🎮 Game Error Codes What it means & how to fix it

Unreal Engine Error Codes

Most crashes in Unreal Engine games (Fortnite, and the wave of UE4/UE5 titles) show the same engine-level messages no matter which game you're playing — the 'Out of video memory trying to allocate a rendering resource' popup, the generic 'LowLevelFatalError' / 'Fatal error!' crash reporter, and GPU-reset errors like DXGI_ERROR_DEVICE_REMOVED. They almost always trace back to the GPU driver, VRAM/page-file headroom, or a corrupt shader cache rather than the specific game. Here's what each means and the fastest fix.

Error codes

Out of video memory trying to allocate a rendering resource The engine failed to allocate the GPU/rendering memory it needs, even if your card has spare VRAM

Cause: A driver or DirectX 12 shader-compilation fault, too little page file, or (confirmed by Intel in 2026) microcode instability on some 13th/14th-gen Intel CPUs — not necessarily a truly full VRAM.

Fix:
  1. Add the launch option -dx11 (Steam: Properties -> Launch Options; Epic: Manage game -> Additional Command Line Arguments) to bypass the DX12 shader path.
  2. Clean-install the latest GPU driver and increase the Windows page file / virtual memory to give the engine headroom.
  3. Lower texture quality / resolution and close other apps; if you run a 13th/14th-gen Intel CPU, update your motherboard BIOS to the microcode fix.
LowLevelFatalError / Fatal error! The Unreal crash reporter caught an unhandled engine exception and closed the game

Cause: A corrupt shader cache or game file, an out-of-date/overclocked GPU, or a missing runtime.

Fix:
  1. Verify/repair the game files to rebuild a corrupt engine DLL or shader cache.
  2. Clean-install the GPU driver and remove any GPU overclock (including factory OC).
  3. Install the latest Visual C++ Redistributables (x86 + x64) and DirectX runtime, then disable overlays.
DXGI_ERROR_DEVICE_REMOVED / 'video driver crashed and was reset' (0x887A0006) Windows reset the graphics driver because the GPU stopped responding (a TDR event)

Cause: An unstable/overclocked GPU, an overheating or power-starved card, or a buggy driver.

Fix:
  1. Remove all overclocks and undervolt or cap the frame rate to reduce GPU spikes.
  2. Clean-install the latest driver and make sure the card has adequate cooling and PSU power.
  3. As a test, add -dx11 to force DirectX 11, which some cards handle more stably than DX12.
Assertion failed: Unreal Engine is exiting due to D3D device being lost (Error: 0x887A0005 - 'REMOVED') The graphics driver stopped responding and Windows reset the GPU, so UE lost its Direct3D device mid-frame and aborted.

Cause: Driver timeout (TDR), an unstable GPU overclock, an overheating card, or a too-aggressive power/undervolt profile.

Fix:
  1. Clean-install the latest WHQL GPU driver with DDU, and remove any GPU overclock/undervolt (including factory OC on the card).
  2. Disable overlays and GPU monitoring/tuning tools (MSI Afterburner, RivaTuner, vendor apps) that hook the driver.
  3. Cap the framerate and improve case airflow so the GPU does not spike/overheat and trip a TDR reset.
GPU crashed or D3D device removed — GPU page fault / TDR timeout A draw or dispatch either read GPU memory that was not mapped (page fault) or ran longer than the ~2s Windows TDR limit, forcing a device-removed reset.

Cause: A GPU hang from a driver/shader bug, unstable memory clocks, or a genuine hardware fault; Epic documents this under "Dealing with a GPU Crash".

Fix:
  1. Launch with -d3ddebug (or -gpucrashdebugging) to get a breadcrumb of the failing draw and confirm whether it is a page fault or timeout.
  2. Roll back to a known-stable GPU driver and test at stock clocks to rule out an unstable overclock.
  3. Lower graphics settings / resolution and switch DX12 to DX11 (-dx11) to reduce GPU load until the crash stops.
Unhandled Exception: EXCEPTION_ACCESS_VIOLATION (0xC0000005) The game tried to read or write a memory address it does not own, so Windows killed the process — the CPU-side equivalent of a GPU page fault.

Cause: A corrupt install/shader cache, faulty system RAM, an incompatible mod/plugin, or an out-of-date runtime.

Fix:
  1. Verify/repair the game files and delete the shader cache so a corrupt DLL or cached shader is rebuilt.
  2. Run a memory test (Windows Memory Diagnostic / MemTest86) and disable any XMP/EXPO overclock if errors appear.
  3. Reinstall the latest Visual C++ Redistributables (x86+x64) and .NET/DirectX runtimes, and remove recently added mods.
Unhandled Exception: EXCEPTION_STACK_OVERFLOW (0xC00000FD) The engine recursed or allocated too deep on the call stack and overran it, crashing the process.

Cause: A broken Blueprint/loop or plugin in modded builds, or a corrupted engine binary in shipped games.

Fix:
  1. Verify/repair game files to restore a clean engine executable.
  2. Disable or remove recently installed mods/plugins that may cause an infinite loop.
  3. Update the GPU driver and Visual C++ runtimes, then relaunch with default settings.

Fixes that work for most Unreal Engine 4 / 5 games (PC — Windows, any UE title e.g. Fortnite, Palworld, STALKER 2) errors

  1. Update your GPU driver with a clean install (NVIDIA: use the 'Clean install' option; AMD: use AMD Cleanup Utility then reinstall) — stale or corrupt display drivers cause most launch crashes and GPU-reset errors.
  2. Install the latest Microsoft Visual C++ Redistributables (BOTH x86 and x64) and the DirectX End-User Runtime — missing runtimes behind the game cause 'DLL not found' and 0xc000007b launch failures.
  3. Verify/repair the game files (Steam: Properties -> Installed Files -> Verify integrity; Epic: three-dots -> Verify; Battle.net/EA: Scan and Repair) to replace a corrupted engine DLL or shader cache.
  4. Increase the Windows page file / virtual memory (System -> Advanced -> Performance -> Virtual memory) and close background apps (browsers, overlays) to give the engine the memory headroom it needs.
  5. Disable in-game overlays (Discord, Steam, GeForce Experience, Xbox Game Bar) and any GPU overclock (including factory OC and MSI Afterburner) — both commonly trigger device-removed and fatal-error crashes.

FAQ

How do I fix 'Out of video memory trying to allocate a rendering resource' in an Unreal Engine game?

Despite the wording it's usually not a truly full VRAM — it's a DirectX 12 shader-compilation fault, too little Windows page file, or (confirmed by Intel in 2026) microcode instability on some 13th/14th-gen Intel CPUs. Add the launch option -dx11 to bypass the DX12 path, clean-install your GPU driver, and increase your virtual memory/page file. Lower texture quality and close background apps, and if you have a 13th/14th-gen Intel CPU update your motherboard BIOS to the microcode fix.

What does DXGI_ERROR_DEVICE_REMOVED or 'video driver crashed and was reset' mean?

It means Windows reset your graphics driver because the GPU stopped responding (a TDR event), usually from an unstable or overclocked GPU, overheating, insufficient PSU power, or a buggy driver. Remove all overclocks (including factory OC), cap your frame rate, clean-install the latest driver, and make sure the card is cool and properly powered. Forcing DirectX 11 with the -dx11 launch option is a useful stability test.

What does "D3D device removed" or "GPU crashed" mean in Unreal Engine?

It means the graphics driver stopped responding and Windows reset the GPU (a TDR timeout or GPU page fault), so Unreal lost its Direct3D device and closed. It is usually fixed by a clean driver reinstall, removing any GPU overclock, and switching DX12 to DX11.

Other games