Kernel-Power 41 (63): What It Means and How to Fix It
Your PC restarted with no warning, and Event Viewer shows a critical Kernel-Power error with Task Category 63. Here is what that entry actually tells you, the five real causes ranked by likelihood, and the fixes that work for each one.
Kernel-Power Event 41 with Task Category 63 means Windows woke up and discovered it had not shut down properly last time. It is not a diagnosis on its own, it is Windows telling you something knocked the system out before it could log what happened.
The most common causes, in order, are a failing or undersized power supply, a graphics driver crash, overheating, a Fast Startup or Windows Update conflict, and, less often, failing RAM or a loose connection. Check BugcheckCode in your log first: a value of 0 points toward power loss, while any other number points toward a driver or software crash.
What Kernel-Power 41 actually means
Every time Windows starts up, it checks whether the previous session ended cleanly. If it did not, and the shutdown looks abrupt rather than a normal restart or sleep, Windows logs Event ID 41 from the source Microsoft-Windows-Kernel-Power, filed under Task Category 63. The description reads something like "the system has rebooted without cleanly shutting down first."
The important thing to understand is what this event is not. It is not a description of the cause. Windows did not have time to write a proper crash report, because whatever happened cut the system off before the operating system could react. Event 41 is closer to a smoke alarm than a fire investigation report: it tells you something went wrong and roughly when, but the actual cause has to be worked out from the surrounding clues in your log and your system's behavior pattern.
That is exactly the gap a tool like ReadMyLog is built to close. Paste your full Event 41 entry, including the BugcheckCode and any parameter values, and it reads the specific fields rather than repeating the generic Microsoft description back at you.
How to read your own Event 41 entry
Open Event Viewer with Win + R, type eventvwr, and press Enter. Under Windows Logs, click System, then look for a Critical entry with Source Kernel-Power. Click the Details tab and switch to XML View to see the full data, including the fields that actually matter for diagnosis.
Two fields do most of the work:
- BugcheckCode: If this reads
0, Windows never got a chance to record a specific crash reason, which usually points to a sudden power interruption or a hard system hang rather than a software fault. If it shows any other number, that number corresponds to a real Windows bugcheck (stop code), and the crash was caused by something Windows itself detected, typically a driver. - BugcheckParameter1 through 4: When the bugcheck code is non-zero, these parameters often point to a memory address or a specific subsystem, which narrows down whether the fault sits in a graphics driver, a storage driver, or elsewhere.
If your entry shows all zeros across the board, that pattern by itself is meaningful. It is the signature of either a hard power loss or a freeze so complete that Windows never regained enough control to log anything more specific.
Paste your full Event 41 entry and get the specific cause named, not a generic checklist.
Decode it with ReadMyLogThe five real causes, ranked by likelihood
Based on the pattern of reports from PC builders, gamers, and the repair cases I have personally worked through, these five causes account for the overwhelming majority of Kernel-Power 41 events. They are ordered from most to least common.
A power supply that cannot keep up under load
This is by far the most frequent cause, especially when the crash happens during gaming or another GPU-intensive task rather than at idle. A power supply that is undersized for the graphics card, aging, or has a failing capacitor can momentarily drop voltage when demand spikes, which cuts the system without warning. If your crashes cluster around load-intensive moments and your BugcheckCode reads 0, this is the first thing worth ruling out.
A graphics driver crash Windows could not recover from
Modern GPU drivers can usually recover from a hang on their own, showing a brief flicker and continuing. Occasionally the recovery itself fails and takes the whole system down with it, which produces an Event 41 rather than the more informative driver timeout warning you would otherwise see. This shows up more often after a recent driver update, or on older driver versions running new games.
Overheating that triggers a hard shutdown
Motherboards and CPUs have thermal protection that cuts power abruptly rather than risk damage, which looks identical to a power failure from Windows' point of view. Dust buildup, dried thermal paste, or a fan that has quietly stopped spinning are the usual culprits, and this cause tends to correlate with warmer rooms, longer play sessions, or a system that has not been cleaned in a year or more.
Fast Startup or a Windows Update conflict
Fast Startup hibernates the kernel session instead of fully shutting it down, and on some hardware combinations this process fails silently, which Windows then reports as an unclean shutdown the next time it boots. This cause is more likely if your crashes happen specifically around shutdown or wake from sleep, rather than during active use, and it is one of the few causes on this list that is a pure software fix.
Failing RAM or a loose internal connection
Least frequent, but worth ruling out if the first four do not match your pattern. A RAM module developing errors under load, or a power or PCIe cable that is not fully seated, can produce the exact same signature. This is more likely if the problem started right after moving the PC, opening the case, or a recent hardware change.
Step by step: how to fix it
Work through these in order. Each step is quick to test and rules out one cause before you move to the next, rather than changing several things at once and not knowing what actually fixed it.
- Check Task Manager during the activity that triggers the crash Watch CPU, GPU, and memory usage while doing whatever normally causes the restart. If usage spikes to the low 90s or higher right before a crash with BugcheckCode 0, that supports the power supply theory.
- Update your graphics driver through the manufacturer's site, not Windows Update Download directly from NVIDIA, AMD, or Intel rather than relying on the automatic driver Windows installs, since manufacturer packages tend to be more current and more stable for this specific issue.
- Check your temperatures with a monitoring tool HWMonitor or a similar free tool run alongside your normal use for a day shows whether CPU or GPU temperatures are climbing into unsafe territory before a crash. Anything consistently above 85 to 90°C under load is worth investigating further.
- Disable Fast Startup Open Control Panel, go to Power Options, then Choose what the power buttons do, and untick Turn on fast startup. This costs a few seconds on your next boot and rules out an entire category of cause at zero risk.
- Reseat your power cables and RAM With the PC unplugged, firmly press down on RAM sticks until the clips click, and reseat the main 24-pin and CPU power connectors. A cable that looks plugged in is not always making full contact.
- If the pattern still points to the PSU, test with a different unit This is the step that actually costs money, so it is deliberately last. If a borrowed or spare power supply of similar or higher wattage stops the crashes, you have your answer with certainty.
Never open your PC case or handle internal cables while the system is plugged in, even if it is switched off. Unplug it fully at the wall first.
When it is not a software problem
A few signs point clearly toward failing hardware rather than something a setting or a driver update can fix. If the crashes happen at idle, with nothing demanding, that rules out a load-related power supply issue and shifts suspicion toward a motherboard or RAM fault. If they happen at a fairly consistent time interval regardless of what you are doing, that pattern is more typical of a thermal or hardware fault than a driver crash, which tends to be tied to specific triggers like a particular game or app.
If you have worked through the software steps above and the crashes continue with no clear pattern, that is the point where a proper hardware diagnostic, either your own multimeter and stress-testing tools, or a technician, becomes the right next move rather than continuing to guess.
Frequently asked questions
It means Windows woke up and found that it had not shut down properly the last time it ran. The system did not get a chance to log what caused the shutdown, so Event 41 is a symptom, not a diagnosis. The real cause is almost always one of five things: a failing or undersized power supply, a driver crash, overheating, a Windows update or fast startup conflict, or in rare cases a failing component.
A BugcheckCode of 0 means Windows did not get to write a crash dump before it lost control. This usually points away from a software crash and toward a sudden loss of power or a hard hang, since a normal driver crash almost always leaves a bugcheck code behind.
No. A failing or undersized PSU is the single most common cause, especially under gaming load, but it is not the only one. Graphics driver crashes, overheating, and a botched fast startup or update are common secondary causes with their own distinct patterns you can check before replacing hardware.
Yes, in many cases. If the cause is fast startup, an outdated graphics driver, or a Windows power management setting, a software fix resolves it completely with no new hardware. If the cause is a failing PSU or overheating component, software changes can reduce frequency but will not fix the underlying hardware fault.
Not sure which of these five causes fits your system? Paste your exact log and get a specific verdict instead of a guess.
Decode your Event 41 now