STATUS_GRAPHICS_ADAPTER_WAS_RESET (0xC01E0003) – Fix Guide

You're in the middle of a game or a GPU-heavy render, and suddenly your screen goes black for a few seconds. Then it comes back, but you get a notification or application crash with error 0xC01E0003 – STATUS_GRAPHICS_ADAPTER_WAS_RESET. This is the Windows graphics subsystem telling you the display driver stopped responding, the OS killed it, and the driver restarted. It's not a BSOD, but it's the same family of problem: the GPU hit a timeout (TDR – Timeout Detection and Recovery).

What's actually happening here is Windows has a watchdog timer for GPU operations. If the GPU doesn't respond within 2 seconds (default), the OS assumes it's hung, resets the driver stack, and your screen goes black momentarily. The error code 0xC01E0003 is the return from that reset – the driver says "everything we were doing is gone, start over."

Why Does This Happen?

The root cause is almost always one of three things:

• GPU overclock too aggressive – memory or core clock pushed past the chip's stability limit for sustained workloads. The GPU can't finish a shader operation in time.
• Unstable or outdated display driver – a buggy driver version, especially after a Windows update or driver update that didn't clean previous install.
• Power delivery hiccup – the GPU isn't getting consistent voltage, often from a PSU that's aging, underpowered, or sharing a rail with hungry components.

Less common but real: overheating (GPU hitting 85°C+ for long periods), memory errors, or a failing GPU. But 9 times out of 10, it's driver or overclock related.

The Fix – Step by Step

1. Roll back any recent driver or Windows update.
   Go to Settings > Windows Update > Update history > Uninstall updates. If the error started after a Windows Update or GPU driver update, remove the update. Reboot. Test. If that fixes it, block that update from reinstalling using the Microsoft Show/Hide tool.
2. Do a clean reinstall of your GPU driver using DDU.
   Don't just run the installer. Download Display Driver Uninstaller (DDU). Boot into Safe Mode. Run DDU – select your GPU vendor (Nvidia or AMD) and let it wipe the driver completely. Reboot into normal Windows, then download the latest official driver from your GPU vendor – not from Windows Update. Install it with the "Clean Install" checkbox if Nvidia, or "Factory Reset" if AMD. Reboot again. This removes any leftover junk that causes conflicts.
3. Drop your GPU overclock to stock.
   If you've overclocked even slightly, return core clock and memory clock to factory defaults. Test with a known stable game or benchmark (Unigine Heaven or FurMark for 15 minutes). If the error disappears, you know the overclock was the cause. You can then try a milder overclock (e.g., +50 MHz core, +100 MHz memory) and test again. Some chips just can't hold a high boost under sustained load.
4. Check your power supply.
   Open the case. Look at the GPU power cables. Are you using a single daisy-chained cable to power a GPU with two 8-pin connectors? If yes, that's often the culprit. Use separate cables from the PSU for each connector. The daisy-chain can't deliver enough current under spikes. Also check your PSU wattage – if it's below the GPU manufacturer's recommendation plus 100W headroom, upgrade. A 650W unit for a card that recommends 750W will cause resets under load.
5. Increase the TDR timeout as a band-aid.
   This doesn't fix the root cause, but it stops the crash from triggering while you diagnose. Open Registry Editor (regedit), go to HKEY_LOCAL_MACHINE\SYSTEM\CurrentControlSet\Control\GraphicsDrivers. Create a DWORD (32-bit) named TdrDelay and set it to decimal 8 (seconds). Reboot. This gives the GPU more time to respond before Windows resets it. The real fix is above – use this only if you're stuck.
6. Test with a different application or benchmark.
   Sometimes a specific game or app triggers the error due to its rendering pipeline. Run a benchmark like 3DMark Time Spy or FurMark. If those run fine, the issue is isolated to that one app. Update the app, or adjust its graphics settings (disable ray tracing, lower texture quality, cap FPS).

If It Still Fails

If you've done steps 1 through 4 and the error persists, you're looking at hardware degradation. Start with the easiest test: reseat the GPU. Power down, unplug the GPU, remove it, clean the PCIe slot contacts with isopropyl alcohol, put it back firmly. Also reseat the RAM – bad memory timing can cause GPU resets on some motherboards.

Next, monitor your GPU temperatures using HWMonitor or GPU-Z during a stress test. If the card hits 85°C or higher, you have a thermal problem. Clean the fans, replace thermal paste if you're comfortable, or improve case airflow.

If all else fails, test the GPU in a different PC. If the error follows the card, the GPU itself is failing – often the memory modules or a voltage regulator. Time for a replacement. If the error stays on the original PC with a different GPU, your motherboard's PCIe slot or PSU is the issue.

A final note on Windows 11 vs Windows 10: Windows 11 has stricter TDR behavior with WDDM 3.0 drivers. If you're on Windows 11 with an older GPU (pre-2018), try rolling back to Windows 10 or using the TdrDelay registry tweak above. The error is more common there.