STATUS_GRAPHICS_ALLOCATION_BUSY (0XC01E0102) Fix: GPU memory stuck in use

Quick answer: Run net stop wcnfs && net start wcnfs in an admin command prompt, then restart your graphics driver with Win+Ctrl+Shift+B. If that doesn't work, reboot.

This error — STATUS_GRAPHICS_ALLOCATION_BUSY (0XC01E0102) — means the GPU's video memory manager can't release an allocation because something still holds a reference to it. Usually it's a stuck DirectX or Vulkan application that crashed but left the driver in limbo. You'll see this most often when alt-tabbing out of a game like Call of Duty: Warzone or Cyberpunk 2077, or after a remote desktop session that used GPU acceleration. The culprit here is almost always a process that didn't clean up its GPU resources on exit.

How to fix it

1. Restart the graphics driver. Press Win+Ctrl+Shift+B. Your screen will flash black for a second — that's normal. Try whatever app was failing again.
2. Restart the Windows Graphics Service. Open Command Prompt as admin. Type:
   

   net stop wcnfs
   net start wcnfs

   Then press Win+Ctrl+Shift+B again.
3. Kill the hung process. Open Task Manager (Ctrl+Shift+Esc). Look for anything using high GPU — game executables, Discord with hardware acceleration, Chrome with GPU video decode. Right-click and choose End task. If you're not sure, sort the GPU column and kill the top user.
4. Reboot. If steps 1–3 don't work, just restart the machine. This flushes all GPU allocations from kernel memory.
5. Update or roll back the graphics driver. Go to Device Manager, find your GPU under Display adapters, right-click, choose Properties, go to the Driver tab. If you recently updated, click Roll Back Driver. If you haven't updated in months, download the latest from NVIDIA or AMD directly — not Windows Update.

If the main fix fails

Sometimes the issue is deeper. Here's what else to try:

• Disable Hardware Acceleration in apps that use GPU rendering — browser (Chrome/Edge: Settings > System > Use hardware acceleration when available), Discord, Slack, or any video player. Restart the app after.
• Run a full GPU stress test with FurMark or OCCT. If the error pops up during the test, your GPU might have faulty VRAM. Run MemTest86 for system RAM too — bad system RAM can corrupt GPU allocations.
• Check for malware. Some cryptominer malware grabs a GPU allocation and never releases it. Scan with Malwarebytes, then reboot.
• Reset the Windows Graphics Stack completely:
  

  Dism /Online /Cleanup-Image /RestoreHealth
  sfc /scannow

  Then restart. Rarely needed, but worth a shot if you're stuck.

How to prevent it

This error pops up when you push the GPU hard then abruptly switch contexts. A few habits cut it down:

• Don't alt-tab out of fullscreen games. Use borderless windowed mode instead. It adds a tiny latency penalty but avoids the driver context switch that triggers the bug.
• Close GPU-heavy apps before sleep or remote desktop. If you lock your PC while a game is running, the driver might not handle the power transition cleanly.
• Keep your graphics driver clean. When you update, use DDU (Display Driver Uninstaller) in Safe Mode to wipe old driver files. Driver stacking causes these kinds of allocation conflicts.
• Set a TdrDelay registry tweak. If you see this error during heavy rendering, your GPU is timing out. Increase the timeout:
  

  reg add "HKLM\SYSTEM\CurrentControlSet\Control\GraphicsDrivers" /v TdrDelay /t REG_DWORD /d 8 /f

  Default is 2 seconds. Setting it to 8 gives the GPU more time to finish before Windows declares it hung.

I've seen this error on everything from an RTX 3090 to an old GTX 970. In 90% of cases, step 2 (restarting the graphics service) fixes it without a reboot. Don't waste time reinstalling drivers unless steps 1–4 fail. And if you're still hitting it after all this, your GPU might have a hardware fault — time to RMA it.