Windows 11 freezing randomly — where the mouse stops moving, the keyboard inputs nothing, and the only option is a hard restart — is one of the most disruptive PC problems precisely because it’s unpredictable. You can’t know when it’ll happen, you lose unsaved work, and the cause could be hardware, drivers, Windows itself, or some combination. We go deeper on the whole subject in our Complete Guide to Fixing Windows, Browser, and Software Errors.
The most revealing diagnostic question: when does it freeze? Every few minutes regardless of what you’re doing points at RAM or temperature. Specifically during graphics-heavy tasks points at GPU driver or GPU overheating. Only when certain applications are open points at those applications or their drivers. After sleep or waking from standby is often a driver issue with power state transitions. Freezes that began immediately after a Windows update point at a driver regression or an update bug. Note the pattern — it cuts through a lot of guesswork.
Check Temperatures First
Thermal throttling and thermal shutdown cause freezes that look identical to software crashes. When a CPU or GPU overheats, it either throttles (runs slower) or shuts itself down — which Windows experiences as a hard freeze. This is more common than most guides acknowledge, and ruling it out early saves time.
Download HWiNFO64 (free) and run it while doing the kind of work that causes freezes. Watch CPU Package Temperature and GPU Temperature in the sensor panel. CPU temperatures above 90°C under sustained load, or GPU temperatures above 95°C, are into throttling territory. A machine that used to work fine and now freezes under the same load as before is a strong indicator that something changed with cooling — dust buildup in the heatsink is the most common culprit. Compressed air through the vents often resolves temperature-related freezes without any software changes.
Also check NVMe SSD temperature if you have one. NVMe drives running above 70°C throttle their throughput dramatically — Windows operations that depend on fast storage suddenly become extremely slow and Windows can appear to freeze while waiting for a disk operation that’s been throttled to a fraction of normal speed.
Fix 1: Check for a Bad Driver Update
Windows updates frequently include driver updates, and a new driver version that’s less stable on specific hardware is a common cause of freezes that began on a specific date. If you can correlate the freezing to when Windows Update last ran, rolling back the most recent driver changes is the highest-probability fix.
Device Manager → Display Adapters → right-click GPU → Properties → Driver → check the Driver Date. If it matches when freezes started: Roll Back Driver. Same process for Storage controllers, Network adapters, and Audio if any of those also updated recently.
To check which updates Windows installed when: Settings → Windows Update → Update history. Any update with a date matching when freezes started is worth investigating. Uninstalling recent driver updates or rolling back to previous driver versions through Device Manager resolves a significant portion of sudden-onset freeze problems. If a recent update left the Settings app itself refusing to launch, our guide to fixing Windows 11 Settings not opening covers that specific failure.
Fix 2: Run Memory Diagnostics
Faulty RAM produces random freezes, blue screens, and application crashes in a pattern that looks completely random because memory errors can occur on any memory access — not just specific operations. If the freezes are truly random with no identifiable trigger, RAM is the first hardware component to test.
Windows Memory Diagnostic (built-in): search in Start → Windows Memory Diagnostic → Restart now and check for problems. The test runs on next startup before Windows loads. Results appear in Event Viewer (Windows Logs → System → look for MemoryDiagnostics-Results) after the machine boots back into Windows.
MemTest86 (more thorough): free download, runs from a USB drive, tests RAM completely independently of Windows. Run at least two complete passes. Any errors at all indicate faulty RAM.
If RAM errors are found, first try reseating the RAM sticks (remove them from the slots, clean the contacts with a dry cloth, reseat firmly). If errors persist after reseating, the RAM itself is faulty. Also check whether XMP or EXPO is enabled in BIOS — these profiles overclock the RAM and occasionally cause instability that presents as random freezes. Disabling XMP and running at standard speeds is a useful test even if errors don’t show in diagnostics.
Fix 3: SFC and DISM for Corrupted System Files
Corrupted Windows system files cause freezes when Windows tries to access a damaged file during routine operations. This is more likely after a failed update, an interrupted installation, or a disk error event. SFC and DISM repair the damage:
DISM /Online /Cleanup-Image /RestoreHealthRun this first (needs internet, takes 15–20 minutes). Then:
sfc /scannowBoth from an administrator Command Prompt. Restart after both complete. If SFC reports “found corrupt files and repaired them” — system file corruption was contributing to the freezes. The repair doesn’t fix hardware problems, but it eliminates software causes from the suspect list.
Fix 4: Disable Fast Startup
Fast Startup saves the Windows kernel session to disk during shutdown and resumes from it on boot, making startup faster. The problem: when drivers or system state become inconsistent between the saved state and the current hardware configuration, Fast Startup produces freezes and instability that only appear after a shutdown-and-start (not after a true restart, which bypasses Fast Startup).
Test this: does the machine freeze after a full restart but not after sleeping and waking? Or does it freeze specifically after shutting down and starting up — not after a restart? If yes, Fast Startup is involved. Disable it: Control Panel → Power Options → Choose what the power buttons do → Change settings that are currently unavailable → uncheck “Turn on fast startup.” Shutdowns now perform a full power cycle, and freezes caused by stale Fast Startup state stop occurring.
Fix 5: Check the Event Viewer for the Actual Cause
After a freeze that requires a hard restart, Windows logs what happened before the crash in Event Viewer. This is the most direct path to identifying the real cause — rather than guessing.
Event Viewer → Windows Logs → System → look for Critical entries around the time of the last freeze. Key events:
- Kernel-Power Event ID 41 — System restarted without a clean shutdown. Always appears after a hard restart from a freeze. The reason codes within this event can indicate power supply issues (reason 5) vs general crash (reason 63).
- LiveKernelEvent 141 or 193 — GPU driver timeout. The GPU stopped responding for too long. Points directly at the display driver.
- SYSTEM_THREAD_EXCEPTION_NOT_HANDLED with a specific driver file named — that driver is the cause.
Event Viewer → Windows Logs → Application → also check for application-level errors just before the system events that preceded the freeze. Sometimes an application error cascades into a system freeze.
Fix 6: GPU Driver Full Reinstall With DDU
When the GPU driver is identified as the likely cause through temperature checks, event viewer, or by pattern (freezes during graphics work), a clean driver reinstall with DDU (Display Driver Uninstaller) removes every trace of the current driver before installing the new one — something the standard installer doesn’t do.
- Download DDU from the official Guru3D site and the latest GPU driver from NVIDIA/AMD/Intel
- Boot into Safe Mode (hold Shift + Restart → Troubleshoot → Advanced options → Startup Settings → Enable Safe Mode)
- Run DDU → select GPU manufacturer → Clean and restart
- After the clean restart (in normal mode), install the new driver
This resolves driver-related freezes that survive standard update/uninstall/reinstall cycles because DDU removes registry entries and files that the standard uninstaller leaves behind.
Fix 7: Check Disk Health
A failing storage drive produces freezes that are almost indistinguishable from RAM failures — Windows operations stall waiting for disk reads that the failing drive can’t deliver in time. The freeze typically lasts 30–120 seconds and then either recovers or requires a hard restart.
CrystalDiskInfo (free) reads S.M.A.R.T. data from the drive and displays health status. Look for: Reallocated Sectors Count (any non-zero value is concerning), Pending Sectors (sectors with read errors waiting to be reallocated), and Uncorrectable Sector Count (non-zero values mean permanent data damage). Also check the drive’s temperature — SSDs above 70°C and HDDs above 55°C are in the problematic range.
For SSDs specifically: check drive utilisation percentage (how much of the write endurance has been consumed). Drives approaching 100% TBW (Total Bytes Written) spec become less reliable and can cause intermittent freezes as the controller manages the failing flash cells.
When Overclocking Is Involved
If the CPU, GPU, or RAM is overclocked (either manually or through XMP/EXPO for RAM), overclocking instability is an extremely common cause of random freezes — particularly ones that don’t correlate with temperature. An overclock that appears stable under light use can fail under specific computational patterns that stress different execution units.
The test: reset everything to stock/default in BIOS (look for “Load Optimized Defaults” or “Reset to Default”) and run the machine without any overclocks. If freezes stop completely, the overclock was unstable. You can then re-enable overclocking incrementally to find the stable ceiling, or leave it at stock if the performance difference isn’t critical.
Our guide on Windows 11 black screen covers the GPU driver crash that sometimes precedes or overlaps with freeze scenarios — Win+Ctrl+Shift+B for the display driver reset applies here too. For the post-update freezes where the Windows update itself is the trigger, our Windows 11 slow after update guide covers the driver rollback process in more detail. Microsoft’s Windows 11 stability documentation covers the Reliability Monitor tool — accessible by searching “reliability history” — which plots crashes on a calendar and makes it easy to see exactly when freezing started and correlate it with system changes.
Sleep and Wake Freezes
Freezes that happen specifically when the machine wakes from sleep — where Windows comes back from sleep but then freezes within a few minutes, or the wake process itself results in a frozen, unresponsive desktop — are almost always driver-related. The GPU driver, network adapter driver, and USB controller drivers all need to handle power state transitions correctly, and any one of them failing to reinitialise properly after sleep produces the post-wake freeze.
The Win+Ctrl+Shift+B shortcut is particularly useful here: when the machine wakes and then freezes shortly after, pressing this combination sends a reset signal to the display driver. It sometimes unfreezes the machine without a hard restart, and if it works consistently, it confirms the GPU driver is failing during the wake transition. The permanent fix is updating the GPU driver — sleep/wake stability is one of the things that improves significantly between driver versions for both NVIDIA and AMD.
For USB-related sleep/wake freezes: Device Manager → Universal Serial Bus controllers → right-click each USB Root Hub → Properties → Power Management → uncheck “Allow the computer to turn off this device to save power.” USB controllers powering down during sleep and failing to come back properly is a real cause of post-wake system freezes, and disabling the power management for USB hubs specifically (not the same as USB selective suspend) resolves a specific category of sleep-related instability.
Minidump Analysis for Persistent Unexplained Freezes
When freezes produce blue screens (BSODs) before the hard reset, Windows writes minidump files to C:WindowsMinidump. These files capture the exact state of the system at the moment of the crash — which driver was executing, which memory addresses were involved, and the stop code. Analysing these provides much more information than Event Viewer alone.
The free WinDbg tool from Microsoft’s developer tools can open minidump files and output an analysis. The key line to find in the analysis output is “Probably caused by” — this directly names the driver or component that caused the crash. A dump pointing at ntoskrnl.exe (the Windows kernel itself) typically indicates RAM or hardware issues rather than a specific driver. A dump naming a third-party driver (nvlddmkm.sys for NVIDIA, atikmpag.sys for AMD, or any vendor-named .sys file) points directly at that driver as the culprit.
If minidump analysis consistently points at the same driver across multiple crashes, that driver is definitively the cause. Updating, rolling back, or reinstalling that specific driver — rather than working through all possible fixes — resolves the freezes efficiently.
NVMe Thermal Throttling
This cause is underappreciated and worth addressing explicitly. NVMe SSDs — the M.2 drives most modern machines use — generate significant heat during sustained reads and writes. Without adequate cooling (heatsink, thermal pad, or adequate airflow), temperatures climb during intensive storage operations. At around 70°C, most NVMe controllers begin throttling performance. At 80°C+, throttling becomes severe enough that Windows operations waiting on disk take many seconds rather than milliseconds.
The symptom: Windows freezes specifically during large file operations, game loading, or application installations — but not during idle or light use. HWiNFO shows the NVMe temperature spiking during these operations. The fix is an M.2 heatsink (a few dollars, screws onto the drive) or improving case airflow to reduce ambient temperatures. After the thermal fix, the same intensive operations complete without freezing because the drive maintains acceptable temperatures throughout.
High RAM Usage Causing Pseudo-Freezes
Windows manages memory by moving less-used data from RAM to the pagefile on disk when RAM gets full. On machines with traditional HDDs (not SSDs), this paging process is extremely slow. When Windows runs out of RAM and begins paging heavily to an HDD, the machine becomes unresponsive for periods of time — sometimes 30 seconds to several minutes — while Windows waits for the disk to deliver the paged data. This feels identical to a freeze but is actually Windows working, just extremely slowly.
Check whether this is the cause: open Task Manager during a freeze (if possible — sometimes Ctrl+Alt+Delete still works even when the taskbar is unresponsive) and look at Memory and Disk columns. If Memory shows 95%+ and Disk shows near 100%, paging is the issue. Solutions: close memory-hungry applications before doing intensive work, add more RAM if the machine consistently hits high memory usage, or upgrade from HDD to SSD to make paging less catastrophically slow. An SSD upgrade specifically eliminates the “pseudo-freeze” category of slowdowns almost entirely, because SSD paging is fast enough to be largely invisible. Related: Excel Keeps Freezing.
