Fix DirectX Error 0x887A0005 in Games on Windows 10/11

What is Error 0x887A0005?

Error 0x887A0005, also written as DXGI_ERROR_DEVICE_REMOVED, is the DirectX way of saying your graphics card stopped responding or got removed from the system. In practice, it happens when the GPU takes too long to finish a frame, and Windows kills the process to protect the system. You'll see it in games like Overwatch 2, Fortnite, Cyberpunk 2077, and many Unreal Engine titles.

I've seen this error on both NVIDIA and AMD cards, on laptops and desktops. The trigger is often a demanding scene, a driver bug, or an unstable overclock. Let's fix it from the most common cause down.

Cause 1: Outdated or Corrupted GPU Drivers

The most common reason for this error is a graphics driver that doesn't play nice with the game or the OS. Windows updates can break things, or you might have a driver that's months old and missing optimizations.

Fix: Clean Reinstall Your GPU Driver

Don't just update. Do a clean install. That wipes out old settings that can cause conflicts.

1. Download DDU (Display Driver Uninstaller) from the official Guru3D site. It's free and trusted.
2. Boot Windows into Safe Mode: press Win+R, type msconfig, go to the Boot tab, check "Safe boot" with Minimal, click OK, and restart. (After the driver install, come back and uncheck it.)
3. Run DDU. Choose "Clean and restart" — it will remove the current driver completely.
4. After restart, go to your GPU manufacturer's site: NVIDIA or AMD. Download the latest driver for your exact model and Windows version.
5. Install the driver with a Custom installation and check "Perform a clean installation" if offered.
6. Restart again.

After this, launch the game and see if the error appears. If it's gone, you're done. If not, move to the next cause.

Cause 2: GPU Overclock or Undervolt Instability

If you've overclocked your GPU using MSI Afterburner or EVGA Precision X, or undervolted it, that stability can be marginal. When the GPU hits a clock speed it can't sustain, it resets the driver, and you get this error.

Fix: Set GPU to Stock Clocks

Reset your GPU to factory settings. In MSI Afterburner, click the reset button (the circular arrow) to set core clock and memory clock to defaults. Also reset the power limit to 100%. Apply and test.

If you're not overclocking, skip this step. But if you are, this fix often resolves it instantly. I've had a 10 MHz overclock crash a game that ran fine otherwise — not worth it.

Cause 3: Windows TDR (Timeout Detection and Recovery) Too Short

Windows has a watchdog called TDR that checks if the GPU responds within a certain time — normally 2 seconds. If a heavy game frame takes longer, Windows kills the driver, and you get error 0x887A0005. You can extend that timeout.

Fix: Increase TDR Delay via Registry

This is a safe tweak, but it masks the real problem if the GPU is genuinely crashing. Still, many users report it fixes the error permanently.

1. Press Win+R, type regedit, and hit Enter.
2. Go to HKEY_LOCAL_MACHINE\SYSTEM\CurrentControlSet\Control\GraphicsDrivers.
3. Right-click on the right pane, choose New → DWORD (32-bit) Value.
4. Name it TdrDelay.
5. Double-click it, set the base to Decimal, and enter 10 (seconds).
6. Also add a DWORD named TdrDdiDelay with value 10.
7. Close regedit and restart your computer.

After restart, the driver gets 10 seconds to respond before Windows steps in. This solves most timeout issues, but if your GPU is actually overheating or dying, it won't fix that.

Cause 4: Overheating or Power Supply Issues

Heat is a silent killer. When the GPU hits a thermal limit, it throttles hard, and that can trigger a device-removed error. Also, a weak power supply can cause voltage drops under load.

Fix: Monitor Temps and Check PSU

• Install HWiNFO64 or MSI Afterburner and log your GPU temperature during gaming. If it's above 85°C on most models, that's concerning.
• Clean dust from fans and heatsinks. Check that all case fans spin.
• If you have a laptop, elevate the back for airflow.
• For PSU: if your GPU requires 8-pin connectors and you're using adapters or a cheap PSU, consider upgrading to a quality unit with enough wattage (recommended: 650W or more for mid-range cards).

This fix isn't a quick software change, but it's often the real cause on older systems.

Quick Reference Table

Start with the driver reinstall — it solves the majority of cases. If you're still stuck after trying all these, check Windows Event Viewer for the exact live kernel event 141, which will point to the hardware component. That's a deeper rabbit hole, but at least you'll know it's not a software fix.