Why Does My PC Restart Randomly Under Load? (September 2026)

Most of the time, a PC that restarts randomly under load has a power supply that cannot deliver stable voltage during peak demand, a CPU or GPU hitting thermal shutdown limits, or faulty RAM causing memory errors. If your screen goes black and the system reboots with no blue screen or error message, power delivery is the prime suspect.

I have spent years helping users in buildapc and computerhelp communities diagnose this exact problem. The frustration is real. You run a stress test and it passes. You play a game and the PC restarts with zero warning. No crash dump, no BSOD, no clue what just happened.

This guide covers every cause of random restarts under load, including the ones no other article talks about. Specifically, I will break down transient load spikes, undervolting instability, and motherboard power delivery issues that explain why your system crashes in games but passes every diagnostic you throw at it.

Here is the short version of what causes a PC to restart randomly under load:

• Power supply (PSU) cannot handle peak power draw or is degrading
• CPU or GPU hits thermal shutdown temperature (usually 90 to 105 degrees Celsius)
• Transient load spikes overwhelm a PSU during rapid usage changes
• Motherboard VRM or CPU power connector (4-pin vs 8-pin) limits delivery
• Faulty or unstable RAM, especially with XMP profiles enabled
• Aggressive undervolting makes the system unstable under load
• Corrupted drivers, outdated BIOS, or Windows settings triggering reboots

By the end of this article, you will know exactly which of these applies to your situation and what to do about it. Let us start with a symptom-to-cause mapping table so you can jump straight to your specific scenario.

Quick Summary: Symptom-to-Cause Mapping

Different restart patterns point to different root causes. Before diving into detailed explanations, use this table to narrow down your problem based on what you are actually experiencing.

Restart Pattern	Likely Cause	First Check
Restart during gaming, no BSOD, passes FurMark	Transient load spike or PSU wattage limit	PSU wattage and 12V rail stability in HWiNFO
Restart only in certain games, not others	Power delivery during variable load patterns	Frame rate caps and power draw logs
Restart during load transitions (lobby to game, alt-tab)	Transient power spike overwhelming PSU	PSU age, wattage rating, and tier quality
Restart with no crash dump file generated	Hardware power-off, not a software crash	Event Viewer for Kernel Power 41 errors
Restarts getting more frequent over weeks	PSU capacitor degradation	PSU age, replace if over 4 to 5 years old
Restart after enabling XMP or overclocking RAM	Memory instability	Disable XMP, run memtest86
Restart after undervolting GPU or CPU	Undervolt too aggressive	Revert undervolt to stock, retest
Restart with high temps before crash	Thermal shutdown	HWiNFO temperature logging
Restart only on budget motherboard with 4-pin CPU power	VRM power delivery limit	Check CPU power connector type
Restart stopped after removing surge protector	Faulty surge protector or power strip	Plug directly into wall outlet

If your symptom is not listed here, do not worry. I cover every scenario in detail below, starting with the most common cause.

Power Supply Issues: The Number One Cause of Random Restarts Under Load

Your power supply is the single most common reason a PC restarts randomly under load. When your CPU and GPU spike to maximum draw, the PSU must deliver clean, stable voltage across its rails. If it cannot, the system powers off instantly to protect itself.

This is why most random restarts under load produce no blue screen and no crash dump. The system does not crash in software. It loses power at the hardware level and reboots.

Signs Your PSU Is the Problem

Several patterns strongly indicate a PSU issue. Restarts that happen only under load and never at idle are the classic sign. Restarts that produce no BSOD or crash dump point to power delivery. Restarts that get progressively worse over months suggest capacitor aging inside the PSU.

Another telltale sign is when your PC passes synthetic stress tests like FurMark or AIDA64 but restarts during actual gameplay. This happens because stress tests apply a steady, predictable load, while games cause rapid power draw changes that expose PSU weaknesses. I cover this in detail in the transient load section below.

How to Check if Your PSU Is Failing

Start by checking your PSU wattage against your system requirements. A common mistake is using a PSU that barely meets the rated wattage of the system. Your PSU should have at least 20 to 30 percent headroom above your system’s peak power draw. If your system pulls 450W at full load, a 500W PSU gives you almost no margin for transient spikes.

Use HWiNFO to monitor your 12V, 5V, and 3.3V rail voltages under load. The 12V rail should stay between 11.4V and 12.6V. If it drops below 11.4V during a game or stress test, your PSU is struggling. Any reading outside ATX specification ranges is a red flag.

Also consider the quality tier of your PSU. Budget units like the Thermaltake Smart series (80 Plus White) are known for inconsistent power delivery under transient loads. Forum users repeatedly identify these entry-level units as the cause of restart problems. A higher-quality unit from a reputable brand with proper voltage regulation will handle transient spikes far better.

PSU Degradation Over Time

PSUs do not last forever. Electrolytic capacitors inside the unit degrade over years of use, reducing the PSU’s ability to deliver clean power during peak demand. If your system ran fine for a year or two and then started restarting under load, PSU aging is a prime suspect.

This progressive failure pattern is one of the most confusing aspects for users. The PSU worked before, so they assume it is not the problem. But a PSU that is three to five years old, especially a budget model, may have lost significant capacity. I have seen cases where a Thermaltake Smart 600W that ran fine for 18 months started causing restarts as capacitors degraded.

If your PSU is over four years old and you are experiencing random restarts under load, replacing it should be one of your first actions. Even if it is not the cause, an aging budget PSU puts your other components at risk.

Overheating and Thermal Shutdown

Overheating is the second most common cause of random restarts under load. Modern CPUs and GPUs have built-in thermal protection. When they reach a critical temperature, typically between 90 and 105 degrees Celsius, they shut the system down to prevent permanent damage.

This shutdown is instant. No blue screen, no error message, no crash dump. The screen goes black and the system reboots. This makes it easy to confuse with PSU problems, which is why temperature monitoring is essential.

How to Check Your Temperatures

Download HWiNFO or HWMonitor and log temperatures while gaming or running a stress test. Pay attention to the maximum temperatures, not just the current readings. Key thresholds to watch:

• CPU thermal shutdown: typically 100 to 105 degrees Celsius (varies by model)
• GPU thermal limit: typically 83 to 90 degrees Celsius before throttling, 105 to 110 degrees for hard shutdown
• VRM temperatures on the motherboard: over 100 degrees Celsius can cause instability
• Hotspot temperature: more important than core temperature on modern GPUs

If your CPU is hitting 95 degrees or higher before a restart, thermal shutdown is almost certainly the cause. Check your CPU cooler mounting, thermal paste application, and case airflow.

Common Overheating Causes

Dust buildup is the most common overheating cause. Over months of use, dust clogs heatsink fins, case fans, and GPU coolers. This restricts airflow and causes temperatures to climb gradually. If your temperatures have been creeping up over time, dust is likely the reason.

Dried or poorly applied thermal paste is another frequent cause. Thermal paste degrades over years and loses its heat transfer efficiency. If you are using the stock cooler that came with your CPU and it has been two or three years, repasting can drop temperatures by 5 to 15 degrees.

VRM overheating is a less obvious but important cause. The VRM (Voltage Regulator Module) on your motherboard delivers power to your CPU. If VRM components overheat, they can throttle or cause system instability even when your CPU temperature looks fine. Budget motherboards with minimal VRM heatsinks are especially vulnerable.

Transient Load Spikes: The Hidden Culprit Nobody Explains

This is the section I wish every troubleshooting guide included. If your PC passes FurMark and AIDA64 but restarts during games, transient load spikes are almost certainly the cause. No other article I have found explains this properly.

Here is what happens. When you run a synthetic stress test, the load is steady and predictable. Your GPU sits at 100 percent usage drawing a consistent amount of power. Your PSU handles this fine because the power draw is constant.

Games are different. GPU usage fluctuates constantly as you move between menus, lobbies, cutscenes, and gameplay. When GPU usage spikes from 30 percent to 100 percent in a fraction of a second, the power draw spikes too. This is a transient load spike.

Why Transient Spikes Cause Restarts

Your PSU has internal capacitors that smooth out voltage delivery. During a transient spike, the PSU must instantly deliver a large burst of current. If the PSU’s capacitors cannot handle the sudden demand, voltage drops below the level your components need to stay stable. The system powers off.

This explains a pattern I have seen dozens of times in forums. A user reports that Fallout 4 causes restarts at 100 percent GPU usage, but Resident Evil 3 runs fine at the same usage level. The difference is not the sustained load. It is how rapidly the load changes during gameplay.

It also explains why restarts happen during load transitions. Going from a game lobby to a match, alt-tabbing, or experiencing a scene change can trigger a sudden power draw shift. These transition moments are exactly when transient spikes occur.

Modern GPUs and Transient Spikes

Modern GPUs, especially higher-end models, are notorious for transient power spikes. A GPU rated at 250W sustained draw can briefly pull 350W or more during a transient spike. If your PSU was sized for the sustained draw without headroom for transients, those spikes will cause restarts.

This is why I recommend at least 20 to 30 percent headroom on your PSU wattage. If your system draws 500W at sustained full load, you want a 650W or 700W PSU to absorb transient spikes without breaking a sweat.

The fix for transient load restarts is straightforward. Replace the PSU with a higher-quality, higher-wattage unit that can handle peak transient loads. Frame rate caps can also help by reducing the severity of load transitions, but they treat the symptom, not the cause.

Motherboard Power Delivery: 4-Pin vs 8-Pin CPU Connectors

This is another topic no competitor covers. Your motherboard’s CPU power connector plays a significant role in power delivery stability, and budget boards with 4-pin connectors are a known cause of restart problems under load.

Most motherboards have either a 4-pin or 8-pin EPS power connector for the CPU. An 8-pin connector can deliver up to 336W to the CPU. A 4-pin connector delivers half that, up to 168W. For lower-power CPUs, 4-pin is fine. For higher-power chips, especially under load, 4-pin can become a bottleneck.

The Budget Board Problem

Budget motherboards like the Gigabyte H310M often feature 4-pin CPU power connectors and minimal VRM cooling. Forum users have specifically identified this combination as a source of restart problems. The VRM components are delivering power at their limit, and the 4-pin connector restricts how much clean power reaches the CPU.

When the CPU demands more power than the VRM and 4-pin connector can deliver, the system becomes unstable. This can manifest as a restart, a freeze, or unpredictable behavior under load.

VRM Quality Matters

Even if your motherboard has an 8-pin connector, VRM quality varies dramatically between board tiers. High-end motherboards have robust VRM designs with multiple phases, quality components, and large heatsinks. Budget boards cut corners here, using fewer phases and minimal cooling.

Under sustained load, cheap VRMs heat up and their power delivery becomes less stable. This can cause restarts even when your CPU temperature is within safe limits. The CPU is not overheating, but the VRM feeding it power is struggling.

If you are running a power-hungry CPU on a budget motherboard, VRM and power connector limitations may be contributing to your restart problems. Unfortunately, the only real fix is a motherboard upgrade. There is no software solution for inadequate hardware power delivery.

RAM and Memory Instability

Bad RAM can absolutely cause random restarts under load. Memory errors become more frequent and severe when the system is working hard, which is why RAM problems often appear only during gaming or intensive tasks.

Memory instability is especially common when XMP (Extreme Memory Profile) is enabled. XMP runs your RAM at speeds higher than its base JEDEC specification. If your RAM kit or motherboard cannot handle the XMP settings stably, you get memory errors that can cause crashes, freezes, or restarts.

How to Test Your RAM

The gold standard for RAM testing is memtest86. Download it, put it on a USB drive, and boot from it. Let it run for at least one full pass, though four passes is ideal. Any errors mean your RAM is unstable and needs replacement or configuration changes.

A quicker test is to disable XMP in your BIOS and run at stock JEDEC speeds. If the restarts stop with XMP disabled, your RAM cannot run stably at its advertised XMP profile. You can try manually tuning the speed and voltage, or accept lower speeds for stability.

Also try reseating your RAM sticks. Remove them, clean the contacts with isopropyl alcohol, and reseat them firmly. Try running with one stick at a time to isolate a faulty module. If one stick consistently causes issues while the other does not, you have found your problem.

Undervolting Instability: When Fixes Cause Crashes

Undervolting is a popular technique to reduce temperatures and power draw. You lower the voltage delivered to your CPU or GPU, which reduces heat and can actually improve performance by allowing higher boost clocks without thermal throttling.

But here is the catch. An undervolt that is too aggressive causes instability under load. The component does not have enough voltage to remain stable at high clock speeds, and the system crashes or restarts.

The Forum Experience

This is a well-known phenomenon in enthusiast forums. One user reported that undervolting their GPU by 50mV completely eliminated restarts, pointing to a PSU power delivery issue. The undervolt reduced peak power draw enough that the PSU could handle it. But another user found that pushing their undervolt further caused new restarts, this time from voltage instability rather than power delivery.

This creates a confusing situation. Undervolting can fix restarts caused by PSU limitations. But push it too far and you create a new restart cause. The line between stable and unstable varies by individual chip quality, a concept known as the silicon lottery.

How to Tell if Undervolting Is Causing Restarts

If you applied an undervolt and restarts started or changed pattern, revert to stock voltages and test. If the restarts stop, your undervolt was too aggressive. If the restarts continue at stock, the problem is elsewhere.

When undervolting, test stability with a demanding stress test before trusting it for daily use. Use OCCT or AIDA64 for at least 30 minutes. If it passes, gradually increase the undervolt in small steps, testing each time. The goal is to find the lowest stable voltage, not the lowest voltage that boots.

Driver, BIOS, and Software Conflicts

Software issues are less likely to cause instant power-off restarts compared to hardware problems. But they can cause blue screens that look like restarts if automatic restart is enabled. They can also cause instability under specific load conditions.

GPU Driver Problems

Outdated or corrupted GPU drivers are a common cause of instability during gaming. If your PC restarts only in one specific game or after a driver update, the GPU driver is a likely suspect. Use DDU (Display Driver Uninstaller) to completely remove your current driver, then install the latest version clean.

Driver conflicts between your GPU driver and other system software can also cause issues. Overlays, recording software, and monitoring tools all hook into the graphics pipeline and can cause instability under load.

BIOS Updates

An outdated BIOS can cause stability issues, especially with newer CPUs installed on older motherboards. BIOS updates often include microcode updates that improve CPU stability and power delivery. Check your motherboard manufacturer’s website for the latest BIOS version.

Be careful with BIOS updates. A failed update can brick your motherboard. Follow the manufacturer’s instructions exactly, use a UPS if possible to prevent power loss during the update, and never interrupt the process.

Windows Settings That Trigger Restarts

Windows has several settings that can cause automatic restarts. Fast Startup can cause conflicts with hardware initialization. Automatic Restart on System Failure reboots your PC immediately after a blue screen, which makes it look like a random restart rather than a crash.

To disable automatic restart, go to System Properties, Advanced tab, Startup and Recovery Settings, and uncheck “Automatically restart.” Now if a blue screen occurs, you will actually see it instead of the system rebooting instantly. The error code on the blue screen tells you what failed.

Windows Update can also force restarts. Check Update and Security for pending updates that might be scheduled to install. Set active hours to prevent updates from restarting your PC during use.

Kernel Power 41: What Event Viewer Tells You

When your PC restarts unexpectedly, Windows Event Viewer logs the event. The most common log entry is Event ID 41, Kernel Power. This event means the system rebooted without cleanly shutting down first.

Kernel Power 41 is not a root cause. It is a symptom. It tells you the system lost power or shut down unexpectedly, but it does not tell you why. However, the absence of other error events before the Kernel Power 41 entry is itself diagnostic.

How to Use Event Viewer

Open Event Viewer by pressing Windows key plus R, typing eventvwr.msc, and pressing Enter. Navigate to Windows Logs, then System. Look for Event ID 41 entries around the time of your restart.

Check the events immediately before the Kernel Power 41 entry. If there are error entries about specific drivers or components, those point to the cause. If there is nothing except the Kernel Power 41 entry with no preceding errors, the shutdown was almost certainly a hardware power-off. This means the system lost power before Windows could log anything, pointing to PSU or overheating.

The absence of crash dump files confirms this. When Windows crashes with a blue screen, it writes a dump file. When the system loses power at the hardware level, no dump file is created. If you check C:\Windows\Minidump and find no recent files despite restarts happening, you are dealing with a hardware power issue, not a software crash.

Step-by-Step Troubleshooting Guide

Now let me walk you through a diagnostic process in the right order. Following these steps systematically will save you time and money compared to guessing.

Step 1: Disable Automatic Restart

Before anything else, disable automatic restart in Windows. Go to System Properties, Advanced, Startup and Recovery, and uncheck “Automatically restart.” This ensures that if your restarts are actually blue screens, you will see the error code instead of an instant reboot.

Step 2: Monitor Temperatures Under Load

Download HWiNFO and run it with logging enabled while you game or run the application that triggers restarts. Watch CPU and GPU temperatures, VRM temperatures, and voltage rail readings. If temperatures are hitting thermal limits, overheating is your cause. If voltage rails are sagging under load, your PSU is struggling.

Step 3: Check Event Viewer

Open Event Viewer and look for Event ID 41 Kernel Power entries. Check for any errors preceding them. Note whether crash dump files are being generated in C:\Windows\Minidump. No dump files means hardware power-off, not a software crash.

Step 4: Test Your RAM

Disable XMP in BIOS and test. If restarts stop, XMP instability was the cause. Run memtest86 overnight for thorough testing. Reseat your RAM sticks and test each one individually if errors appear.

Step 5: Revert Any Overclocks and Undervolts

Set everything to stock speeds and voltages in your BIOS. CPU overclocks, GPU overclocks, RAM XMP, and undervolts. Test at completely stock settings. If the restarts stop, one of your modifications was causing instability. Add them back one at a time to identify the culprit.

Step 6: Update Drivers and BIOS

Use DDU to clean-install your GPU driver. Check your motherboard manufacturer’s website for BIOS updates. Run Windows Update to ensure your system is current. These steps eliminate software-related causes.

Step 7: Run System File Checker

Open Command Prompt as administrator and run sfc /scannow. Follow it with DISM /Online /Cleanup-Image /RestoreHealth. These commands repair corrupted Windows system files that can cause instability.

Step 8: Stress Test with the Right Tools

Use the right tool for the right test. Here is a comparison to help you choose:

• OCCT: Best overall stability tester. Tests CPU, GPU, memory, and power delivery. Its power test simultaneously loads CPU and GPU, which is excellent for stressing your PSU.
• AIDA64: Good for sustained stability testing with configurable load types. The system stability test provides detailed real-time monitoring.
• FurMark: GPU-only stress test. Useful for thermal testing but applies an unrealistic steady load that may not reproduce game-related restarts.
• HWiNFO: Not a stress test, but essential for monitoring. Run it alongside any stress test to log temperatures, voltages, and power draw.
• memtest86: Memory-only test. Boot from USB and run for thorough RAM diagnostics.

Run OCCT’s power test for 30 minutes while monitoring voltages in HWiNFO. If your 12V rail sags significantly or the system restarts, your PSU is the problem.

Step 9: Test Without Surge Protectors and Power Strips

Plug your PC directly into the wall outlet. A faulty surge protector or power strip can cause voltage drops under load that trigger restarts. If the restarts stop with a direct wall connection, replace your surge protector.

Step 10: Consider Breadboarding

If you have tried everything above and the restarts continue, breadboarding is the next step. Breadboarding means removing all components from the case and testing them on a non-conductive surface outside the case.

This process isolates short circuits caused by case standoffs, mounting issues, or debris. Remove the motherboard, CPU, cooler, RAM, GPU, and PSU. Set them on a wooden table or cardboard box. Connect only the essentials and test. If the system is stable outside the case but restarts inside it, you have a grounding or short circuit issue.

Step 11: Replace the PSU

If all diagnostics point to power delivery and your PSU is a budget unit, over four years old, or undersized for your system, replace it. This is the most common fix for random restarts under load. Choose a quality unit from a reputable brand with adequate wattage headroom for your system plus transient spikes.

Free Diagnostic Tools You Need

Here is your complete toolkit for diagnosing random restarts. All of these are free:

• HWiNFO: Comprehensive hardware monitoring with logging. Tracks temperatures, voltages, fan speeds, and power draw.
• HWMonitor: Simpler alternative to HWiNFO for basic temperature and voltage monitoring.
• OCCT: All-in-one stability testing for CPU, GPU, memory, and power delivery.
• AIDA64: System stability tester with detailed monitoring (free trial available).
• FurMark: GPU stress test for thermal testing.
• Heaven Benchmark: GPU benchmark that applies variable load, useful for testing transient behavior.
• memtest86: Bootable RAM diagnostic tool. The industry standard for memory testing.
• DDU (Display Driver Uninstaller): Completely removes GPU drivers for clean reinstallation.
• Event Viewer: Built into Windows. Essential for checking restart logs and error patterns.

Download HWiNFO and OCCT first. These two tools alone will tell you more about your restart problem than anything else.

When to Replace vs Repair

Here is a decision framework to help you decide whether to replace a component or keep trying to repair it.

Replace your PSU if it is over four years old, is a budget tier unit, or if HWiNFO shows voltage rail sagging under load. PSU problems only get worse over time and a failing PSU can damage other components. This is not a component to gamble with.

Replace your thermal paste and clean dust if temperatures are high but your hardware is otherwise functional. This is a cheap fix that costs under $10 and an hour of time.

Replace your RAM if memtest86 shows errors after you have already tried reseating, cleaning contacts, and disabling XMP. One faulty stick is enough to cause system-wide instability.

Upgrade your motherboard only if you have confirmed VRM or power delivery limitations are causing restarts, and only if your CPU actually needs better power delivery. This is an expensive fix and should be a last resort.

Repair software issues through driver reinstalls, BIOS updates, and Windows repair tools (SFC and DISM) before spending any money on hardware.

FAQs

Conclusion: Finding Your PC Restart Cause

Understanding why your PC restarts randomly under load comes down to systematic elimination. Start with Event Viewer and HWiNFO monitoring to determine whether you are dealing with a hardware power-off or a software crash. Check your temperatures, voltages, and RAM stability. Test with the right tools, remembering that stress tests and games stress your system differently due to transient load patterns.

If your PSU is a budget unit, over four years old, or undersized for your system, it is the most likely culprit. No other component causes the specific pattern of no-BSOD restarts under variable load quite like a struggling power supply. Address the PSU first if your diagnostics point to power delivery, and always choose a quality replacement with adequate headroom for transient spikes.

The key insight from this guide that you will not find elsewhere is this: your PC can pass every synthetic stress test and still restart during games because transient load spikes behave completely differently from sustained loads. If FurMark passes but your games crash, stop second-guessing your GPU and look at your power supply.