Steve H.

Reviewed for technical accuracy

Specialities: Sudden desktop restart triage (Windows 10/11), Event Viewer & Reliability Monitor analysis, overheating vs power loss isolation, GPU driver crash patterns.

Desktop keeps rebooting randomly — how to tell power loss vs overheating vs driver crash (safe checks before BIOS changes or replacing parts)

Technician

Hey! A desktop computer restarting without warning can feel scary, but it usually becomes manageable once we identify what type of restart it is. “Random reboot” can mean anything from a Windows crash to a brief power interruption or a temperature protection event.

Before we touch settings, can you describe what you see in the moment?
• Blue screen / stop code even for a second?
• Freeze first (mouse stuck) and then restart?
• Instant black screen like the power got cut, then it boots back up?

Also: Windows 10 or 11? And did this start after something specific — a Windows Update, a graphics driver update, moving the tower, a new game/app, or totally out of nowhere?

Customer

Windows 11. It’s basically a black screen and then it restarts. I haven’t noticed a blue screen. Started about a week ago. I didn’t change hardware.

Technician

Got it — that “instant black screen → reboot” pattern is important. People search this as PC rebooting randomly, Windows 11 restarts without warning, desktop restarts by itself, or unexpected restart no blue screen.

Two quick questions, in a slightly unusual order (because it helps separate the buckets early):
1) Does it ever restart while completely idle (just sitting on the desktop), or mostly under load (gaming, streaming, lots of tabs)?
2) When it happens, do you hear a click from the PC or power strip, like power briefly dropped — or does it just go dark and come back without any sound change?

And last: are you plugged into a wall outlet directly, a power strip, or a UPS?

Customer

Mostly while gaming, but it did happen once just browsing. No clear click. It’s plugged into a power strip.

Technician

Thanks — that mix (mostly under load, sometimes light use) is common when the underlying issue is borderline. The safest next move is to pull evidence from Windows (no “fixes” yet), so we don’t guess and accidentally make it harder to diagnose.

Safe check #1: Reliability Monitor
In the Windows search box, type reliability → open View reliability history. Click the day/time of the restart. Look for lines like “Windows was not properly shut down,” “Hardware error,” or any driver/app crash right before it.

Safe check #2: Event Viewer (System)
Open Event Viewer → Windows Logs → System. Find the same time and see if there’s a Critical entry (often Kernel-Power, Event ID 41). That just means Windows didn’t get a clean shutdown — it can happen from power loss, a hang, or a crash.

If you can paste the exact wording of the Reliability Monitor red X (or the event name/ID you see right at the reboot time), we’ll know which direction is most likely: power path, overheating, driver crash, or memory stability.

Customer

Reliability Monitor shows “Windows was not properly shut down.” Event Viewer has Kernel-Power and Event ID 41 at the same time.

Technician

Perfect — that confirms the category: an unexpected restart, not a normal Windows restart. Now the goal is to narrow it without doing anything irreversible.

Here are the big “buckets” we’re separating (still diagnostic only):
1) Power interruption / power delivery (power strip/UPS/outlet/cable, or PSU not coping with brief spikes).
2) Overheating (CPU or GPU hitting limits under load; gaming is a common trigger).
3) Driver or system crash (sometimes Windows auto-restarts so you never see a stop code).
4) Memory stability (RAM/XMP/EXPO can be “almost stable” until a game pushes it).

Let me ask a “risk” question now: any burning smell, visible sparking, or the power strip getting hot? If yes, stop and don’t keep testing. If no, we can safely do a couple of isolation checks.

Also, does your PC have a dedicated GPU (NVIDIA/AMD) and do the fans get very loud right before it reboots?

Customer

No burning smell or anything like that. I do have a dedicated GPU. Fans get pretty loud during games, and it restarted twice while gaming.

Technician

That makes heat + power spikes the two most likely suspects, but we’ll prove it with safe tests instead of guessing.

Safe isolation test A: power path
For one controlled gaming session, plug the desktop directly into a wall outlet (bypass the power strip). This doesn’t “fix” the PC — it just tells us if the strip is contributing to instability. Also make sure the PC power cable is firmly seated at both ends.

Safe isolation test B: load pattern
Note whether it restarts after a similar amount of time under load (example: 10–25 minutes into gaming) or if it’s truly random. Overheating tends to correlate with time + load. Power/PSU issues can feel more instant and unpredictable (especially when the GPU suddenly boosts).

Safe observation C: hidden blue screen
Windows can be set to automatically restart on system failure, so you never see the stop code. We don’t need to change a bunch of settings — just confirm whether Windows is currently hiding a valuable clue.

One more quick thing people forget: is the PC dusty inside (front intake, CPU cooler fins, GPU fans)? Dust can raise temps fast even if nothing “broke.”

Customer

It’s definitely dusty. I can clean it carefully. I can also try the wall outlet test. How do I check if Windows is hiding a blue screen?

Technician

You’re thinking about it exactly right — we want information, not a pile of random “fixes.” To check the “auto-restart” behavior (so you can actually see a stop code if it happens):

Go to Settings → System → About → Advanced system settings (or open “View advanced system settings” from search). Under Startup and Recovery, click Settings, and look for Automatically restart under “System failure.”

You don’t have to change it if you’re not comfortable — the key is simply knowing whether a crash could be getting hidden. If it’s enabled, that explains why you might never see a BSOD even if the root cause is a driver crash.

For the dust clean: power the PC off, unplug it, and use short air bursts. If you see fans spinning too fast from the air, gently hold the fan blades still while cleaning so they don’t overspin. That’s just safety/maintenance — not a “guaranteed fix.”

After you do the wall-outlet test + basic dust clean, keep an eye on whether the restart behavior changes. If it still restarts, the next safest step is to capture the exact warnings right before Event ID 41 (sometimes you’ll see driver resets, disk warnings, or WHEA entries).

Customer

I saw some warnings around the restart time but I didn’t understand them. I think one said WHEA-Logger. Could RAM or XMP cause a reboot like this without a blue screen?

Technician

Yes — and you’re also right to treat WHEA-Logger as “meaningful,” not just noise.

WHEA-Logger is Windows recording a hardware-related warning (often CPU, memory controller, PCIe/GPU link, or power/voltage instability). It still doesn’t automatically name “the bad part,” but it strongly supports “instability under load” rather than a simple app problem.

And RAM/XMP can contribute to sudden restarts. Even if XMP is commonly used, some systems end up “almost stable” — fine for weeks, then a Windows update, a new game build, or warmer temps pushes it over the edge. Sometimes you’ll see a blue screen; sometimes it can look like a hard reboot depending on what fails and whether Windows can catch it cleanly.

For now, the safest next action is not “turn XMP off” or “replace RAM.” It’s: capture the exact WHEA event details (event ID + the short description) and note what you were doing at that moment. That tells us whether we should lean toward power/PSU behavior, GPU driver/hardware, or memory stability — without making big changes prematurely.

Quick answer: the accepted fix

Accepted first fix for desktop computer restarts without warning: avoid jumping straight to BIOS tweaks, “network/Windows resets,” or swapping parts. Start with safe diagnostics that keep your system stable. Check Reliability Monitor for the exact restart timestamp and review Event Viewer → System for Kernel-Power (Event ID 41) plus any warnings immediately before it (driver resets, WHEA-Logger, disk events). Then run one clean isolation test: bypass the power strip by plugging the PC directly into a wall outlet for a controlled session and see if the pattern changes. In parallel, look for a heat correlation (restarts during gaming with loud fans, dusty intake/coolers). If Windows is set to auto-restart on failure, you may be missing a stop code — knowing that setting helps you tell a crash from a pure power cut. This evidence-first approach helps you narrow the cause safely before any “big” changes.

FAQ

Sources

• Microsoft Learn — Event ID 41: The system has rebooted without cleanly shutting down first
• Microsoft Tech Community — Using Reliability Monitor for troubleshooting
• Microsoft Support — Troubleshooting Windows unexpected restarts and stop code errors
• MemTest86 — Official site (memory testing overview)

Related articles

• Lenovo computer restarts without warning — easy help for seniors
• Dell computer restarts without warning — quick help (safe checks)
• Asus computer restarts without warning — simple explanation (safe help)
• Asus computer restarts without warning — quick help (safe checks)
• Desktop screen stays black but power light is on — what to do next (safe checks)
• Windows computer screen stays black but power light is on — what to do next
• Windows computer screen stays black but power light is on — easy help for seniors (what to check first)
• Lenovo computer blue screen error (BSOD) — what to check first without making it worse

Steve ● replies in < 1 min

Technician Expert

⭐⭐⭐⭐⭐ Trusted by 26,294 customers

Alex · Expert's Assistant

Start Chat

5 Tech Support Specialists are online right now.