Skip to main content
Excessive heat is one of the most damaging conditions a PC can operate under. Left unaddressed, chronic overheating shortens the lifespan of your CPU, GPU, and storage devices, and can trigger data-corrupting crashes at the worst possible moment. This guide walks you through identifying overheating, measuring temperatures accurately, and resolving the underlying cause.
1

Recognise the Symptoms of Overheating

Before opening any monitoring tool, confirm that overheating is the likely culprit. The following symptoms are strong indicators:
  • Unexpected shutdowns or restarts — modern CPUs and GPUs perform an emergency shutdown when they exceed their thermal threshold (typically 95–105 °C).
  • Performance drops during load — thermal throttling reduces clock speeds to dissipate heat, causing stuttering in games or slow rendering in creative apps.
  • Loud or continuously running fans — fans at 100% RPM under light workloads suggest the cooling system is already overwhelmed.
  • System instability only under load — errors or crashes that appear during gaming, video encoding, or compilation but not during idle use.
Overheating and insufficient power delivery share many symptoms. If temperatures appear normal but instability persists, investigate the PSU and power rails as a parallel track.
2

Monitor Temperatures With Diagnostic Tools

Obtain a temperature baseline before making any physical changes. This lets you confirm the problem and later verify that your fix worked.Windows — HWiNFO64 or HWMonitorDownload and run HWiNFO64 (recommended) or HWMonitor. In the sensors view, watch these values:macOS — iStatMenus or StatsInstall iStatMenus (paid) or the free Stats app. Monitor CPU die temperature, GPU temperature, and fan RPM from the menu bar.Linux — sensors (lm-sensors)
Run a deliberate stress test — such as Prime95 for the CPU or FurMark for the GPU — while watching temperatures. A well-cooled system should plateau below danger-zone thresholds within a few minutes rather than climbing continuously.
3

Clean Dust From Vents, Fans, and Heatsinks

Dust acts as an insulating blanket over heat sinks and fan blades. A single year of use in an average office can reduce cooling capacity significantly.
  1. Power off the PC completely and unplug it from the wall.
  2. Take the PC to a well-ventilated area (preferably outdoors) to avoid redistributing dust inside the room.
  3. Use a can of compressed air in short bursts to blow dust out of:
    • CPU heatsink fins (blow through the fins, not across them)
    • GPU heatsink and fans
    • Case intake and exhaust fan grilles
    • Power supply intake (blow from the outside in — never open the PSU)
  4. Use a soft brush or cotton swabs for stubborn dust buildup on fan blades.
  5. For laptops, focus on the exhaust vents on the underside or rear edge with short compressed-air bursts.
Hold spinning fans stationary with a finger or a pen while applying compressed air. Allowing fans to over-spin from air pressure can damage the fan bearings.
After cleaning, power on and recheck temperatures. In dusty environments, cooling improvements of 10–20 °C are common after a thorough clean.
4

Check Case Airflow and Fan Orientation

Clean fans cannot compensate for a poorly configured airflow path. Air must move from intake to exhaust in a predictable direction, flowing over the major heat sources along the way.
  • Standard airflow convention: intake fans at the front and bottom, exhaust fans at the rear and top.
  • Confirm each fan is blowing in the correct direction — the label/sticker side of the fan frame faces the direction air is pushed toward (the exhaust side), while the open blade side faces the direction air is drawn from (the intake side). For an intake fan mounted at the front of the case, the label faces inward; for an exhaust fan at the rear or top, the label faces outward.
  • Ensure cables inside the case are bundled and routed away from the airflow path (cable management matters for thermals, not just aesthetics).
  • Verify that slot covers are installed in unused PCIe and expansion slots — open gaps short-circuit airflow.
Laptops have fixed, sealed airflow paths. If a laptop overheats after cleaning, focus on thermal paste (Step 5) and fan curve settings (Step 6) rather than airflow configuration.
5

Reapply Thermal Paste on the CPU (and GPU if Needed)

Thermal paste between the CPU die and heatsink conducts heat away from the processor. After 2–5 years, most pastes dry out, crack, or pump out from the interface, sharply reducing thermal transfer.You will need: Isopropyl alcohol (90%+), lint-free wipes, new thermal paste (Arctic MX-6, Thermal Grizzly Kryonaut, or equivalent).
  1. Remove the CPU cooler by unscrewing its mounting hardware evenly in a cross pattern to avoid warping the bracket.
  2. Clean the old paste from both the CPU heat spreader and the cooler base plate using isopropyl alcohol and a lint-free cloth. Repeat until no residue remains.
  3. Apply a pea-sized (3–4 mm) dot of new paste to the centre of the CPU heat spreader. Do not spread it manually — the mounting pressure will distribute it.
  4. Re-mount the cooler, again tightening screws in a cross pattern to apply even pressure.
  5. For a GPU, the process is the same but involves removing the GPU shroud first — consult the manufacturer’s teardown guide for your specific card.
Never power on a CPU or GPU without thermal paste installed. Even a few seconds of operation without paste can cause permanent damage to the chip.
Expect CPU temperatures to drop by 5–20 °C after a fresh paste application on a system more than two years old.
6

Adjust Fan Curves in BIOS or Software

If temperatures are acceptable but fans remain loud, or if fans are not spinning up fast enough under load, tune the fan curve.Via BIOS/UEFI:
  1. Restart and press the BIOS key during POST (typically Delete, F2, or F10 — shown briefly on the boot screen).
  2. Navigate to the Hardware Monitor, Fan Control, or Q-Fan section (label varies by board manufacturer).
  3. Select the relevant fan header (CPU_FAN, SYS_FAN1, etc.) and choose a PWM or temperature-based control mode.
  4. Set a curve that keeps fans quiet below 50 °C and ramps aggressively above 70 °C.
Via Software (Windows) — Fan Control app or MSI Afterburner (GPU):
For GPU fan curves on NVIDIA or AMD cards, use MSI Afterburner → Fan tab → Enable user-defined software control → drag curve points.
Save your fan curve changes to a BIOS profile before exiting. If a future BIOS update resets your settings, you can reload the saved profile rather than re-configuring from scratch.
7

Escalate to Tier 2 Support

If temperatures remain in the danger zone after completing all of the above steps, further intervention is required.
  • CPU or GPU temps exceed 100 °C even at idle after cleaning and paste replacement
  • System shuts down within minutes of booting regardless of load
  • Fan(s) are completely non-functional (zero RPM in sensors, no audible spin-up)
  • Visible burn marks, discolouration, or a burning smell from any component
  • Laptop heatsink pipe is cracked, kinked, or the vapour chamber has lost its vacuum (flat, room-temperature feel when under full load)
Provide the Tier 2 technician with your temperature logs from HWiNFO64 or iStatMenus, the steps you completed, and when the overheating issue first appeared. Component-level repairs — such as replacing a heatsink, a fan, or the liquid cooling loop — require physical access and are handled at the bench.