PC restarts while gaming

• Estimated time: 2 min
• Easy

Separate restart, shutdown and game closure using what already happened.

1. Record whether the startup logo and Windows return, it stays off, a stop code appeared or only the game closed. If it stays off, follow the related gaming-shutdown guide; if it cannot stay powered on, follow the immediate power-on/off guide.

2. If only the game closes and Windows remains active, investigate that application. If it really restarts, record time, game and recent change. Save important data while stable; do not repeat the failure to get a photo.

• Estimated time: 5 min
• Easy
• Low risk

1. Once Windows is stable, open Event Viewer > Windows Logs > System and review around the recorded time. Keep BugCheck, WHEA-Logger and Kernel-Power 41 details without assuming the last visible event caused it. Event 41 is commonly recorded during startup after the failure.

2. If BugcheckCode is nonzero, Windows recorded a stop: keep the code and existing dump file for support. In Event 41 the number can be decimal; do not confuse it with the hexadecimal stop code. Do not publish complete dumps without privacy review.

3. If WHEA appears, keep component/error type and whether the same description recurs, then follow memory/stability. Without WHEA or bugcheck, hardware and drivers are not ruled out: a fast shutdown can prevent writing them. Follow changes, temperature or power according to the other findings.

• Estimated time: 5 min
• Easy
• Low risk
• Administrator required
• Restart required

1. Use the usage history: did other demanding games already run without restarting, or also fail? One game prioritizes its version, mods/overlays and driver; it does not prove healthy hardware because workloads differ. Do not install benchmarks to complete the comparison.

2. If it began after a game or graphics-driver update, compare version and symptom with official release notes. Apply only the relevant correction or a supported compatible previous version; do not reinstall Windows or start with cleanup tools. If a new overlay/mod was used, disable only that nonessential component before also changing the driver.

3. After a concrete correction, observe normal use only if the PC is stable without electrical/thermal danger signs. If it restarts again, stop the game and keep the result; do not run a test series until it fails again. If several workloads or idle also fail, prioritize general stability.

• Estimated time: 5 min
• Easy
• Low risk

1. Review existing temperature warnings and records from the PC/GPU’s official tool. Compare CPU, GPU and hotspot sensors only with their model limits: they are different measurements and this guide has no universal threshold. Intel publishes processor-specific limits; do not apply that value to every GPU or CPU.

2. A thermal warning, frequency drop near a temperature limit or blocked airflow prioritizes cooling. With the PC off and unplugged, externally check that vents are unobstructed and it has ventilation space. Do not insert objects or open the PSU; internal cleaning or fan work requires the manufacturer/technician procedure.

3. A normal reading after restarting does not describe the failure moment. If prior records show no excess, temperature becomes less likely but is not ruled out; follow other findings. For abnormal heat or a required fan not working, stop and seek inspection before another load.

• Estimated time: 5 min
• Easy
• Low risk

1. Record CPU, GPU, RAM and PSU model from documentation, plus any recent GPU/upgrade. Compare the PC/GPU manufacturer’s published power and connector requirements with the installed configuration. Printed wattage does not prove PSU health or compatibility of every combination.

2. An immediate restart under load with few logs makes power supply and assembly worth reviewing, especially after hardware expansion. But drivers, memory or motherboard can also contribute. FPS limiting or a lower-power GPU avoiding the failure does not separate these causes.

3. If shutdowns recur, ask a technician for controlled substitution with a known compatible PSU using that PSU’s correct cables, and connection/cooling inspection. Do not buy a PSU solely because of Event 41, open its enclosure, bridge pins or take electrical measurements. Without a new hypothesis, stop retries.