VIDEO_TDR_FAILURE Error: Causes, Diagnosis, and Step-by-Step Fix

Quick Answer

VIDEO_TDR_FAILURE is a Windows blue screen caused by your GPU driver hanging and Windows giving up on it. The fix involves updating or rolling back the graphics driver, checking GPU temps, testing RAM, and confirming your PSU isn't undersized. Most SA users resolve it in under 30 minutes with a clean driver reinstall using DDU.

What Triggers VIDEO_TDR_FAILURE

TDR stands for Timeout Detection and Recovery. Windows expects the GPU to respond within 2 seconds, and when it doesn't, Windows resets the driver and throws a blue screen. The error name in the BSOD usually points at nvlddmkm.sys for Nvidia cards, atikmdag.sys or amdkmdag.sys for AMD, and igdkmd64.sys for Intel integrated. Common root causes are corrupted or mismatched drivers, GPU overheating past 85 degrees, an undersized or failing PSU, faulty system RAM, or aggressive overclocks. Less often it's a dying graphics card, especially on cards from 2018 and earlier where VRM failure is creeping in.

Step-by-Step Driver Fix

Start with a clean driver reinstall. Download the latest Nvidia GeForce or AMD Adrenalin driver matching your card and Windows 11 24H2 or 25H1. Boot into Safe Mode by holding Shift while clicking Restart, then choose Safe Mode with Networking. Run DDU (Display Driver Uninstaller) and pick the Clean and Restart option for your GPU brand. Once Windows boots normally, install the freshly downloaded driver. Skip Windows Update graphics drivers, they're often older and conflict with manufacturer releases. If the blue screen returns within a day, roll back to the driver from 2 versions ago since brand new releases sometimes carry regressions.

Hardware Checks: Temps, Power and RAM

Open HWiNFO64 or MSI Afterburner and watch GPU temperatures during a 10 minute Furmark or 3DMark Time Spy run. Anything above 85 degrees on the core or 95 on hotspot points at thermal paste degradation or fan failure, common on cards 3 years and older. Next, verify your PSU wattage matches the GPU's requirement. An RX 7800 XT or RTX 4070 Super needs 700W minimum, and a tired 600W unit causes random TDR resets under load. Run Windows Memory Diagnostic or MemTest86 overnight to rule out faulty RAM, since bad sticks often masquerade as GPU failures. Loadshedding-related dirty power can also damage components over time, so a UPS with surge protection saves repair bills.

When It's Time to Replace the GPU

If clean drivers, fresh thermal paste, a new PSU and good RAM all check out, and TDR persists, the card itself is likely on its way out. Visual artefacts, random black screens at idle, or BSODs even in light browsing point at VRAM or VRM failure. SA gamers often nurse a card for 5 to 7 years, so an RX 580 or GTX 1070 throwing TDR errors in 2026 has earned its retirement. New mid-range options like the RX 7600, RTX 4060 or RTX 4060 Ti at R5,999 to R9,999 deliver a major lift and ship with free SA delivery on qualifying orders.

Frequently Asked Questions

Can VIDEO_TDR_FAILURE damage my GPU permanently?

The crash itself doesn't damage hardware, it's a recovery mechanism. But repeated crashes often indicate underlying issues like overheating or bad VRMs that, if ignored, will eventually kill the card. Fix the root cause within a few weeks of first seeing the error.

Does VIDEO_TDR_FAILURE happen more on Nvidia or AMD?

Both vendors trigger TDR equally often, the cause is usually driver mismatch or hardware issue rather than brand. Recent driver releases from both Nvidia (Game Ready 560 plus) and AMD (Adrenalin 25 plus) have been stable for most SA users on Windows 11 24H2.

Will undervolting my GPU stop VIDEO_TDR_FAILURE crashes?

Undervolting can help if the crash is caused by power delivery issues or marginal silicon. Drop the core voltage by 50 to 80mV in MSI Afterburner and stress test, but if the crashes return, the issue lies elsewhere and undervolting is masking the symptom.