Frustrated by the DXGI_ERROR_DEVICE_REMOVED game crash? Discover the ultimate 2026 troubleshooting guide to permanently fix device hung errors and stabilize your GPU today.
You are in the middle of a tense boss fight or a ranked match, the screen freezes for a split second, and boom, you are staring at your desktop. An error message pops up displaying DXGI_ERROR_DEVICE_REMOVED.
Whether your specific prompt says dxgi_error_device_hung, fatal d3d error 24, or dxgi_error_invalid_call, the sheer frustration is exactly the same. You just want your game to work seamlessly without the constant fear of a sudden crash to the desktop ruining your progress.
Most tech sites will just tell you to update your graphics drivers and leave it at that, which is rarely enough. But if you are here reading this, you already know that basic troubleshooting does not always work.
Based on extensive community testing, rigorous hardware analysis, and deep-dive Reddit troubleshooting, we have compiled the definitive, step-by-step guide to permanently fixing the DXGI error in Windows 10 and Windows 11.
We are going to explore the root causes, the software conflicts, and the hardware adjustments necessary to get you back into your favorite games with complete stability.
What Does DXGI_ERROR_DEVICE_REMOVED Mean?
DXGI stands for DirectX Graphics Infrastructure, which is a core component of the Windows operating system designed to handle the communication between your game, the graphics API like DirectX 11 or DirectX 12, and the physical graphics hardware.
When you get this error, which is often accompanied by specific hexadecimal codes like 0x887a0005 or 0x887a0006, it means Windows temporarily lost communication with your graphics card. In the eyes of the operating system, the GPU stopped responding to commands.
To prevent your entire PC from crashing and throwing the dreaded Blue Screen of Death, the Windows kernel forcefully removes or suspends the GPU device instance. This is a protective measure.
Because the game engine suddenly has no graphics card to talk to, it results in an immediate crash to the desktop.
The game attempts to render the next frame, looks for the hardware, fails to find it, and issues the device removed or device hung error. You might also see this presented as a fatal d3d error 24 depending on the specific game engine, but the underlying mechanism is identical across the board.
What Causes the DXGI Error Device Removed Crash?
This terrifying error is rarely a sign of a dying GPU, so you can breathe a sigh of relief regarding your hardware investment. Instead, it usually comes down to a few very specific software and hardware tuning conflicts.
One of the absolute most common culprits is unstable factory overclocks. Many modern AMD and Nvidia graphics cards produced by board partners push their boost clocks incredibly aggressively out of the box to win performance benchmarks.
Sometimes, the specific silicon on your card simply cannot handle those extreme frequencies under heavy gaming loads, leading to a momentary processing failure that triggers the disconnect.
Another major cause is system memory instability, specifically related to Extreme Memory Profiles like XMP for Intel systems or EXPO for AMD systems. Pushing your RAM to extreme frequencies can cause tiny, almost imperceptible micro-stutters in data delivery to the CPU and GPU.
These micro-stutters can delay the rendering pipeline just long enough to trip the DirectX timeout protocol.
Speaking of timeouts, the Windows Timeout Detection and Recovery system, known as TDR, is another core factor. By default, Windows only gives the GPU a maximum of two seconds to process any given graphical instruction.
If the graphics card is overwhelmed by a massive explosion or an unoptimized shader compilation and takes two and a half seconds to respond, Windows assumes the GPU is dead and resets the display driver.
Finally, hardware conflicts such as multi-monitor setups running at different refresh rates or conflicting integrated graphics on gaming laptops can confuse the DirectX allocator, resulting in the exact same crash.
Step-by-Step Fix for DXGI Error
We are going to work our way down this troubleshooting list from the easiest, least invasive tweaks to the more advanced hardware adjustments. You should test your game after applying each level to see if stability has been restored.
Level 1: The Easiest Fixes to Start With
The very first thing you should try is capping your framerate. Pushing unlimited frames maximizes your GPU load, running the silicon at maximum voltage and maximum heat continuously. Cap your FPS to 60 or match your monitor’s exact refresh rate in your game’s internal settings.
If the game does not offer a reliable framerate limiter, you can enforce a strict cap via the Nvidia Control Panel by navigating to the Max Frame Rate setting, or by using the AMD Radeon Chill feature in the Adrenalin software.
Capping the framerate gives your graphics card crucial breathing room between rendering frames, significantly reducing the chances of a timeout.
Next, consider disconnecting secondary monitors during heavy gaming sessions. If your second monitor lags right before the game crashes, the Windows Desktop Window Manager might be struggling to sync different refresh rates across multiple displays. Unplug the secondary display directly from the GPU rather than just turning it off, and see if the crashes stop.
Additionally, you should always try running your games as an Administrator. You can bypass launchers like Steam, Epic Games, or the EA App by navigating directly to the game’s installation folder on your storage drive.
Find the main executable application file, right-click it, select Properties, navigate to the Compatibility tab, and check the box to run this program as an administrator. This grants the game elevated permissions to communicate with your hardware drivers without being interrupted by user account control restrictions.
Level 2: Intermediate Hardware Tweaks from the Community
If the easy fixes fail, these community-proven methods cure the problem for the vast majority of players. The absolute most successful solution across Reddit forums is the underclock fix.
You will need to download and install a free utility called MSI Afterburner, which works perfectly for both Nvidia and AMD graphics cards from any manufacturer. Once installed, open the software and look for the Core Clock slider.
You want to lower your GPU’s Core Clock by negative 50 megahertz to negative 100 megahertz. Click the checkmark to apply the setting. This simple adjustment forces factory-overclocked cards to stay within a highly stable, lower power limit. It sacrifices perhaps one or two frames per second but stops the crash entirely for many users by preventing voltage instability.
Another critical adjustment is changing the Nvidia Power Management Mode. Open the Nvidia Control Panel by right-clicking your desktop and go to the Manage 3D Settings menu. Scroll down the list of features until you find Power Management Mode.
If this is set to Prefer Maximum Performance, you need to change it back to Normal or Optimal Power. Setting it to maximum performance forces the GPU to maintain high clock speeds and voltages even during loading screens or menu transitions, which can cause transient power spikes that trip the power supply’s protections or the DXGI timeout limits.
Changing it to Normal stops the GPU from over-boosting itself into an unstable state.
Furthermore, you must perform a completely clean driver install. Using the standard update button in GeForce Experience or AMD Adrenalin often leaves corrupted registry keys and broken driver files behind.
You need to download a free tool called Display Driver Uninstaller, commonly known as DDU. Boot your computer into Windows Safe Mode so that no graphics drivers are actively running, and use DDU to completely wipe your old display drivers, audio drivers, and physics software. Once the system restarts into a normal Windows environment, install the latest stable graphics drivers downloaded directly from the manufacturer’s website.
Level 3: Advanced System Fixes
If you are still experiencing the device removed error, we need to look at deeper system configurations. You should try disabling your RAM XMP or EXPO profiles. Restart your PC and repeatedly press the Delete or F2 key to enter your motherboard’s BIOS interface.
Locate the memory settings and completely turn off your RAM’s Extreme Memory Profile, reverting the memory to its base JEDEC speed. High RAM frequencies can cause systemic instabilities that manifest uniquely as DXGI crashes because the CPU fails to feed data to the GPU fast enough. While you will lose some memory bandwidth, achieving total system stability is far more important.
The final advanced fix is the TdrLevel Registry Edit, which directly modifies how Windows handles driver timeouts. Press the Windows key and the R key simultaneously, type regedit into the run box, and hit Enter to open the Registry Editor.
Carefully navigate through the folder tree to HKEY_LOCAL_MACHINE, then SYSTEM, then CurrentControlSet, then Control, and finally GraphicsDrivers. In the right-hand pane, right-click on an empty space, select New, and then choose DWORD 32-bit Value.
Name this new entry exactly TdrLevel with a capital T and L. Double-click your new entry and set the value data to 0. This completely disables the Windows timeout detection. Note that if your system still hangs after this, your entire computer might freeze instead of just crashing to the desktop, in which case you should return to the registry and delete the TdrLevel key.
Game-Specific DXGI Error Solution
Sometimes the fault lies not with your hardware at all, but with the game engine itself. Different rendering pipelines handle graphics API calls uniquely, meaning some games are incredibly prone to this error while others run flawlessly on the exact same computer. Here are specific fixes for the most heavily searched and problematic titles currently on the market.
Baldur’s Gate 3
Baldur’s Gate 3 offers two different rendering APIs, and switching between them is often the key to stability. Switch your rendering API from Vulkan to DirectX 11 in the Larian launcher before hitting play.
If you absolutely must use Vulkan for performance reasons, you can apply a configuration file fix. Navigate to your local app data folder, find the Larian Studios directory, open the Baldur’s Gate 3 folder, and locate your vkDeviceConfig.lsx file.
Open this file using Notepad and change the UseAsyncCompute value to true. Save the file and launch the game. This changes how the Vulkan API queues compute tasks, often bypassing the specific bottleneck that causes the device removed crash.
Battlefield 6 and EA FC
Titles built on the Frostbite engine, such as Battlefield 6 and the EA FC sports series, are notorious for throwing directx function failed errors. The most effective fix here is to clear your DirectX shader cache.
Open the Windows Disk Cleanup tool, select your primary C drive, check the box for DirectX Shader Cache, and run the cleanup. This forces the Frostbite engine to rebuild its shaders from scratch, removing any corrupted files that were causing the crash.
Additionally, you must completely disable the EA in-game overlay, as its visual hooking mechanism frequently conflicts with the Frostbite rendering pipeline.
Monster Hunter Wilds and Capcom Games
If you are playing Capcom titles like Monster Hunter Wilds, the Resident Evil 4 Remake, or Devil May Cry 4, and you see the fatal d3d error 24 dxgi_error_device_removed message, you are almost certainly running out of Video RAM.
The RE Engine attempts to allocate memory dynamically, and if it asks for space that doesn’t exist, it crashes instantly. You need to dive into the graphics settings and significantly lower your texture pool sizes.
Furthermore, turn off Ray Tracing entirely, as Ray Tracing requires massive amounts of VRAM to store bounding volume hierarchies. Lowering volumetric fog quality will also ease the burden on the memory controller.
Unreal Engine 5 Titles
Games utilizing the newest Unreal Engine 5 technologies, including Tekken 8, Remnant 2, Wuthering Waves, and Arc Raiders, heavily utilize compute shaders for the Lumen lighting and Nanite geometry systems. These can sometimes overwhelm graphics drivers.
You can try to force the game to run on an older, more stable rendering path. Right-click the game in your Steam library, go to Properties, and in the General tab, look for the Launch Options text box.
Type -d3d11 or -dx11 into this box and close the menu. This forces the Unreal Engine to fallback to DirectX 11 instead of DirectX 12, which often circumvents the specific compute crashes associated with the newer API.
FiveM and Grand Theft Auto V
Players using the popular FiveM modification framework for Grand Theft Auto V frequently encounter an error stating the gpu device instance has been suspended. Because FiveM streams custom server assets dynamically, it relies heavily on your system’s storage and memory paging.
The best solution is to clear your FiveM cache folder completely. After doing that, you need to increase your Windows Virtual Memory, also known as the Pagefile.
Set a custom, large Pagefile on your fastest NVMe SSD to ensure that when FiveM runs out of physical RAM or VRAM, it has a high-speed fallback option that won’t cause the graphics driver to stall and suspend the device instance.
PlayStation PC Ports
Sony’s blockbuster PC ports, including The Last of Us Part 2, Ghost of Tsushima, Marvel’s Spider-Man 2, and Final Fantasy 7 Rebirth, are incredibly CPU and GPU intensive during their initial setup phases.
You must never skip the initial shader compilation process that happens on the main menu. Let the percentage counter hit one hundred percent before attempting to load your save file.
Trying to play the game while shaders are compiling in the background creates massive CPU usage spikes that can easily starve the graphics card of instructions, leading directly to a DXGI timeout error.
Device-Specific Troubleshooting
The type of hardware you are playing on can also dictate how you should approach fixing this issue. Desktop computers have different power delivery methods than portable electronics, requiring tailored troubleshooting steps.
Handheld Gaming PCs
If you are gaming on portable handhelds like the Steam Deck or the ASUS ROG Ally, you are using an APU, which is a processor that shares system memory with the graphics cores.
Ensure your APU drivers are updated via the manufacturer’s specific software hub, like Armoury Crate, rather than relying solely on Windows Update.
More importantly, you should consider entering the device BIOS to increase your UMA Frame Buffer Size. By default, some handhelds only allocate a small amount of RAM to the graphics cores. Increasing this allocation to four gigabytes or even eight gigabytes gives the integrated graphics enough memory to prevent the DXGI error in demanding modern titles.
Gaming Laptops
Gaming laptops face unique challenges regarding power delivery and graphics switching technologies. First and foremost, ensure your laptop is plugged directly into a wall outlet and not running on battery power. Battery power cannot supply the necessary wattage for peak GPU performance, causing immediate crashes.
If your laptop features a MUX switch or Advanced Optimus technology, open your control software and ensure the laptop is set to use the dedicated Nvidia or AMD GPU explicitly. Relying on auto-switching modes can cause the display driver to crash if the system attempts to hand rendering duties back and forth between the integrated CPU graphics and the dedicated graphics card while a game is actively running.
Software and Overlay Conflicts to Avoid
Beyond hardware tuning and game settings, background software is a massive contributor to DXGI crashes. Modern gaming PCs are often loaded with peripheral management software and social overlays that try to hook into the game’s rendering pipeline to display frames per second counters, social notifications, or crosshair overlays.
When multiple programs fight for control over the DirectX presentation layer, the graphics driver can panic and reset.
You should systematically disable the in-game overlays for applications like Discord, GeForce Experience, AMD Adrenalin, and even the Steam overlay itself. Furthermore, RGB control software such as Corsair iCUE, ASUS Armoury Crate, or Razer Synapse utilize hardware polling to monitor system temperatures and adjust lighting.
This constant polling can interrupt the GPU’s data stream. Try closing all non-essential background applications and RGB software completely via the Windows Task Manager before launching a highly demanding game.
This ensures the cleanest possible communication channel between the game engine and your graphics card.
Hardware Maintenance and Power Delivery
While we established that a dying GPU is rarely the cause, physical power delivery issues can absolutely trigger a device removed error. Modern high-end graphics cards require massive amounts of power, and they often experience transient power spikes that demand double their rated wattage for a fraction of a millisecond.
If your power supply unit is aging or lacks the necessary ATX 3.0 transient response capabilities, the voltage may drop slightly during these spikes. When the GPU detects a voltage drop, it shuts down to protect itself, resulting in the DXGI error.
Ensure that you are using separate, dedicated PCIe power cables from your power supply to your graphics card. Never use the daisy-chained or split ends of a single cable to power a high-wattage GPU, as a single cable cannot safely deliver the required amperage.
If you are using a newer Nvidia card with the 12VHPWR connector, make absolutely certain that the cable is fully and deeply seated into both the graphics card and the power supply, as even a slightly loose connection can cause power delivery fluctuations and subsequent driver timeouts.
Additionally, keep your hardware clean. A GPU packed with dust will overheat quickly. When a graphics card hits its thermal limit, it aggressively throttles its clock speeds down.
A sudden, massive drop in clock speed can destabilize the driver and cause a crash, so maintaining healthy temperatures through good case airflow and regular dusting is essential for software stability.
FAQ: Frequently Asked Questions
Does dxgi_error_device_removed mean my graphics card is dying?
No. In ninety-nine percent of cases, this is a software conflict, a driver timeout, or an unstable factory overclock. Your hardware is physically fine. It simply encountered a mathematical or power delivery situation that forced the operating system to sever the connection for safety. Following the underclocking and registry steps above will resolve the issue for nearly all functional hardware.
Why does the error still happen after formatting Windows 11?
Because resetting Windows does not reset your BIOS settings or your GPU’s firmware. If your RAM XMP profile is fundamentally unstable, or your graphics card’s factory-programmed boost clock is simply too high for the silicon to handle, the error will easily survive a fresh Windows installation. You must apply the hardware-level fixes, such as underclocking via third-party software and disabling extreme memory profiles, to achieve lasting stability.
Can undervolting fix the DXGI error instead of underclocking?
Yes, undervolting is actually an excellent alternative to underclocking. By using the curve editor in MSI Afterburner, you can tell the GPU to operate at its normal clock speeds but at a slightly lower voltage limit. This often smooths out power delivery and prevents the transient spikes that cause the driver to crash, while allowing you to maintain full gaming performance. However, undervolting requires more precise testing than a simple underclock, as providing too little voltage will cause a different type of system crash.
What should I do if absolutely none of these fixes work?
If you have tried every software fix, underclocked your GPU, disabled memory profiles, and ensured perfect power delivery, and you are still receiving the DXGI error across multiple different games, you might be looking at a rare case of actual hardware degradation. In this scenario, you should attempt to test your graphics card in a completely different computer to see if the issue follows the GPU. If it does, you should contact the manufacturer to begin the warranty return merchandise authorization process.
Visit Our Post Page: Blog Page