How to Fix a Computer That Randomly Restarts

Published 2026-10-14 · LuxuriousComputers Repair

Your computer restarts without warning — no blue screen, no error message, it just suddenly goes black and boots back up as if someone pressed the reset button. This is different from a controlled restart (where Windows shows the shutting down screen) and different from a BSOD (where Windows displays an error before restarting). A sudden restart with no warning usually points to a hardware problem: power delivery failure, overheating, or a faulty component triggering a protection mechanism.

Check if Windows is hiding the error

By default, Windows is configured to automatically restart when a BSOD occurs. This means the blue screen may be appearing for less than a second before the restart — too fast to see. To check whether there is actually a BSOD behind the restart:

  1. Right-click This PC (or My Computer), select Properties
  2. Click Advanced System Settings on the left
  3. Under Startup and Recovery, click Settings
  4. Under System Failure, uncheck Automatically Restart

With this setting disabled, the next time Windows crashes it will stay on the blue screen instead of restarting. This lets you read the stop code and search for the specific error. If the computer still restarts instantly without showing any blue screen, the problem is below the operating system level — hardware or power.

Check the Event Viewer

Even if you do not see a blue screen, Windows logs crash information. Open Event Viewer (type it in Start menu search), expand Windows Logs, System. Look for Critical events (red circle with X) around the time of the last restart. Common entries:

  • Kernel-Power Event ID 41: The system rebooted without cleanly shutting down first. This is the generic log entry for an unexpected restart and does not by itself tell you the cause — it just confirms the restart was not intentional
  • BugCheck entries: If there was a BSOD before the restart, the BugCheck event will contain the stop code and faulting module. This tells you exactly what crashed
  • WHEA-Logger: Hardware error events logged by the Windows Hardware Error Architecture. These point to CPU, RAM, or PCIe bus errors

Power supply problems

A failing or insufficient power supply is the single most common hardware cause of random restarts. The computer draws more power than the PSU can deliver (typically during gaming, rendering, or other heavy loads), and the PSU either shuts down briefly or its output voltage drops below what the motherboard needs, triggering an immediate reboot.

Signs the PSU is the cause

  • Restarts happen mostly under load — gaming, video editing, running benchmarks, or when multiple drives and USB devices are all active
  • The restart is instant — no warning, no shutdown animation, just black screen to boot sequence. This pattern suggests power was interrupted, not that the OS decided to restart
  • The restarts started after adding new hardware (a more powerful GPU, more drives, more fans) that increased the power draw beyond what the PSU can sustain
  • The PSU is more than 5 years old. PSU capacitors degrade over time, and the actual wattage a PSU can deliver decreases as it ages

What to do

If you have access to a known-good PSU of equal or greater wattage, swap it temporarily and see if the restarts stop. If they do, replace the PSU. When choosing a replacement, calculate your system's actual power needs (CPU TDP plus GPU TDP plus roughly 100 watts for everything else) and add 20 to 30 percent headroom.

Overheating

CPUs and GPUs have built-in thermal protection. When the temperature exceeds a critical threshold (typically 100 to 105 degrees Celsius for CPUs and 105 to 110 for GPUs), the system shuts down or restarts immediately to prevent physical damage. This is a safety feature, not an error — the hardware is working correctly by refusing to operate at a temperature that would destroy it.

Checking temperatures

Install HWiNFO and monitor CPU and GPU temperatures during the tasks that trigger restarts. Key temperature readings to watch:

  • CPU Package temperature: Should stay below 85 to 90 degrees Celsius under sustained load. Above 95 is in the danger zone
  • GPU temperature: Normal gaming range is 65 to 85 degrees depending on the cooler. Above 95 is problematic
  • CPU VRM temperature: The voltage regulators on the motherboard that power the CPU. If VRM temperature exceeds 100 to 110 degrees, the motherboard may throttle or shut down. VRM overheating is common in budget motherboards paired with high-power CPUs

Fixing overheating

  • Clean dust from the CPU cooler, GPU heatsink, and case fans. Dust is the most common cause of gradual temperature increase over months. Compressed air clears most buildup
  • Check that all fans are spinning. A dead CPU fan, a dead case fan, or a GPU with one of its fans failed can push temperatures into the shutdown range
  • Replace thermal paste. The paste between the CPU and cooler (and between the GPU die and its heatsink) dries out over 3 to 5 years. Dried paste transfers heat poorly. Replacing it with fresh paste can drop temperatures by 10 to 20 degrees
  • Improve case airflow. Make sure the case has both intake fans (front) and exhaust fans (rear and top). Cases with restricted airflow build up internal heat that affects every component

RAM problems

Defective RAM can cause random restarts, especially when the system accesses a specific faulty memory address. Unlike RAM errors that cause BSODs with specific stop codes, some RAM failures cause an immediate system reset with no error logged — the memory controller detects an unrecoverable error and signals a hardware reset.

Test RAM with MemTest86 (boot from USB, run at least 4 passes). If errors are found, test each RAM stick individually to identify the faulty stick. If you recently installed new RAM or changed RAM settings in BIOS (like enabling XMP/EXPO profiles), try running the RAM at default speeds — XMP profiles are technically overclocks and are not always stable on every motherboard.

BIOS and firmware settings

Disable automatic restart on failure

Some motherboards have their own restart-on-error settings independent of Windows. Check your BIOS settings for options related to automatic restart, hardware monitoring thresholds, and power failure recovery. Set the system to halt on error rather than restart so you can observe what happens when the next crash occurs.

Reset overclocks

If the CPU, RAM, or GPU has been overclocked, reset all clocks to default and test for stability. An overclock that was stable when first configured can become unstable as components age, when ambient room temperature increases, or after BIOS updates change voltage delivery behavior.

Update BIOS

Motherboard BIOS updates often fix stability issues, improve memory compatibility, and resolve power management bugs that can cause random restarts. Check the motherboard manufacturer's website for the latest BIOS version. Follow the manufacturer's BIOS update instructions carefully — a failed BIOS update can make the computer unbootable.

Software causes

While random restarts are more often hardware-related, some software issues can cause them:

  • Driver conflicts: A driver crash can trigger a restart fast enough that no blue screen is visible even with automatic restart disabled. Update or cleanly reinstall GPU, chipset, and storage drivers
  • Windows Update restart policy: Windows can restart the computer to install updates. Check Windows Update, Active Hours settings and Update History to see if the restarts coincide with update installations
  • Scheduled tasks or group policy: In enterprise environments, group policies can trigger scheduled restarts. Check Task Scheduler for any restart-related tasks

Random restart diagnosis in Marion

If your computer keeps restarting randomly and you cannot pin down the cause, LuxuriousComputers in Marion can run hardware diagnostics — PSU load testing, thermal monitoring, memory testing, and event log analysis — to identify the exact component or condition causing the restarts and fix it.

← Back to Blog