DXGI Error Device Hung | Common Fixes

If a game or 3D application suddenly freezes and then crashes with DXGI_ERROR_DEVICE_HUNG, your graphics card didn’t fail — Windows simply gave up waiting for it to respond and reset the display driver. This guide walks through why it happens and nine fixes that have worked consistently for PC gamers on Windows 10 and Windows 11.

What DXGI_ERROR_DEVICE_HUNG Actually Means

DXGI stands for DirectX Graphics Infrastructure, the layer that sits between your games and your graphics driver. When Windows sends a rendering command and the GPU doesn’t reply within its expected time window, the operating system assumes the driver has locked up. It then tears down and restarts the display driver — this is called a TDR (Timeout Detection and Recovery) event — and your game receives the DXGI_ERROR_DEVICE_HUNG error as it loses its connection to the GPU.

It’s worth knowing this is different from its sibling error, DXGI_ERROR_DEVICE_REMOVED, which means Windows believes the GPU has disconnected or failed outright rather than just being slow to respond.

Games and apps where this shows up most

• Warzone, Fortnite, and other high-refresh competitive shooters
• Flight simulators such as Prepar3D and Microsoft Flight Simulator
• Unreal Engine projects and the Unreal Editor itself
• Story-driven titles with heavy shader compilation, such as Star Wars Jedi: Survivor

Common Causes

There’s rarely a single cause. Most cases come down to one of these:

• Outdated or corrupted GPU driver — the single most common trigger
• Unstable overclocking on the GPU, CPU, or RAM
• Monitor refresh-rate mismatch when a game tries to control the display directly
• Overheating or a GPU under sustained heavy load
• Corrupted game files or an outdated DirectX runtime
• Failing or unstable system memory

9 Fixes, Step by Step

1. Update your graphics driver

Go to your GPU manufacturer’s official site (NVIDIA, AMD, or Intel), download the latest driver for your exact model, and choose a clean install if the option is available. This resolves the issue in the majority of reported cases.

2. Turn off all overclocking

Reset your GPU to factory clock speeds using the same tool you used to overclock it (MSI Afterburner, AMD software, or similar), and disable any CPU or RAM overclock profile (XMP/DOCP) in BIOS if you suspect memory instability.

3. Update DirectX and Windows

Install the latest DirectX End-User Runtime and make sure Windows Update has installed all pending updates, including optional driver updates.

4. Cap your frame rate or fix refresh-rate sync

Cap the game to a fixed FPS (60 or your monitor’s native refresh rate) in its graphics settings, and switch from borderless windowed to exclusive fullscreen mode, or the reverse, if the crash coincides with alt-tabbing.

5. Clear the DirectX shader cache

Go to Settings > System > Storage > Temporary files, select DirectX Shader Cache, and remove it. A corrupted cache can cause rendering hiccups that look identical to a hung device.

6. Run the game with administrator rights

Right-click the game’s .exe file, open Properties > Compatibility, and enable “Run this program as an administrator.” Some titles need elevated access to talk to the GPU driver correctly.

7. Force the correct GPU on laptops

On laptops with both integrated and dedicated graphics, open your GPU control panel’s 3D settings and set the game to always use the dedicated GPU, since an accidental switch to integrated graphics mid-session can trigger the timeout.

8. Increase the TDR delay (advanced users)

This is a registry-level change that gives the GPU more time before Windows declares it hung. It’s a useful diagnostic step but not a permanent fix — treat a longer delay as confirmation the GPU is genuinely overloaded, not just slow.

9. Reinstall the game and test your RAM

Uninstall and reinstall the affected title to rule out corrupted files, and if the crash still persists across multiple games, run a memory diagnostic tool such as Windows Memory Diagnostic or MemTest86 overnight to check for failing RAM.

Fix Difficulty & Success Rate

A quick-reference table to help you prioritize which fix to try first.

Preventing It From Coming Back

• Keep GPU drivers current, but avoid jumping to brand-new “day one” drivers before checking user reports
• Monitor GPU temperatures during long sessions and improve case airflow if temps consistently exceed manufacturer limits
• Avoid running overclocks you haven’t stress-tested with dedicated stability software
• Close background overlays (recording software, chat overlays) that can conflict with DirectX rendering

Frequently Asked Questions

Is DXGI_ERROR_DEVICE_HUNG a sign of GPU hardware failure?

Not usually. In most reported cases it traces back to a driver or software-level timeout rather than permanent hardware damage. Hardware failure becomes more likely only if the error persists after drivers, overclocking, and system files have all been ruled out.

What’s the difference between DEVICE_HUNG and DEVICE_REMOVED?

DEVICE_HUNG means the GPU stopped responding in time and Windows reset it. DEVICE_REMOVED means Windows believes the GPU disconnected or failed entirely, which more often points to a hardware or connection issue.

Should I disable TDR completely?

It’s not recommended as a permanent setting, since TDR exists to recover your system from a genuinely frozen driver. Increasing the delay for diagnostic purposes is safer than disabling the feature outright.

For the official Microsoft explanation of how Timeout Detection and Recovery works at the driver level, see the Microsoft Learn documentation on TDR.

If you’re also dealing with driver-related crashes on other titles, our guide on fixing common Windows app and game errors covers several related troubleshooting steps.