How to Fix BSOD THREAD_STUCK_IN_DEVICE_DRIVER (0x000000EA)
What error 0x000000EA means
Windows constantly watches your graphics card, and it has a watchdog timer for exactly this situation. If the GPU stops responding to commands for about two seconds, Windows tries to reset the driver and recover gracefully instead of crashing — that is the "TDR" (Timeout Detection and Recovery) mechanism you may have seen in forum threads. THREAD_STUCK_IN_DEVICE_DRIVER (stop code 0x000000EA) is what happens when that rescue attempt fails: a thread inside the device driver spun itself into an infinite loop, the GPU never came back, and Windows had no choice but to halt everything with a blue screen.
Despite the generic name, the "device driver" is the graphics driver in the overwhelming majority of cases — NVIDIA, AMD, or Intel GPU drivers. That is why the crash so often happens while gaming, editing video, or scrolling hardware-accelerated apps: those are the moments the GPU is working hardest, and a bug, a bad install, or a flaky clock setting finally shows itself. The good news is that this error rarely means a dead card. Corrupt or incompatible drivers, Windows Update silently swapping in an older graphics driver, overheating, unstable overclocks, or shaky power delivery cause most 0x000000EA crashes, and all of them can be checked and fixed at home. Work through the steps below in order and stop as soon as the crashes end.
Common causes
- Corrupt or incompatible graphics driver: a botched update, leftover files from previous driver versions, or Windows Update silently replacing your working driver is the single most common trigger.
- Windows Update driver conflicts: Windows has a habit of installing its own older GPU driver over a newer manual install, creating a mismatch that locks the driver thread.
- Overheating GPU: dust-clogged fans, dried thermal paste, or poor case airflow push the card past its safe limits, and an overheated GPU stops responding mid-task.
- Unstable overclock or undervolt: pushed GPU/memory clocks, aggressive undervolts, or XMP memory profiles can stall the driver exactly the way this blue screen describes.
- Power delivery problems: a loose PCIe power connector, daisy-chained splitter cables on high-power cards, or an undersized PSU starves the GPU under load, which freezes its driver thread.
How to fix error 0x000000EA — step by step
Work through these fixes in order. The Safe Mode check and the DDU clean-install end the majority of cases, so resist the urge to touch hardware before you've ruled out software.
- Boot into Safe Mode to confirm it is the driver. Interrupt the boot process three times in a row (hold the power button while the Windows logo shows) to reach the Recovery screen. Go to Troubleshoot > Advanced options > Startup Settings > Restart, then press 4 for Safe Mode. Safe Mode loads only Microsoft's basic display driver, so if the PC is perfectly stable here, your GPU driver is the culprit and steps 2–3 are very likely to end the crashes. If it crashes even in Safe Mode, system files, RAM, or hardware become the prime suspects.
- Clean-install the graphics driver with DDU. A plain driver reinstall is often not enough, because corrupted leftovers from previous versions are a classic 0x000000EA trigger. Download Display Driver Uninstaller (DDU) and the latest driver from NVIDIA, AMD, or Intel — not the version Windows Update offers. Boot into Safe Mode, run DDU, select your GPU brand, and choose "Clean and restart." After rebooting normally, install the downloaded driver, pick Custom installation, and tick "Perform a clean installation." Then stress the GPU with a game or a benchmark for a while before declaring victory.
- Stop Windows from reinstalling the bad driver. This step is easy to skip and costly to regret: Windows Update can silently replace your freshly installed driver with an older one overnight and bring the crashes back. Search the Start menu for "Device installation settings," open it, select "No (your device might not work as expected)," and save. If you are comfortable with the registry, you can also set the
ExcludeWUDriversInQualityUpdatepolicy to block driver delivery through updates entirely. - Check GPU seating, power connectors, and temperatures. Shut down, switch off the PSU, open the case, and reseat the graphics card firmly in its PCIe slot until the latch clicks. Make sure every power connector on the card is fully seated, with cables running directly from the PSU — avoid split daisy-chain adapters on power-hungry cards. Back in Windows, watch temperatures with HWMonitor during a game: a GPU sustaining 85°C or more under load can lock its driver thread, in which case cleaning dust from the fans and improving case airflow are the real fix.
- Reset any overclock and test at stock settings. Open your tuning tool — MSI Afterburner, Intel XTU, AMD Software, or the BIOS — and restore default clocks for the GPU, CPU, and memory, disabling XMP or EXPO memory profiles as well. Overclocked or aggressively undervolted GPUs are a textbook cause of stuck driver threads, and this test is free. If the crashes stop at stock settings, your previous clocks were never stable; re-apply them gradually, or accept stock stability.
- Update chipset and BIOS, then test the hardware itself. Install the latest chipset drivers from your motherboard maker's support page, and check for a BIOS/UEFI update — PCIe and power-management fixes in BIOS updates have resolved this exact error for many users. If a clean driver, cool temperatures, and stock clocks still crash, move the card to another PCIe slot, or test it in a different PC. If the crashes follow the card, the hardware itself is failing and needs repair or replacement; if they don't, the motherboard or PSU is the suspect.
Did this fix it? Quick checklist
- PC is stable in Safe Mode and crashes only with the GPU driver loaded
- GPU driver clean-installed with DDU and driver auto-updates blocked
- Card reseated, power connectors secure, GPU under 85°C under load
- Overclocks reset to stock and chipset/BIOS updated
If crashes continue after all six steps, open Event Viewer (Win + X > Event Viewer), go to Windows Logs > System, and look for BugCheck or Display driver timeout entries around the crash time — the faulting module tells you exactly what hung. You can also check C:\Windows\Minidump files with a tool like BlueScreenView. As a last resort, a clean Windows install removes every driver and damaged file at once: back up your files first and choose "Keep nothing" so nothing suspect survives.
Frequently asked questions
What does THREAD_STUCK_IN_DEVICE_DRIVER (0x000000EA) mean?
It means a thread inside your graphics card driver got stuck in an infinite loop and stopped responding. Windows normally tries to reset the GPU after about two seconds of unresponsiveness, and when that reset fails it halts the system with this blue screen. The "device driver" in the name is almost always the GPU driver: NVIDIA, AMD, or Intel.
Does 0x000000EA mean my graphics card is dying?
Not necessarily. In most cases it is a software problem: a corrupt or incompatible GPU driver, a botched driver update, or Windows Update silently replacing your driver. Overheating, an unstable overclock, or weak power delivery are the next most common causes. A dying card is possible but should only be assumed after a clean DDU driver install and stock-clock testing still crash.
Why does 0x000000EA mostly happen in games or video?
Games and hardware-accelerated video push the GPU hardest, so a driver bug or an unstable clock that stays hidden on the desktop finally shows up under load. Overheating also peaks during gaming. If the crash only happens in games, suspect the GPU driver, heat, or an overclock before anything else.
Will reinstalling Windows fix 0x000000EA?
It often helps, because it removes every corrupt driver and conflicting setting at once — but it is a last resort. Most cases are fixed much earlier by a DDU clean-install of the GPU driver in Safe Mode, checking the card's power connectors and temperatures, and resetting any overclock. Only reinstall if a clean driver, stock clocks, updated BIOS, and hardware testing all fail.