Skip to content
WhySoGeek.
Gaming

Fix DirectX 12 Device Hung Crashes on PC 2026

DXGI_ERROR_DEVICE_HUNG crashes a game to desktop when Windows loses the GPU. Here are the proven fixes, from drivers to frame caps.

Sam Carter 8 min read
Cover image for Fix DirectX 12 Device Hung Crashes on PC 2026
Photo: Schill / flickr (BY-NC 2.0)

You are deep into a session when the screen freezes for a second and the game vanishes to the desktop with a message about DirectX function GetDeviceRemovedReason failing, often with a code like 0x887a0006 or the words DEVICE_HUNG, DEVICE_REMOVED, or DEVICE_RESET. It is one of the most common and most frustrating PC gaming crashes, and it spiked again with several big 2026 releases.

The error means Windows momentarily lost communication with your graphics card.

Quick answer

DXGI_ERROR_DEVICE_HUNG (often code 0x887a0006) means your GPU stopped responding and Windows dropped the connection, crashing the game to desktop. The fixes in order: clean-install or roll back the GPU driver, cap your frame rate to your monitor's refresh, turn off frame generation (DLSS or FSR), and strip out every overclock including your XMP or EXPO memory profile. One of those resolves the large majority of cases. For stubborn single-game crashes, verify files and force DirectX 11.

Key takeaways

  • DXGI_ERROR_DEVICE_HUNG means the GPU stopped responding and Windows dropped the connection, crashing the game.
  • A misbehaving or mismatched GPU driver is the single most common cause.
  • Capping your frame rate and disabling frame generation resolve many cases instantly.
  • An unstable overclock, including XMP or EXPO memory, is a frequent hidden trigger.
  • Verifying game files and forcing DirectX 11 are useful fallbacks for specific titles.

What the error actually means

When you see GetDeviceRemovedReason fail with DXGI_ERROR_DEVICE_HUNG or DEVICE_REMOVED, Windows is telling you the graphics card stopped responding to commands. The game engine suddenly has no GPU to talk to, so it crashes to the desktop immediately. The trigger can be a driver fault, an unstable overclock, a thermal spike, or a specific game pushing the GPU in a way it does not like. Work through the fixes below in order, testing after each.

Here is how to prioritize, ranked by how often each cause is the real one and how long the fix takes:

CauseHow commonFixTime
Bad or mismatched GPU driverVery commonClean-install or roll back the driver15-20 min
Uncapped frame rateCommonCap FPS to monitor refresh1 min
Frame generation (DLSS/FSR)Common in new gamesToggle it off1 min
Unstable overclock or XMP/EXPOCommon, overlookedReset to stock, disable memory profile5 min
Corrupt game filesSingle-game casesVerify files, force DirectX 1110-20 min
Overheating or weak PSULess commonImprove airflow, check power30+ min

Fix 1: Sort out your GPU driver

This error is most often a graphics driver crashing under load.

    1. Update to the latest stable GPU driver from NVIDIA, AMD, or Intel.
    2. If the crashes started right after a driver update, roll back to the previous known-good version instead. Some specific driver and game combinations are known to misbehave.
    3. For a clean slate, use a Display Driver Uninstaller in Safe Mode, then install the driver fresh.
    4. Test the game again before moving on.

Note

If a recent driver caused the problem, a clean install of a slightly older, stable driver often fixes it faster than waiting for a hotfix. Note which driver version was stable so you can return to it.

Fix 2: Cap your frame rate

An uncapped frame rate can drive the GPU to wild load swings that provoke a hang, especially in menus or simple scenes where the card spikes to thousands of frames. Capping smooths the load.

In the game's settings, cap your frame rate to 60 or to your monitor's exact refresh rate. You can also enforce a cap in the NVIDIA or AMD control panel. This simple change resolves a surprising number of device-hung crashes on its own.

Fix 3: Disable frame generation

Frame generation, the DLSS and FSR feature that interpolates extra frames, is a frequent trigger of DirectX 12 instability in newer games. If you run it, turn it off and test. For many players hitting these crashes, that single toggle is the fix. You can read more about how the feature works in our DLSS, FSR, and XeSS upscaler guide.

A gaming PC graphics card under load, where device-hung crashes originate
Photo: TJStamp / flickr (BY 2.0)

Fix 4: Remove every overclock

Overclocks that seem perfectly stable in benchmarks can still cause DXGI_ERROR_DEVICE_HUNG under the specific load a game imposes. This is one of the most overlooked causes.

Reset your CPU and GPU to stock clocks, including any MSI Afterburner profile. Crucially, also test with your memory profile disabled, because an unstable XMP or EXPO setting can manifest as a GPU hang even though the RAM is the real culprit. If disabling it stops the crashes, your memory overclock was unstable; see our XMP and EXPO guide for getting it stable. If you undervolt, a too-aggressive curve can also cause hangs, as covered in our GPU undervolting guide.

Fix 5: Verify files and try DirectX 11

If the crash only happens in one game, verify its files through Steam or the Epic Games Store to repair any corrupted data. Many titles also offer a launch option to run in DirectX 11 instead of DirectX 12. Forcing the DirectX 11 path often sidesteps the specific compute crashes tied to the newer API, at a small performance cost. It is a reliable workaround while you wait for a patch.

Fix 6: Check temperatures and power

A GPU that overheats or briefly loses stable power can hang. Monitor your GPU temperature during play; if it climbs past safe limits, improve case airflow and clean dust from the card. On the power side, a failing or underpowered PSU can cause hangs under peak load, particularly with high-end cards that have sharp transient spikes.

Frequently asked questions

What causes DXGI_ERROR_DEVICE_HUNG?

It happens when the GPU stops responding to commands and Windows drops the connection. Common causes are a crashing or mismatched driver, an unstable overclock, frame generation bugs, overheating, or power delivery problems.

Does capping FPS really stop the crash?

For many people, yes. An uncapped frame rate causes large, rapid GPU load swings that can provoke a hang. Capping the frame rate evens out the load and prevents the spike that triggers the crash.

Should I update or roll back my driver?

If you are on an old driver, update. If the crashes began immediately after a recent update, roll back to the previous stable version. Both are worth testing because specific game and driver combinations can be the problem.

Is this a sign my GPU is dying?

Usually not. It is far more often a software or overclock issue. However, if the crash persists at completely stock settings across many games and you also see artifacts, it is worth testing the card in another system to rule out hardware failure.

What to do right now

Crashing every session and need it fixed tonight? Run this in order and test after each step:

  • Cap your frame rate to your monitor's refresh in the game or the GPU control panel (fastest possible win).
  • Turn off frame generation (DLSS Frame Gen or FSR Frame Gen) and retest.
  • Reset CPU, GPU, and any Afterburner profile to stock, and disable your XMP or EXPO memory profile.
  • If still crashing, clean-install the GPU driver with DDU in Safe Mode, or roll back if it started after an update.
  • For a single problem game, verify its files and add the launch option to force DirectX 11.
  • Monitor GPU temperature during play and improve airflow if it runs hot.

The bottom line

DXGI_ERROR_DEVICE_HUNG looks scary but is rarely fatal. Start with your driver, cap your frame rate, and disable frame generation, then strip out every overclock including your memory profile. One of those almost always resolves it. For stubborn single-game cases, verify files and force DirectX 11 while you wait for a patch.

#gaming#troubleshooting

Sources & further reading

Keep reading