A PC that restarts while gaming usually has a hardware-protection, power-delivery, driver, or stability problem—not necessarily a failing game. The important clue is whether the system instantly loses power, shows a blue-screen error, freezes first, or restarts only in one title. Modern games can stress the CPU, GPU, memory, storage, and power supply at the same time, exposing a weakness that is invisible during web browsing or light work.

Start by preserving data and avoiding repeated forced tests if you smell burning, hear electrical crackling, or see swelling capacitors. For ordinary gaming restarts, work from the least invasive checks upward: record the error, remove overclocks, monitor temperatures, test power behavior, then isolate hardware and software variables.

Most Common Causes

Cause Symptom Likelihood
Power supply instability or inadequate capacity Instant black screen and reboot during GPU-heavy scenes; no blue screen High
CPU or GPU overheating Restart after several minutes of play; fans become loud; case feels hot High
GPU driver fault or corrupted driver installation Game crash, black screen, display-driver message, or reboot after an update High
Unstable RAM, XMP/EXPO, CPU tuning, or GPU overclock Random restarts, blue screens, game-specific crashes, errors under load High
Windows corruption, firmware issue, or conflicting background software Restarts in multiple games after a system change; Event Viewer errors Medium
Loose GPU, RAM, or power connection Problem began after moving, cleaning, or upgrading the PC Medium
Failing GPU, motherboard, RAM, or storage drive Increasing crashes, artifacts, boot failures, or errors outside games Low to medium

Diagnose Step by Step

  1. Check whether Windows recorded a crash

    Press Win + X, select Event Viewer, then open Windows Logs > System. Choose Filter Current Log and check Critical and Error. A Kernel-Power, Event ID 41 entry confirms Windows detected an unexpected shutdown, but it does not identify the failed component. Look immediately before it for display-driver, disk, WHEA-Logger, or BugCheck events.

    Also open Control Panel > Security and Maintenance > Reliability Monitor, or run perfmon /rel from the Start menu. Note the game, time, and any named failing application or driver.

  2. Stop automatic restart so blue-screen information remains visible

    Press Win + R, type sysdm.cpl, and select Advanced > Startup and Recovery > Settings. Clear Automatically restart. Under “Write debugging information,” select Small memory dump (256 KB) and confirm the path is %SystemRoot%\Minidump. If the next failure displays a stop code instead of immediately rebooting, write down the code and any driver filename.

  3. Return every performance setting to stock

    Disable CPU overclocking, undervolting, Precision Boost Overdrive, Intel performance tuning, GPU overclocks, and memory profiles temporarily. In UEFI/BIOS, use the option commonly named Load Optimized Defaults, Load UEFI Defaults, or Restore Defaults. Then save and restart. In GPU software, reset the tuning page to default values.

    This is a diagnostic step, not an admission that the hardware cannot run its rated settings. XMP and EXPO profiles are overclocked memory configurations in practice, and a marginal CPU memory controller can fail only under demanding games.

  4. Monitor CPU and GPU temperatures during the failure window

    Use a hardware-monitoring utility that can log sensor values to a file. Watch CPU package temperature, GPU temperature, GPU hot-spot temperature if available, clock speeds, fan speeds, and power readings. Start logging before launching the game, then review the final values after a restart.

    Do not rely on one universal “safe” temperature because limits differ by processor and graphics card. What matters is whether temperatures rapidly approach the device’s stated thermal limit, clocks collapse, or the restart occurs after heat builds over time rather than during the first minute of play.

  5. Separate a power event from a software crash

    Set the game to a lower GPU load: cap frame rate to 60 FPS, disable ray tracing, reduce resolution, and select a lower preset. If the PC becomes stable only when GPU power demand falls, suspect the power supply, GPU power cables, GPU tuning, or heat. Test more than one game. A restart in every demanding title points to a system-level cause; a restart in one game points first to that game’s files, anti-cheat component, mod, or graphics API setting.

  6. Run basic Windows, memory, and drive checks

    Open Windows Terminal (Admin) from the Win + X menu. Run sfc /scannow, then run DISM /Online /Cleanup-Image /RestoreHealth. For memory, press Win + R, enter mdsched.exe, and choose Restart now and check for problems. Check drive health in the SSD manufacturer’s utility and ensure the Windows drive has free space for updates, shader caches, and crash dumps.

Fixes by Cause

Power supply, power cable, or outlet instability

What causes it: A weak, aging, defective, or overloaded PSU can trip protection when CPU and GPU demand spike together. Loose modular cables, daisy-chained GPU power connectors, damaged extension leads, and unstable power strips can create similar symptoms.

How to check: Look for instantaneous restarts with Kernel-Power 41 and no blue-screen code. Inspect the 24-pin motherboard cable, CPU EPS cable near the top of the motherboard, and every GPU power connector with the PC unplugged. Confirm connectors are fully seated and latched.

What to do: Connect the PC directly to a known-good wall outlet for testing. Remove questionable adapters and extensions. If the GPU accepts multiple PCIe power connectors, use separate PSU PCIe cables when the PSU documentation supports that configuration; do not mix modular cables from another PSU, even if they fit. Temporarily set the GPU power limit to 80% to 90% in its tuning utility. If that stops restarts, test with a known-good, properly sized PSU.

How to undo: Restore the GPU power limit to 100% and reconnect your normal surge protector only after confirming it is not involved. Keep any correctly seated cable arrangement in place.

CPU, GPU, or case overheating

What causes it: Dust-clogged heatsinks, failed fans, dried thermal compound, poor cooler mounting, blocked intake airflow, or aggressive overclock settings can trigger thermal protection.

How to check: Compare logged temperatures with the CPU and GPU manufacturer’s thermal limits. Check whether all CPU cooler, GPU, and case fans spin. Remove the side panel briefly as a test; a large improvement suggests airflow restriction.

What to do: Shut down, switch off the PSU, unplug it, and clean dust with compressed air while preventing fan blades from spinning freely. Confirm the CPU cooler fan is connected to the CPU_FAN header. In UEFI/BIOS, use the fan-control page to choose a standard or performance curve. Improve cable routing and verify front intake and rear/top exhaust fans are oriented correctly.

How to undo: Restore the previous fan curve if noise is unacceptable after stability is confirmed. Reinstall the side panel after testing; it is not a permanent cooling solution.

GPU driver or game software corruption

What causes it: A damaged driver install, incompatible overlay, shader-cache issue, or game update can cause display failures that occasionally escalate into a restart.

How to check: In Reliability Monitor, look for hardware errors or display-driver failures. Test the game with Discord overlay, Steam overlay, recording tools, RGB suites, and GPU overlays disabled. Check whether another demanding game is stable.

What to do: Install a current stable GPU driver using its Custom or Clean installation option if offered. Verify the game’s installed files in its launcher. Reset in-game graphics settings, remove mods, and clear only the game’s shader cache if the game provides that option. In Windows, open Settings > System > Display > Graphics and set the affected game to High performance if the PC has both integrated and discrete graphics.

How to undo: Re-enable overlays one at a time. If the newest driver created the problem, use Device Manager > Display adapters > [GPU] > Properties > Driver > Roll Back Driver when available.

RAM, XMP/EXPO, or overclock instability

What causes it: Memory profiles and manual CPU or GPU tuning can appear stable in light tasks while failing under game loads.

How to check: Run Windows Memory Diagnostic and inspect Event Viewer afterward for MemoryDiagnostics-Results. More importantly, test with XMP/EXPO disabled and all tuning reset. WHEA-Logger errors are especially relevant for CPU, memory, and PCIe instability.

What to do: Leave memory at default settings for several gaming sessions. If stable, re-enable XMP/EXPO but choose a lower memory-speed option if the firmware provides one. Reseat RAM: power off, remove the modules, and reinstall them in the motherboard’s recommended slots, commonly A2 and B2. Test one module at a time if errors persist.

How to undo: Re-enable the original profile only after confirming stability. Keep notes of the prior frequency, voltage, and timing settings before changing them.

When to Replace or Reinstall

Replace the PSU when restarts continue at stock settings, GPU power limiting improves stability, cables and outlet tests are clean, and a known-good PSU resolves the issue. Replace or repair cooling hardware when fans fail, the cooler mount is damaged, or temperatures remain excessive after cleaning and restoring stock settings.

Consider RAM replacement when one module repeatedly fails diagnostics or the PC is stable only with that module removed. Consider a GPU or motherboard warranty claim if there are visual artifacts, repeated PCIe/WHEA errors, crashes in multiple clean operating systems, or faults reproduced with known-good power and memory.

Reinstall Windows only after hardware, temperature, driver, and memory checks. First try Settings > System > Recovery > Fix problems using Windows Update if available, or Reset this PC with “Keep my files.” A clean installation is reasonable when Reliability Monitor shows widespread application failures, system files cannot be repaired, or a restart began after extensive driver and utility changes. Back up documents, game saves, encryption recovery keys, and browser data before any reset. This order remains practical for PC troubleshooting in 2026 because reinstalling Windows cannot repair a failing PSU or unstable memory profile.

FAQ

Why does my PC restart instead of showing a blue screen?

An instant restart often means power was interrupted or hardware protection activated before Windows could save an error. It can also be a blue screen hidden by automatic restart. Disable automatic restart through sysdm.cpl and check Event Viewer for Kernel-Power 41 and earlier errors.

Can a GPU cause the entire PC to restart?

Yes. A GPU can trigger a restart through excessive heat, a driver failure, a defective card, or a sudden power-demand spike that exposes a PSU or cable problem. Lower the GPU power limit and frame rate as a reversible test; do not assume the GPU is defective until power and driver checks are complete.

Is Kernel-Power 41 the cause of my gaming restart?

No. Kernel-Power Event ID 41 means Windows noticed that the previous shutdown was unexpected. It is evidence of the restart, not a diagnosis. Read the events immediately before it and compare them with temperature logs and the exact point at which the game failed.

Should I update the BIOS if my PC restarts while gaming?

Update firmware only after restoring default settings and checking the motherboard support page for fixes relevant to CPU compatibility, memory stability, or system crashes. Use the board’s built-in UEFI flash tool, avoid power interruptions, and do not update BIOS merely because a single game crashes. Record existing settings first because updates commonly reset them.

Related guides

Browse all Troubleshooting guides →