How to Fix an Overheating Laptop: The Practical Guide

A laptop that overheats isn’t just uncomfortable to use — it’s throttling itself to avoid damage, and every hour it runs hot is an hour it’s degrading components faster than it should. The good news is that the vast majority of overheating problems have one of three causes, all of them fixable. The bad news is that people regularly chase the wrong cause and waste time and money.

This guide cuts through the noise. Here’s how to confirm your laptop is actually overheating (not just warm), identify which of the three root causes you’re dealing with, and fix it the right way.

Step One: Confirm It’s Actually Overheating

A laptop that feels warm is not necessarily overheating. Modern CPUs run hot by design — they’re designed to extract every bit of performance within safe temperature limits. “Warm to the touch” is normal. Thermal throttling is not.

Use HWiNFO64 to Read Real Temperatures

Download HWiNFO64 — it’s free, portable, and the most complete sensor monitoring tool available for Windows. Run it in Sensors-only mode and look for the following values in the CPU section:

CPU Package Temperature: The overall CPU temperature. This is your primary number.
CPU Core temperatures: Individual core readings. On a multi-core CPU these vary; the hottest core is the one that matters.
Thermal Throttling: HWiNFO64 shows a boolean indicator — “Yes” or “No” — in the sensor list. If this is showing Yes under any real workload, your CPU is actively throttling.
Power Limit Throttling: Separate from thermal throttling. This means the CPU is hitting a power limit configured in firmware, not a temperature ceiling. It’s related but distinct.

Run a real workload while watching the sensor window. Gaming, video export, a compilation run, or even a 15-minute Prime95 stress test. Watch the CPU Package temperature and the throttling indicators.

Temperature Ranges: What’s Normal, What’s Not

Under 75°C under sustained load: Normal. The cooling system is doing its job. If performance feels fine and no throttling is active, you don’t have a problem.

75–88°C under sustained load: Warm but acceptable for most laptop processors. Many thin-and-light machines run in this range under full load by design. If no throttling is active, this is within spec — monitor it, but it doesn’t need immediate intervention.

88–95°C under sustained load: Getting into concerning territory. At this range, many Intel and AMD mobile processors will begin reducing clock speeds to stay below their thermal limits. AMD Ryzen mobile processors typically throttle at 95°C (Tj Max). Intel Core mobile processors from the 12th gen onward have a similar 100°C throttle point, but will often reduce power limits before that to manage heat.

Above 95°C under sustained load: This is overheating. The CPU is spending real time throttled, performance is being sacrificed, and component lifespan is being shortened. Something needs to be fixed.

Hitting 100°C immediately under any load: The cooling system is essentially failed — either the fan isn’t spinning, the heatsink is completely clogged, or the thermal interface has totally failed. This needs immediate attention before more use.

Intel vs AMD Throttling Thresholds

Intel (12th–14th Gen Core mobile): Tj Max is 100°C. The CPU will reduce power limits beginning around 90–95°C to stay under Tj Max. Some SKUs (particularly H-series high-performance chips in thin chassis) are configured with aggressive power limits that cause throttling even before thermal limits — that’s a firmware/design issue, not a cooling failure.

AMD Ryzen mobile (Ryzen 5000 and later): Tj Max is 95°C for most mobile SKUs. AMD tends to let the CPU run harder before throttling compared to Intel’s more conservative approach, which means AMD laptops can run hotter under load without actually throttling — but also means when they do throttle, they’ve been running at a higher sustained temperature.

Checking for throttling in ThrottleStop: ThrottleStop (discussed in the undervolting section) shows throttling reasons in more detail than HWiNFO64. If you see “BD PROCHOT” active, that’s a thermal signal from another component (often the power delivery circuitry or the GPU) forcing the CPU to throttle — which points to a different issue than simple CPU overheating.

The Three Root Causes

Once you’ve confirmed the laptop is actually overheating and throttling, the cause is almost always one of three things. Work through them in order.

Root Cause 1: Dust and Blocked Vents

This is the most common cause, especially in machines that are two or more years old. Laptop cooling systems pull air through intake vents (usually on the bottom or sides), push it through the heatsink fins, and exhaust it out the rear or sides. Dust and lint accumulate on the heatsink fins over time, eventually blocking airflow so severely that the fan is running at full speed and accomplishing almost nothing.

How to diagnose it: Shine a flashlight into the exhaust vents. If you can’t see daylight through the fins, they’re clogged. On some machines you can see the fan blades through the intake and judge how much debris has accumulated.

How to fix it — compressed air first: Before anything else, try compressed air. With the laptop powered off and unplugged, use a can of compressed air to blow through the vents — alternate between the intake and exhaust. Blow in short bursts. If a significant amount of dust comes out, this was the problem or at least part of it.

One important note: hold the fan still while blowing compressed air through the fins. Spinning a fan with compressed air at high velocity can over-speed the motor or push debris backward into the bearing. Use a toothpick, a finger through the vent, or a plastic pick to hold the blade before blasting air through.

When compressed air isn’t enough: If the heatsink fins are packed solid with compacted dust — which happens on machines that haven’t been cleaned in 3–5 years — you need to open the laptop to clean them properly. This means following the same disassembly procedure as a repaste (see the repaste guide for the full process). With the bottom panel off, use a soft brush or low-pressure compressed air to clear the fins from the heatsink side. Packed dust that won’t budge with air can be carefully cleared with a toothpick or plastic pick.

After cleaning, retest. If temperatures drop 10–20°C, that was the problem. If they’re still elevated after a proper cleaning, move to the next cause.

Root Cause 2: Dried or Failed Thermal Paste

Thermal paste is the interface between the CPU die and the heatsink. It’s a thermally conductive compound that fills the microscopic gaps between two metal surfaces that look flat but aren’t at a microscopic scale. When thermal paste degrades — which it does over 3–5 years under normal cycling conditions — those gaps fill with air instead of compound. Air is a terrible thermal conductor. The result is temperatures that run 15–25°C higher than they should even with a clean, unobstructed heatsink.

How to diagnose it: Age is the primary indicator. If the laptop is 3+ years old and has never had the thermal paste replaced, assume it needs it. Under a repaste, you’ll see the old paste — if it’s gray and crumbly or has separated into clumps with clear areas, it was definitely the problem.

When to repaste vs when cleaning alone is enough: If cleaning the vents brought temperatures down into acceptable range, a repaste is optional but still worthwhile on an older machine. If the machine is under two years old and vents are clean but it’s still running hot, the paste is probably fine — look at the third cause instead. If the machine is 3+ years old, vents are clean, and temperatures are still elevated, repaste it.

A proper repaste on a failed compound routinely drops temperatures 10–20°C. On machines where the paste had completely dried out, I’ve seen drops of 25°C. It’s one of the highest-impact, lowest-cost repairs you can do on an aging laptop.

The repaste process — tools, paste selection, application technique, heatsink screw sequence — is covered in detail in the repaste guide. Don’t skip reading that before doing your first one.

Root Cause 3: Inadequate Cooler for the Workload

Some laptops are simply undersized for the workload they’re being asked to run. Thin-and-light ultrabooks paired with 28W or 45W TDP processors often have cooling systems that are adequate for office work but throttle under sustained gaming or video encoding. This isn’t a failure — it’s a design tradeoff the manufacturer made to hit a certain thickness and weight target.

How to identify it: The laptop is clean, the paste is fresh (or the machine is new), and it’s still throttling under heavy loads. Check the manufacturer’s specs. If the machine has a 28W TDP CPU and a thin chassis, sustained heavy workloads will exceed what the cooling system can handle — that’s expected behavior, not a defect.

Software solutions: If you’re dealing with an undersized cooler, you can’t add heatsink mass — but you can reduce what the cooling system has to deal with. The main tool is undervolting.

Undervolting with ThrottleStop

Undervolting reduces the voltage supplied to the CPU at a given performance state. Lower voltage means less heat generated for the same clock speed, which means the cooling system has less work to do. It’s not overclocking — you’re not increasing performance above rated spec, you’re running at rated spec while generating less heat.

ThrottleStop is the standard tool for this on Intel platforms. It’s free, actively maintained, and gives you direct access to CPU voltage and power limit settings.

How to undervolt with ThrottleStop (Intel):

Download and run ThrottleStop. No installation needed — it’s portable.
Click FIVR (Fully Integrated Voltage Regulator) to open the voltage control panel.
Select CPU Core from the left panel.
Check Unlock Adjustable Voltage and set the Offset Voltage to a negative value. Start conservatively: -50 mV to -80 mV is a reasonable starting point.
Click CPU Cache and apply the same offset — the cache runs at the same voltage as the core, and undervolting one without the other is inconsistent.
Click Apply and then OK.
Run a stress test. If the system is stable after 20–30 minutes, try a larger offset in -10 mV increments. Stop when you find the threshold where instability begins, then back off by 20 mV for a stable operating margin.

Realistic gains from undervolting: A -100 mV to -150 mV offset on an Intel Core processor typically reduces temperatures by 8–15°C under sustained load. Combined with a heatsink clean and fresh paste, the cumulative effect can be significant on a machine that was previously thermal-throttling constantly.

AMD note: AMD Ryzen mobile processors don’t support traditional undervolting via ThrottleStop. AMD’s Precision Boost Overdrive (PBO) and Curve Optimizer features in Ryzen Master or the BIOS handle voltage management differently. Some Ryzen laptops allow negative voltage offsets in the BIOS; check your manufacturer’s BIOS for PBO settings if you have a Ryzen machine.

Intel 13th/14th Gen caveat: Intel tightened voltage limit controls starting with 12th Gen. Many 12th and 13th Gen systems allow undervolting in ThrottleStop, but some OEM configurations lock it out at the firmware level. If FIVR is greyed out or changes don’t persist, the manufacturer has locked voltage control. In that case, reducing power limits (the PL1/PL2 sliders in ThrottleStop’s main window) is an alternative approach that trades some peak performance for lower sustained temperatures.

Safe Temperature Ceiling Reference

To summarize the practical numbers:

Situation	Temperature Range	Action
Idle / light browsing	35–55°C	Normal
Under load, thin-and-light	75–88°C	Acceptable, monitor
Under load, any laptop	88–95°C	Investigate — cleaning or repaste likely needed
Under load, throttling active	95–100°C	Fix required — clean, repaste, or undervolt
Immediate 100°C at any load	100°C+	Cooling system failure — do not continue use

AMD Ryzen mobile Tj Max: 95°C. Intel Core mobile Tj Max: 100°C. Both throttle before hitting these ceilings — you’ll see performance reduction before you see absolute temperature limits.

The Correct Order of Operations

Don’t skip ahead. Work through this in sequence:

Confirm the problem with HWiNFO64 — actual temperatures and throttling status under real load.
Clean the vents with compressed air. Retest. If fixed, you’re done.
Disassemble and clean the heatsink fins if compressed air didn’t fully solve it. Retest.
Repaste if the machine is 3+ years old or temperatures are still elevated after cleaning.
Undervolt with ThrottleStop if hardware solutions don’t bring temperatures into the safe range, or if the chassis is simply thermally limited by design.

Each step is additive. A clean heatsink plus fresh paste plus a moderate undervolt on a machine that was previously throttling can easily add up to a 25–35°C reduction in peak temperatures — which means the difference between a machine that struggles through sustained workloads and one that runs at full rated performance indefinitely.

Field Notes — Get weekly tools, techniques, and deals from J.D. Sweeney. No fluff, unsubscribe anytime. Subscribe free →

One More Thing: Fan Health

If you’ve done everything above and temperatures are still wrong, the fan may be failing. A fan that’s failing doesn’t always make noise — sometimes it just spins slower than it should. HWiNFO64 shows fan RPM in the sensor list. Compare it to the manufacturer’s spec or to what the machine sounds like versus historical behavior. A healthy laptop fan under full load is loud. If you’re at 95°C and the fan sounds quiet, the fan is the problem.

A replacement fan for most laptops runs $15–$40 and is a straightforward swap. Once you’ve had the bottom panel off for a cleaning or repaste, you already know the disassembly sequence. Fan replacement is the same procedure with one additional connector.