When This Happens
You're in the middle of something – maybe a game, a video render, or just browsing. Then, without a flicker, your PC shuts off and starts again. No blue screen. No error message. Just a black screen and the BIOS logo again. You check Windows Event Viewer and see "Kernel-Power 41" errors. That's the generic "I lost power" event. It tells you nothing useful.
This happens most often under load, but I've seen it during idle too. The triggers are usually: high CPU or GPU usage, a sudden power spike, or even just moving your desk slightly. The culprit here is almost always RAM instability or a voltage issue on the motherboard.
Root Cause
Think of RAM as the short-term memory for your CPU. If it's got bad timings, wrong voltage, or a physical defect, the CPU can't read or write data reliably. When that happens, the system panics and restarts. No time for a blue screen – it just dies.
Motherboard side? The voltage regulators (VRMs) that feed your CPU and RAM can sag under load. If they can't keep up, the CPU sees a brownout and restarts. Same story. No warning.
Heat can also mess with RAM. If your case airflow is bad or your RAM sticks are running hot (over 85°C), they start throwing errors. That's another trigger.
Fix It – Step by Step
- Reset your BIOS to defaults. Enter BIOS by pressing Del or F2 at startup. Load optimized defaults. Save and exit. This clears any bad overclock or XMP profile you tried. I've seen XMP cause random reboots on cheap motherboards. Try without it first.
- Run a memory test. Download MemTest86 (free version) and put it on a USB stick. Boot from it. Let it run for at least 4 passes. One pass is about 30 minutes on modern systems. If you get any red errors – your RAM is bad. Swap sticks one at a time to find the bad one.
- Check your RAM voltage and timings. Look at the sticker on your RAM sticks. It says something like "1.35V" or "1.2V". Go into BIOS and set the voltage manually. Also set the primary timings (like 16-18-18-36) exactly as printed. Auto settings can be wrong. I've fixed dozens of reboots by just doing this.
- Test with one stick of RAM. Power down, remove all but one stick in slot A2 (usually the second slot from the CPU). Boot and stress test. If it's stable, swap sticks. If only one stick causes reboots, you found the bad one. Replace it.
- Update your motherboard BIOS. Manufacturers fix memory compatibility issues all the time. Go to your motherboard manufacturer's support page, download the latest BIOS, and flash it. Don't skip this – it's saved me hours of debugging.
- Check your PSU if nothing else works. A dying power supply can cause random restarts too. Use a multimeter on the 24-pin connector (yellow wires should be 12V, red 5V) or swap with a known good PSU. I've nailed many reboots to a PSU that was 0.2V low on the 12V rail.
If It Still Fails
If you've tried all that and the PC still restarts, you're looking at a motherboard issue. The VRMs might be shot. Check for bulging capacitors on the motherboard – look for puffy tops or leaked electrolyte. That's a dead board.
Another thing: reseat your CPU. Loose mounting pressure can cause unstable voltage delivery. Take off the cooler, pull the CPU, check for bent pins, then reinstall. Also check your CPU cooler mount – if it's uneven, the CPU overheats and the motherboard triggers a restart. Run HWMonitor to check CPU temps during stress. If it hits 95°C+, that's your problem.
Last resort: try a different power outlet. Bad grounding in your house can cause random resets. I've seen that exactly once in 14 years, but it's worth a check.
If none of that works, replace the motherboard. Start with a B660 or B760 chipset board for Intel, or a B550 for AMD. They're cheap and reliable. Swap the RAM too if you're still unsure. You'll probably fix it by then.