nvlddmkm.sys is the kernel-mode NVIDIA display driver. On Windows 11, a crash that names it usually means the graphics stack exceeded Windows’ Timeout Detection and Recovery (TDR) limit and could not reset the GPU successfully. The most useful fixes are to identify the exact stop code, inspect the System log, install the driver intended for your specific GPU and computer, remove overclocking or other unstable settings, and then use recovery or hardware diagnostics if the failure persists.
Do not download a loose replacement copy of nvlddmkm.sys. It is installed as part of a signed NVIDIA driver package, so replacing one file rarely addresses the underlying driver, power, firmware, thermal or hardware problem.
As an Amazon Associate I earn from qualifying purchases.
What an nvlddmkm crash means
Windows uses TDR to prevent a stalled graphics processor from making the entire system unresponsive. The GPU scheduler watches work submitted through the Windows Display Driver Model (WDDM). If a graphics operation takes longer than the default two-second timeout, Windows treats the GPU as hung and attempts to preempt the work and reset the display driver.
Windows Errors? Fix Them Before They Spread
Repair common Windows errors and clear accumulated junk for a smoother, more stable PC - no reinstall needed.Free scan · no reinstallCrashes, No Sound, or Screen Glitches?
Random freezes, missing sound and display glitches usually trace back to one bad driver. Find and replace yours safely.Free scan · under a minuteA successful reset may briefly blank the screen and then restore the desktop. If recovery fails, Windows can stop with VIDEO_TDR_FAILURE, bug check 0x116. Microsoft’s worked example for that stop identifies nvlddmkm.sys as the image name. A related VIDEO_TDR_TIMEOUT_DETECTED event, 0x117, means Windows detected the timeout; 0x116 means the attempted recovery failed. Repeated hangs can also trigger a GPU-hang bug check after five or more hangs within one minute.
#1 Best Overall
- Powered by Radeon RX 9070 XT
- WINDFORCE Cooling System
- Hawk Fan
- Server-grade Thermal Conductive Gel
- RGB Lighting
The filename identifies the NVIDIA display-driver module involved in the failed recovery; it does not, by itself, prove that the file is corrupt. The initiating condition can be a bad or mismatched driver, an update regression, unstable clock or memory settings, excessive heat, inadequate power delivery, firmware incompatibility, damaged system files, faulty RAM, a motherboard problem or a failing graphics card.
Identify exactly what is failing
Before changing drivers, capture the evidence. A black screen followed by an automatic restart, a blue screen, a game-only crash, a desktop freeze, a Windows Reliability Monitor LiveKernelEvent, and a stop code in a minidump can have different causes even when they all mention NVIDIA.
- Write down the exact message and code:
0x116,0x117, another bug check,LiveKernelEvent, or an application error. - Note when the issue began and what changed immediately beforehand: a driver or Windows update, new game, overclock, BIOS or firmware update, monitor change, sleep/wake event, or newly installed hardware.
- Record the trigger: a particular game, video playback, waking from sleep, switching monitors, or idle use. A crash only under 3D load points to a different test path than one occurring on the desktop.
- Record the exact GPU model, Windows 11 edition and build, computer or notebook model, and whether the machine uses a desktop card or integrated and discrete graphics.
Check Event Viewer
- Right-click Start and open Event Viewer.
- Expand Windows Logs and select System.
- Choose Filter Current Log or scroll to the time of the crash. Look for display-driver resets,
nvlddmkmentries,Kernel-Power, storage errors, WHEA hardware errors and power-related warnings. - Open each relevant event and save its details. The event immediately before the restart can be more informative than the stop-screen text.
Event Viewer does not prove which component is defective, but it can reveal that a graphics timeout occurred alongside a power, storage or hardware-correction error. Keep those observations with the date and driver version so that each later test has a clear before-and-after result.
Do these 3 things before closing this tab:
1Clear out junk files and repair common Windows errors2Scan for outdated or missing drivers - takes under a minute3Repair Windows errors before they cause bigger problemsRank #2
- Powered by the NVIDIA Blackwell architecture and DLSS 4
- Powered by GeForce RTX 5070 Ti
- Integrated with 16GB GDDR7 256bit memory interface
- PCIe 5.0
- WINDFORCE cooling system
Fixes to try, in the order that limits risk
1. Return the system to stock settings
Temporarily undo every performance change before judging a driver. In the GPU utility, motherboard firmware, and memory profile controls, return the GPU core and memory clocks, CPU settings, RAM profile and voltage changes to their default or manufacturer settings. Disable experimental overlays, frame injectors and tuning profiles while testing. A setting that appears stable in one game can still fail when a different workload, resolution or power state is used.
Do not begin by increasing the TDR delay in the registry. A delay edit changes how long Windows waits before declaring a timeout; it does not repair an unstable driver, overheated GPU, failing power path or defective card. It can also make a system hang for longer and obscure the original evidence. If you have already changed a TDR value, record it and restore the Windows default while troubleshooting unless a qualified engineer has a documented reason to test another value.
2. Install the correct driver package
Use the package for the exact GPU and Windows 11 build. For a notebook, check the computer manufacturer’s support page first because its package can include power-management, display-switching and firmware integration that a generic desktop package does not. For a desktop card, use the driver package intended for that exact model and operating system from the GPU manufacturer.
Rank #3
- AI Performance: 767 AI TOPS
- OC mode: 2632 MHz (OC mode)/ 2602 MHz (Default mode)
- Powered by the NVIDIA Blackwell architecture and DLSS 4
- Axial-tech fan design features a smaller fan hub that facilitates longer blades and a barrier ring that increases downward air pressure
- A 2.5-slot design maximizes compatibility and cooling efficiency for superior performance in small chassis
Compare the date of the first crash with the date of the current driver. If the problem began after an update, test a known-good earlier package rather than assuming the newest package is best. If the system has not been updated for a long time, test the current package supplied for the model. Do not install a driver for a similar-looking GPU or a different notebook revision.
3. Reinstall the driver cleanly
- Download the intended package before removing the current one, and save a copy locally so you are not dependent on the display working during the change.
- Disconnect from the internet temporarily if Windows keeps replacing the driver during your test, then close games, monitoring tools and overlays.
- Open Device Manager by right-clicking Start. Expand Display adapters, confirm the GPU model, right-click it and choose Uninstall device. If Windows offers a checkbox to remove the driver package, select it only when you have the correct replacement package ready.
- Restart Windows. The screen may use a basic display driver and a lower resolution during this step; that is expected.
- Run the official package installer for the exact GPU and notebook or desktop model. Choose its clean-install or reset option if the installer provides one, and install only the components you need for testing.
- Restart again, reconnect the network, and test the same workload that previously failed. Change one variable at a time so you can tell whether the reinstall altered the result.
If normal-mode removal is unreliable, perform the uninstall from Windows 11 Safe Mode, then restart and install the prepared package in normal mode. Safe Mode is a diagnostic environment; it is not evidence that the GPU is healthy. If the crash occurs before you can reach the desktop, use Windows Recovery Environment’s Startup Settings to reach Safe Mode.
4. Decide between update and rollback
| When the crashes began | First driver test | What the result means |
|---|---|---|
| Immediately after a driver update | Roll back to the last known-good package for the exact model | If stability returns, the newer package or its interaction with the system is suspect; report the regression and keep the working package until a later version is available. |
| After months without a driver change | Install the current package intended for the GPU and Windows 11 | If the issue remains on a correct current package, investigate settings, temperature, power, firmware and hardware rather than repeatedly changing versions. |
| Only on a notebook after an OEM or Windows change | Test the notebook manufacturer’s package, then compare with the reference package if the manufacturer permits it | A package mismatch can affect switchable graphics, sleep/wake and power states. |
| On multiple known-good packages at stock settings | Stop cycling drivers and move to recovery and hardware checks | Persistence across versions makes a purely package-level explanation less likely. |
5. Undo a recent system change with Windows recovery
If the first crash follows a driver, application or system change and a restore point exists, use System Restore. Press Win+R, enter rstrui.exe, select a restore point dated before the failure, review the affected programs, and complete the restart. System Restore does not erase personal documents, but it can remove drivers and applications installed after the selected point.
Rank #4
- Powered by the NVIDIA Blackwell architecture and DLSS 4. System Requirements: Minimum 850W PSU with 16-pin 12V-2x6 (12VHPWR) connector required. Verify before purchasing.
- Military-grade components deliver rock-solid power and longer lifespan for ultimate durability. Compatibility: 348mm (13.7") length, 3.6 slots, 4.3 lbs. Confirm case clearance and slot spacing. GPU bracket included.
- Protective PCB coating helps protect against short circuits caused by moisture, dust, or debris
- 3.6-slot design with massive fin array optimized for airflow from three Axial-tech fans
- Phase-change GPU thermal pad helps ensure optimal thermal performance and longevity, outlasting traditional thermal paste for graphics cards under heavy loads
When System Restore is unavailable, open Settings > System > Recovery and review the available Windows 11 recovery options. A recent feature or quality update can sometimes be removed from Windows Update > Update history > Uninstall updates, provided Windows lists the update as removable. Back up important files before using reset or reinstall options, and treat those as last-resort software steps rather than a substitute for hardware diagnosis.
6. Check temperatures, airflow and power delivery
Watch GPU temperature, clock behavior and fan operation while reproducing the crash with the system at stock settings. Do not rely on a single temperature number as a universal pass/fail limit; safe operating ranges vary by GPU model, firmware and workload. Instead, look for a rapid thermal rise, clocks collapsing before the timeout, fans not responding, or a crash that appears only after the system has warmed up.
Quick wins for a faster PC:
Scan for outdated or missing drivers - takes under a minuteDriver Scan →Clear out junk files and repair common Windows errorsFree Scan →- Shut down and remove dust from intake and exhaust paths using appropriate, manufacturer-safe cleaning methods.
- Confirm that every required GPU power connector is fully seated and that the power supply and cables meet the card manufacturer’s requirements.
- On a notebook, test on the manufacturer’s power adapter rather than battery power and verify that vents are not blocked.
- Check whether the failure changes when an external monitor, dock or adapter is removed. A changed symptom is a clue, not proof that the accessory is defective.
Power interruptions, WHEA errors or simultaneous storage faults in Event Viewer deserve attention before you continue changing drivers. A technician can test the supply under load when software observations cannot distinguish a card problem from a power-delivery problem.
Best Value
- Next-Gen Intel Arc Graphics: Powered by Intel Arc A580 GPU with Intel Xe HPG microarchitecture, featuring 384 XMX engines for enhanced AI acceleration and content creation.
- High-Performance Memory: 8GB GDDR6 on a 256-bit interface running at 16 Gbps, delivering excellent bandwidth for 1440p gaming and creative workloads.
- Factory Overclocked: Engine clock set at 2000 MHz out of the box, providing optimized performance for smooth gameplay and multimedia tasks.
- Advanced Dual-Fan Cooling: Features a dual-fan design with striped axial fans and an ultra-fit heatpipe for efficient thermal management. 0dB Silent Cooling stops fans completely at low temperatures for silent operation.
- Durable Construction: Includes a stylish metal backplate for enhanced PCB rigidity and a premium aesthetic, backed by ASRock's Super Alloy components for long-term reliability.
7. Test RAM, firmware and other hardware paths
Run Windows Memory Diagnostic or the memory test supplied by the computer manufacturer, and record any error rather than dismissing it as unrelated. Review the motherboard or notebook manufacturer’s firmware notes and install a firmware update only according to its documented procedure; an interrupted firmware update can create a separate boot problem.
Where practical, test with one known-good display cable and monitor, remove unnecessary USB devices, and reseat a desktop graphics card only after shutting down, disconnecting power and following the computer manufacturer’s service guidance. Do not open a notebook or alter a sealed system if that conflicts with its warranty or service requirements.
8. Analyze a crash dump when the failure persists
Windows can save small memory dumps after a blue screen. Check Settings > System > About > Advanced system settings, open Startup and Recovery, and verify that debugging information is configured to write a small dump. After the next failure, inspect the dump with the Windows debugger and look for the bug-check code, the failing image name and the surrounding call stack.
What’s actually slowing this PC down?
Pick the symptom - the matching free tool is one click away.
A dump that names nvlddmkm.sys confirms that the NVIDIA module was active when recovery failed; it does not establish that replacing that file will fix the machine. Compare several dumps if available. The same module on every dump, combined with stock settings and multiple correct driver packages, strengthens the case for examining the GPU, power, firmware, motherboard and RAM. Different modules or different stop codes may point to a wider system problem.
When to stop troubleshooting in software
Escalate to the GPU or computer manufacturer, or to a qualified repair technician, when the crash survives a correct driver rollback or clean reinstall, stock settings, recovery checks and basic hardware inspection. Escalation is especially appropriate when the system fails in more than one application, crashes at idle as well as under load, shows artifacts or persistent black screens, reports WHEA errors, or cannot remain stable long enough to complete testing.
Quick Recap
| Observed pattern | Most useful next action |
|---|---|
| Only one game, with no system errors elsewhere | Check that game’s files, graphics settings and overlays after confirming the driver is correct; compare with another 3D workload. |
| Every 3D workload, even after rollback and stock settings | Have GPU temperature, power delivery and the card tested under load. |
| Crashes during sleep, wake or monitor switching | Prioritize notebook or motherboard OEM drivers, firmware and display-connection tests. |
| Crashes at idle plus WHEA or power events | Investigate the power supply, motherboard, RAM and firmware rather than repeating driver installs. |
| Artifacts, fan failure or immediate black screen | Stop stress testing and arrange hardware service; continued load can worsen a failing component. |
Common mistakes to avoid
- Downloading
nvlddmkm.sysfrom a file-collection site or copying it from another PC. The signed driver package must match the GPU and Windows installation. - Installing a driver for a different model because its name looks similar.
- Changing several settings, drivers and BIOS options at once, which removes the ability to identify the cause.
- Leaving an overclock or memory profile enabled while deciding whether a driver is stable.
- Treating a longer TDR delay as a repair. It changes detection timing and can hide, not solve, the fault.
- Assuming that one successful game session proves the system is fixed. Reproduce the original workload and monitor it over several sessions.
As an optional shortcut after the manual checks, Outbyte Driver Updater can scan more than 100 devices for outdated, corrupted or missing drivers, show the current and recommended versions, install official drivers, and provide driver backup, restore and scheduled scans. It runs on Windows 11, 10, 8 and 7; a 2-day free trial requires no credit card before a subscription. Results vary, and no utility repairs every crash, so keep the manual diagnosis and hardware escalation steps above.
Product prices and availability are accurate as of the date/time indicated and are subject to change. Any price and availability information displayed on Amazon at the time of purchase will apply.
