive 2-second default to 8 seconds. This gives demanding shaders sufficient execution headroom without triggering kernel driver resets:

1. Press Win + R, type regedit, and press Enter to open the Registry Editor with administrative privileges.
2. Navigate to the following key path:

   HKEY_LOCAL_MACHINE\SYSTEM\CurrentControlSet\Control\GraphicsDrivers

3. Right-click in the right pane, select New > DWORD (32-bit) Value, and name it exactly TdrDelay.
4. Double-click TdrDelay, set the Base to Decimal, and enter value 8.
5. Create a second new DWORD (32-bit) named TdrDdiDelay, set its Base to Decimal, and enter value 8 (this controls the software driver timeout limit).
6. Restart your computer for the kernel registry modifications to take effect.

Method 2: Cleanly Purge Corrupted DirectX Shader Caches

Corrupted precompiled shader binaries trigger immediate device hung errors when Unreal Engine 5 or Frostbite attempts to load pre-cached textures. Clean both the driver-level and system-level shader repositories:

1. Close all running game clients, Steam, Epic Games Launcher, and web browsers.
2. Press Win + R, paste the following path, and press Enter:

   %localappdata%\NVIDIA

   (For AMD Radeon users, navigate to %localappdata%\AMD).
3. Delete all folders named DXCache and GLCache. If certain active files are locked, click Skip.
4. Press Win + R, type cleanmgr, select drive C:, check DirectX Shader Cache, and click OK.

Method 3: Stabilize Voltage & GPU Boost Clocks in MSI Afterburner

Factory-overclocked graphics cards from third-party manufacturers (ASUS ROG, MSI Suprim, Gigabyte Aorus) often push boost clocks past the silicon stability margin at stock voltage tables. Locking the core clock curve eliminates the micro-voltage spikes that crash the DirectX engine:

• Open MSI Afterburner and press Ctrl + F to bring up the Voltage/Frequency Curve Editor.
• Locate the 975mV to 1000mV point on the voltage curve.
• Select your target core clock (e.g., 2600 MHz on RTX 40/50 series) and press L to lock the voltage-frequency line. A yellow line will appear, ensuring the GPU core never spikes into unstable voltage zones during gameplay.
• Apply the profile and ensure it activates on Windows startup.

Method 4: Configure PCIe Link State Power Management

When Windows 11 drops the PCIe bus into Active State Power Management (ASPM) L1 or L0s states to conserve electricity, latency spikes can disconnect the GPU driver during scene transitions:

1. Press Win + R, type powercfg.cpl, and press Enter.
2. Click Change plan settings next to your active power profile, then select Change advanced power settings.
3. Expand PCI Express > Link State Power Management.
4. Set the setting to Off for both On Battery and Plugged In.
5. Click Apply and OK.

Technical Comparison: Common DirectX Crash Triggers by Architecture

Different GPU architectures respond differently to DirectX pipeline stresses. The table below details verified failure triggers and the exact countermeasure for each silicon generation:

Frequently Asked Questions (Hardware & GPU Troubleshooting)