GPU Game Crashes: How to Check If It’s Your Graphics Card

Game crashes can stem from many sources, but the GPU is a prime suspect. This guide walks you through a systematic process to determine whether your graphics card is the culprit, using built-in tools and safe testing methods. You’ll need access to your system’s Event Viewer, a GPU monitoring tool, and a stable game to test with.

What You’ll Need

• A Windows PC (the steps reference Windows 10/11; Mac and Linux differ).
• Administrator access to view certain logs.
• A GPU monitoring utility (e.g., MSI Afterburner, HWiNFO, or the built-in Task Manager).
• A game that crashes consistently.
• Optional: a second GPU or integrated graphics for A/B testing.

Step-by-Step

1. Check Event Viewer for GPU-related errors. Press Win + X and select Event Viewer. Navigate to Windows Logs > System. Look for errors with source Display, nvlddmkm (NVIDIA), amdkmdag (AMD), or igfx (Intel). A recurring error like “Display driver stopped responding and has recovered” points to a GPU or driver issue. Note the error codes and timestamps.
2. Monitor GPU temperatures and usage. Open your monitoring tool and run the game in windowed mode. Watch the GPU temperature, clock speed, and usage. If the temperature exceeds 85°C (185°F) under load, overheating is likely. Also note if the GPU usage drops to 0% right before a crash, which suggests the driver crashed.
3. Update or roll back your GPU driver. Go to your GPU manufacturer’s website (NVIDIA, AMD, Intel) and download the latest driver. Use the “Clean Install” option if available. If the crashes started after a recent update, roll back to the previous version via Device Manager (right-click GPU > Properties > Driver > Roll Back Driver).
4. Run a stress test. Use a tool like FurMark or Unigine Heaven to put your GPU under maximum load. If the system crashes or artifacts (visual glitches) appear within 15 minutes, the GPU is likely failing. If it passes, the issue may be game-specific.
5. Test with a different game or benchmark. If only one game crashes, it may be that game’s code, not your GPU. Run a demanding benchmark like 3DMark. If the benchmark runs fine, your GPU is probably healthy.
6. Check power supply and connections. Ensure the GPU’s power cables are firmly seated. If your PSU is underpowered or failing, the GPU may crash under load. Listen for coil whine or see if the system restarts abruptly, which indicates power issues.
7. Disable GPU overclocks. If you’ve overclocked your GPU, revert to stock clocks. Use MSI Afterburner or your motherboard’s software to reset. Overclocks that are unstable can cause crashes in specific games.
8. Test with integrated graphics (if available). Remove your dedicated GPU and plug your monitor into the motherboard’s video port (if your CPU has integrated graphics). Play the same game. If it runs fine, your dedicated GPU is the problem. If it crashes too, the issue is elsewhere (CPU, RAM, or software).

Common Mistakes

• Ignoring event viewer logs: Many users skip this step and guess. The logs often pinpoint the exact driver or hardware error.
• Not monitoring temperatures: A GPU that runs at 90°C will crash, but if you don’t watch temps, you’ll miss the obvious cause.
• Assuming any crash is GPU-related: RAM, CPU, and storage issues can also cause crashes. Test systematically.
• Using outdated or beta drivers: Beta drivers can be unstable. Stick to stable releases unless you need a specific feature.
• Skipping the clean driver install: A simple “update” may leave old files that conflict. Always use the clean install option when possible.

Tips

• Keep a log: Note when crashes happen (in-game, at menu, during loading) and any error messages. Patterns help diagnose.
• Check game-specific forums: If a particular game crashes for many users, it’s likely a game bug, not your GPU.
• Use Windows’ built-in reliability monitor: Type “reliability” in the Start menu to see a timeline of crashes and warnings.
• Underclock your GPU as a test: If you lower the core clock by 100 MHz and crashes stop, it indicates a factory overclock or silicon degradation issue.
• Check for dust: Clean your GPU and case fans. Dust buildup causes overheating and crashes.

Troubleshooting

If the steps above don’t resolve the issue, consider the following:

• If no errors appear in Event Viewer: The crash may be caused by the game engine itself or a corrupted game file. Verify the game’s files via Steam/Epic/other launchers.
• If temperatures are fine but crashes persist: Try lowering the game’s graphics settings. If the game becomes stable at lower settings, the GPU may be struggling with the load, but not necessarily failing—it could be a power delivery issue.
• If the system restarts instead of showing a crash: This is often a PSU or motherboard issue. Test with a known-good PSU if possible.
• If the GPU passes stress tests but crashes in games: The game may be using a specific feature (e.g., ray tracing) that triggers a driver bug. Update the driver or disable that feature.
• If all else fails: Try the GPU in another PC. If it crashes there too, the GPU is likely faulty and may need RMA or replacement.

Frequently Asked Questions

Can a GPU crash without overheating?

Yes. Driver bugs, unstable overclocks, insufficient power, or memory errors can cause crashes even at normal temperatures.

How do I know if it’s a driver issue vs. hardware failure?

If a clean driver install fixes the problem, it was driver-related. If crashes persist across multiple driver versions and clean installs, it’s likely hardware.

Is it safe to run a stress test on my GPU?

Yes, as long as you monitor temperatures and stop if they exceed safe limits (typically 90°C for most GPUs). Stress tests are designed to push hardware, but they shouldn’t damage it if used correctly.

What does “Display driver stopped responding and has recovered” mean?

It means Windows detected that the GPU driver hung, but it recovered. This is often caused by a driver bug, overheating, or an unstable overclock.

Can a failing PSU cause GPU crashes?

Absolutely. If the PSU can’t deliver stable voltage under load, the GPU can crash or the whole system can restart. This is more common with older or low-quality PSUs.

Should I replace my GPU if it crashes?

Not immediately. First, rule out software and driver issues. If the GPU crashes in multiple games, stress tests, and with clean drivers, then replacement or RMA is justified.

Related

• Install Unverified Apps on macOS Sequoia
• Fix Samsung Galaxy Note 4 Won’t Rotate & Gyro Dead
• Recover Data from Formatted Windows 10 Hard Drive