Windows stopped with a VIDEO_TDR_FAILURE (bug check 0x00000116) because the GPU stopped responding and the Timeout Detection and Recovery (TDR) mechanism failed to reset it within the allowed window. Windows tried to recover the display driver but the GPU did not recover in time, forcing a system crash. The offending file is usually atikmpag.sys (AMD) or nvlddmkm.sys (NVIDIA).
The GPU driver (NVIDIA nvlddmkm.sys or AMD atikmpag.sys) has a bug causing the GPU to stop responding under certain workloads.
The GPU is overheating, causing it to throttle or crash under sustained load.
GPU overclocking or a factory-overclocked card is running beyond its stable operating limits.
A corrupted or partially installed GPU driver is causing the display adapter to hang.
The GPU hardware itself is failing — this is especially common on older cards or those that have experienced thermal damage.
Inadequate power delivery from the PSU is causing the GPU to drop out under load.
How to fix it
Clean-reinstall the GPU driver using Display Driver Uninstaller (DDU) in Safe Mode: boot to Safe Mode, run DDU to completely remove the current driver, then install the latest driver from NVIDIA or AMD's website.
Monitor GPU temperatures under load using MSI Afterburner or HWiNFO64. GPU should stay under 85°C under sustained load. Clean dust from the GPU heatsink and fans.
If the GPU is overclocked (from BIOS, Afterburner, or a factory OC card), reduce GPU core and memory clocks to stock speeds.
Try an older GPU driver version if the crash started after a driver update.
Check that PCIe power connectors are fully seated on the GPU.
Test with a different PCIe slot on the motherboard if available.
If the GPU consistently fails regardless of driver or clock settings, the card may be failing and require replacement.
