How to Fix a Computer That Overheats and Shuts Down

Published 2026-10-16 · LuxuriousComputers Repair

Your computer shuts down without warning during gaming, video editing, or other heavy tasks. When you turn it back on, it works fine for a while then shuts down again. Or it runs for decreasing amounts of time before each shutdown. This pattern — working under light use, failing under load — is the signature of thermal shutdown. The computer is overheating, and the hardware is shutting itself off to prevent permanent damage.

How thermal shutdown works

Every CPU and GPU has a maximum safe operating temperature built into its firmware. When the temperature exceeds this threshold (typically 100 to 105 degrees Celsius for CPUs, 105 to 110 for GPUs), the hardware forces an immediate power-off. There is no graceful shutdown — no saving files, no shutdown screen. The computer just dies as if someone pulled the power cord. This is a safety feature designed to prevent the silicon from exceeding temperatures that cause permanent physical damage.

Before reaching the shutdown threshold, the hardware goes through throttling stages — reducing clock speed to generate less heat. If you notice the computer getting slower before it shuts down (frame rates dropping in games, rendering getting slower, general sluggishness), that is thermal throttling happening before the final shutdown.

Check temperatures before anything else

Install HWiNFO (free) and monitor temperatures during the tasks that trigger shutdowns. The key readings:

• CPU Package or CPU Tctl/Tdie: This is the overall CPU temperature. Normal under heavy load is 70 to 85 degrees Celsius depending on the cooler. Above 90 is concerning. Above 95 is in the thermal throttle range. At 100 to 105, the system shuts down
• GPU Temperature: Normal gaming temperature is 65 to 85 degrees depending on the card and cooler. Above 90 indicates a cooling problem. Above 100, most GPUs force a shutdown
• GPU Hotspot: Some GPUs report a hotspot temperature that is higher than the general GPU temperature. This is the hottest point on the die and is what triggers thermal protection
• CPU or GPU Thermal Throttle flag: HWiNFO shows a yes/no flag when throttling is active. If this flag activates before shutdowns, you have confirmed the cause

Dust — the most common cause

Dust buildup is responsible for the majority of overheating problems, especially on computers that are 1 to 3 years old and have never been cleaned internally. Dust accumulates on heatsink fins, fan blades, intake filters, and exhaust vents, forming an insulating layer that traps heat.

Cleaning a desktop

1. Power off the computer and unplug it. Open the side panel
2. Use compressed air (canned air or an electric air duster) to blow dust off the CPU cooler heatsink. Hold the fan blade still with a finger while blowing — letting the fan spin freely from compressed air can damage the bearing or generate voltage on the motherboard header
3. Clean the GPU heatsink the same way. If the GPU is heavily clogged, remove it from the PCIe slot to access both sides
4. Clean all case fans — intake and exhaust. Remove intake dust filters if your case has them and wash them with water (dry completely before reinstalling)
5. Clean the power supply exhaust vent and fan (blow air through the PSU exhaust grille — do not open the PSU)
6. Check that airflow paths are clear. Front-to-back or bottom-to-top airflow is standard. Cables should not block the path between intake and exhaust fans

Cleaning a laptop

Laptop cooling systems are tighter and more prone to dust buildup:

• Use compressed air to blow through the exhaust vents (usually on the side or back). This pushes dust back toward the intake. Alternate between exhaust and intake to work dust out
• If you are comfortable disassembling the laptop, remove the bottom panel to access the fans and heatsink directly. Laptop heatsink fins develop a solid wall of dust that compressed air alone often cannot dislodge
• Laptop fans are smaller and spin faster than desktop fans, so they accumulate dust faster and are more significantly affected by it

Thermal paste replacement

Between the CPU die (the metal surface of the processor) and the cooler's heatsink sits a thin layer of thermal paste. This paste fills the microscopic gaps between the two surfaces to improve heat transfer. Over 3 to 5 years, thermal paste dries out, hardens, and loses its ability to transfer heat efficiently. Dried thermal paste can cause a 10 to 25 degree Celsius increase in temperatures compared to fresh paste.

Desktop CPU thermal paste replacement

1. Remove the CPU cooler (unscrew or unclip it from the mounting bracket)
2. Clean the old paste from both the CPU surface and the cooler base using isopropyl alcohol (90 percent or higher) and a lint-free cloth or coffee filter
3. Apply a small amount of new thermal paste to the center of the CPU (a grain-of-rice-sized dot for most CPUs, a small X pattern for larger CPUs like Threadripper)
4. Reinstall the cooler, tightening mounting screws in a cross pattern (opposite corners) to ensure even pressure

GPU thermal paste replacement

GPU thermal paste also dries out but requires disassembling the GPU cooler from the graphics card PCB. This involves removing the backplate screws, carefully separating the heatsink from the board, and cleaning and reapplying paste on the GPU die. GPU paste replacement also often means replacing the thermal pads on the VRAM chips and VRM components. This is a more advanced procedure than CPU paste replacement but follows the same principles.

Laptop thermal paste replacement

Laptop CPUs and GPUs share a heatsink connected by heat pipes. Accessing them requires significant disassembly — removing the bottom panel, disconnecting the battery, removing fans, and separating the heat pipe assembly. The paste application is the same, but the reassembly must be precise because laptop cooling systems have very tight clearances.

Fan failures

A dead or dying fan cannot move enough air through the heatsink to keep the component cool. Check all fans while the computer is running:

• CPU fan: Should be spinning whenever the computer is on. If it is not spinning or spinning very slowly, the fan has failed or the fan header connection is loose
• GPU fans: Many modern GPUs have a zero-RPM mode where the fans do not spin at idle (this is normal and by design). The fans should spin up under load. If they remain off during gaming or stress testing, they have failed or the fan control circuit on the GPU board has a problem
• Case fans: All installed case fans should be spinning. A dead exhaust fan traps hot air in the case, raising temperatures for everything inside

Desktop fans are usually replaceable — standard 120mm or 140mm fans for case positions, and CPU cooler fans can often be replaced without replacing the entire cooler. Laptop fans are model-specific replacement parts.

Insufficient cooling for the hardware

Some computers overheat because the cooling solution was never adequate for the hardware:

• Stock CPU cooler with a high-power CPU: The basic cooler that comes with some processors is designed for stock speeds under moderate loads. Under sustained heavy loads (long gaming sessions, video rendering, compile jobs), the stock cooler may not dissipate heat fast enough. An aftermarket tower cooler or liquid cooler solves this
• Poor case airflow: Cases with solid front panels and few fans restrict airflow. Adding intake fans, switching to a case with a mesh front panel, or even leaving the side panel off temporarily can determine if airflow is the bottleneck
• Laptop on soft surfaces: Using a laptop on a bed, couch, or blanket blocks the intake vents on the bottom. Always use a laptop on a hard, flat surface. A laptop cooling pad with fans provides additional airflow for laptops that run hot even on a desk

Other causes of overheating shutdowns

• VRM overheating: The voltage regulator modules on the motherboard that power the CPU can overheat on budget motherboards paired with high-power processors. VRM overheating does not always trigger a CPU temperature warning — the motherboard shuts down to protect its own components. Check VRM temperatures in HWiNFO if available
• Ambient temperature: A computer in a hot room starts with a higher baseline temperature. If the room is above 30 degrees Celsius and the computer barely stays cool enough normally, a hot day can push it over the edge
• Overclocking: Overclocked CPUs and GPUs generate more heat than stock. If thermal shutdowns started after applying an overclock, the cooling system cannot handle the increased thermal load. Reduce the overclock or improve cooling

Overheating repair in Marion

If your computer keeps shutting down from overheating and cleaning or paste replacement did not fix it, or if you are not comfortable disassembling the computer yourself, LuxuriousComputers in Marion can diagnose the thermal issue, clean the cooling system, replace thermal paste, test fan operation, and ensure the computer runs within safe temperatures under load.