How to Fix Display Driver Stopped Responding and Has Recovered
Published 2026-10-16 · Marion Remote Fix
Your screen goes black for a few seconds, then comes back with a notification in the system tray: Display driver stopped responding and has recovered. Or worse, the screen goes black and stays black, forcing a hard restart. These display driver crashes — called TDR (Timeout Detection and Recovery) events — happen when the GPU takes too long to respond to a command from Windows. The operating system kills the driver and restarts it, which is why the screen flickers or goes black momentarily. If the driver fails to recover, you get a full blue screen crash. This guide covers the causes and fixes from most to least common.
Update or reinstall the graphics driver
An outdated, corrupted, or incompatible graphics driver is the most common cause of TDR crashes:
- Clean install the latest driver: Download the latest driver from NVIDIA (GeForce Experience or nvidia.com/drivers) or AMD (amd.com/support). During installation, select Custom Installation and check Perform a clean installation (NVIDIA) or Factory Reset (AMD). This removes old driver files that may be causing conflicts
- DDU for persistent issues: If a standard clean install does not fix the crashes, use Display Driver Uninstaller (DDU). Boot into Safe Mode, run DDU, select Clean and restart. After restarting, install the fresh driver. DDU removes all traces of the previous driver including registry entries and leftover files that the standard installer misses
- Try an older driver: If the crashes started after a driver update, the new driver may have a bug with your specific GPU model. Roll back: open Device Manager, expand Display Adapters, right-click your GPU, select Properties, Driver tab, Roll Back Driver. Alternatively, download an older driver version from the manufacturer archive
- Windows Update driver conflicts: Windows Update sometimes installs its own generic GPU driver that conflicts with the manufacturer driver. Check Device Manager to see which driver is currently installed. If it shows a Microsoft Basic Display Adapter or a driver version you did not install, reinstall the manufacturer driver to replace it
Check for overheating
GPUs that overheat become unstable and their driver crashes as a protective measure:
- Download GPU-Z or HWMonitor and monitor GPU temperature during normal use and during the activities that trigger crashes. Desktop GPUs should stay below 85 degrees Celsius under load. Laptop GPUs thermal throttle at 90 to 100 degrees depending on the model
- Clean the GPU: On desktops, power off, unplug, open the case, and use compressed air to blow dust out of the GPU heatsink and fans. Hold the fan blades while blowing to prevent overspinning. A GPU caked with dust can run 15 to 25 degrees hotter than a clean one
- Check fan operation: With the case open and powered on, verify the GPU fans spin under load. Some GPUs use a zero-RPM mode where fans stop at idle (this is normal) but they should spin up when the GPU temperature rises above 50 to 60 degrees. If the fans never spin, the fan header connection may be loose or the fans may be dead
- Reapply thermal paste: On GPUs older than three to four years, the thermal paste between the GPU chip and heatsink may have dried out. This is more involved than a CPU paste replacement because GPU coolers often have thermal pads on memory chips and VRMs in addition to paste on the GPU die. Look up a teardown guide for your specific GPU model before attempting this
- Case airflow: A GPU in a case with poor airflow recirculates hot air and runs hotter. Ensure the case has adequate intake and exhaust fans, and that cables are not blocking the GPU fans
Adjust TDR timeout
Windows gives the GPU driver two seconds to respond before triggering a TDR reset. Some GPUs under heavy load legitimately need more time, especially during video rendering, 3D modeling, or running AI workloads:
- Open Registry Editor (regedit as administrator). Navigate to HKEY_LOCAL_MACHINE\SYSTEM\CurrentControlSet\Control\GraphicsDrivers
- Right-click in the right pane, select New, DWORD (32-bit) Value. Name it TdrDelay. Set the value to 8 (decimal). This gives the GPU 8 seconds instead of 2
- Create another DWORD named TdrDdiDelay and set it to 8 as well
- Restart the computer for the change to take effect
- This does not fix the underlying problem — it just gives the GPU more time to recover. If crashes stop, the GPU was simply taking longer than two seconds for certain operations, which happens with complex shaders, large compute tasks, or during brief thermal throttling
Check power delivery
- PCIe power cables: Mid-range and high-end GPUs require separate power connectors from the power supply (6-pin, 8-pin, or 12-pin). Ensure these are fully seated. A partially connected power cable can cause the GPU to crash under load when it draws more power than the loose connection can deliver
- Use separate cables: If your GPU has two 8-pin power connectors, use two separate cables from the power supply rather than a single cable with a daisy-chain splitter. The daisy-chain shares one cable for both connectors, which can cause voltage drops under heavy load
- Power supply capacity: A power supply that cannot deliver enough wattage for the entire system causes voltage drops under load. If you recently upgraded your GPU to a more power-hungry model without upgrading the PSU, power delivery may be the issue. Check the GPU manufacturer recommended PSU wattage and compare it to your unit
Reseat the GPU
- Power off and unplug the computer. Open the case, unscrew the GPU bracket, press the PCIe slot retention clip, and remove the GPU. Inspect the gold PCIe contacts for dust or corrosion. Reseat the GPU firmly until the retention clip clicks. This fixes contact issues caused by GPU sag, thermal expansion cycles, or installation that did not fully engage the slot
- While the GPU is out, inspect the PCIe slot itself for dust, debris, or bent pins
Hardware acceleration conflicts
- Browser hardware acceleration (Chrome, Edge, Firefox) offloads video decoding and page rendering to the GPU. If driver crashes happen specifically while browsing or watching video, disable hardware acceleration in the browser settings as a test
- Discord, Slack, and other Electron apps also use hardware acceleration. If crashes correlate with these applications, disable hardware acceleration in their settings
- If disabling hardware acceleration stops the crashes, the driver has a bug with certain GPU-accelerated operations. Check for a newer driver version that addresses it
Windows system file corruption
- Corrupted system files can cause display driver instability. Open an administrator Command Prompt and run sfc /scannow. If it finds and repairs files, restart and test. Follow with DISM /Online /Cleanup-Image /RestoreHealth if SFC reports issues it could not fix
- Check Windows Event Viewer (Windows Logs, System) for events from source Display around the time of crashes. The detail text often names the specific driver file that crashed (nvlddmkm.sys for NVIDIA, atikmpag.sys for AMD), confirming a driver-level issue versus a hardware one
RAM and system stability
- Faulty system RAM can corrupt data the GPU driver uses, causing crashes that look like GPU problems. Run Windows Memory Diagnostic (search for it in Start) or boot from a MemTest86 USB drive to test RAM thoroughly. Any errors found mean the RAM should be replaced
- If you have overclocked RAM (XMP/DOCP profile), try disabling the overclock and running at default speed. Unstable memory overclocks cause seemingly random crashes in various subsystems including the display driver
When hardware is failing
If clean driver installs, temperature management, power delivery checks, and system stability tests all come back clean but crashes persist, the GPU itself may be developing hardware failure. Artifacting (visual glitches like colored pixels, lines, or blocks that appear in the display) alongside driver crashes strongly suggests failing GPU memory or a deteriorating GPU chip. At this point, testing with a different GPU in the same system confirms whether the original GPU is the problem.
Remote troubleshooting for driver crashes
Display driver crashes can be tricky to diagnose because the causes range from software conflicts to thermal issues to failing hardware. Our remote technicians can examine your Event Viewer logs, check driver versions, monitor temperatures during the session, and apply the appropriate fix without you needing to bring the computer anywhere.