What thermal throttling is and why it happens

Thermal throttling is when your processor or graphics card deliberately slows itself down to reduce heat. Your CPU or GPU has a maximum safe temperature—usually between 90 and 105 degrees Celsius depending on the chip—and when it hits that limit, the hardware cuts its own speed to cool off. This is a safety feature, not a failure, but it means your computer is running slower than it should.

Thermal throttling happens because heat builds up faster than your cooling system can remove it. This might mean dust clogging your fans, a dried-out thermal paste layer between the chip and heatsink, a fan that has stopped spinning, or straightforward running demanding software in a hot room. The throttling itself is not the problem—it is the sign that something in your cooling chain is not working as designed.

Key Takeaways

  • Thermal throttling shows up as sudden, unexplained drops in performance during gaming or heavy work, even though your CPU or GPU is not fully loaded.
  • You can see throttling happening in real time using free monitoring tools like HWiNFO or GPU-Z, which display both temperature and clock speed side by side.
  • A CPU or GPU that holds a steady high temperature (above 85°C) while under load is a sign throttling is active or about to start.
  • The most common causes are dust buildup in fans and heatsinks, dried thermal paste, or a fan that has stopped working entirely.
  • Thermal throttling is reversible—cleaning your system or reapplying thermal paste usually restores full performance within an hour.

Performance drops that do not match your workload

The first sign of thermal throttling is usually a mismatch between what you are doing and how fast your PC responds. You might be playing a game that normally runs at 100 frames per second, but suddenly it drops to 40 or 50 fps, even though the game settings have not changed. Or you are rendering video or compiling code, and the job that usually takes 20 minutes now takes an hour.

The key is that the slowdown happens without any change in what you are asking the computer to do. If you lower graphics settings and performance improves, that is normal load balancing. If performance drops while the workload stays the same, thermal throttling is the most likely cause. This is especially true if the slowdown is sudden rather than gradual—throttling can kick in within seconds once the temperature threshold is crossed.

Temperature and clock speed readings in monitoring software

The clearest way to confirm thermal throttling is to watch your CPU or GPU temperature and clock speed at the same time. read HWiNFO (free, Windows) or GPU-Z (free, for graphics cards). Run your normal workload—game, video render, whatever causes the slowdown—and open the monitoring tool alongside it.

Look for this pattern: your CPU or GPU temperature climbs to 90°C or higher, and at the exact same moment, the clock speed drops. A healthy CPU running at full load might show 3.8 GHz; a throttled one will drop to 2.5 GHz or lower. The temperature will then stabilize at a lower level because the slower speed produces less heat. If you see temperature spike, clock speed fall, and performance drop all at the same time, throttling is active.

On Windows, you can also check the Event Viewer for thermal throttling logs. Press Windows key + R, type eventvwr.msc, and navigate to Windows Logs > System. Search for events with the source "Kernel-Power" or "Thermal-Zones"—these entries often flag when throttling has occurred, though not all systems log this reliably.

High sustained temperatures under load

Even without seeing the clock speed drop in real time, a CPU or GPU that stays above 85°C while under load is a strong indicator that throttling is either happening or about to start. Most modern processors begin throttling somewhere between 90 and 100°C, so sustained temperatures in the high 80s mean your cooling system is already struggling.

Compare your temperatures to what is normal for your specific hardware. A CPU like the Intel Core i9-13900K can safely reach 100°C under full load without damage, but if it is hitting that temperature during everyday tasks like web browsing or light gaming, throttling is likely active. Check the manufacturer's specifications for your exact chip—the thermal design power (TDP) and maximum junction temperature are listed on Intel's and AMD's websites.

Inconsistent performance in the same task

Thermal throttling often shows up as stuttering or inconsistency rather than a steady slowdown. You might notice frame rates that jump around—100 fps, then 60 fps, then 90 fps—even though nothing on screen has changed. This happens because as the chip throttles and cools slightly, it speeds back up, then heats again and throttles once more. The system is cycling between full speed and reduced speed every few seconds.

This pattern is different from a graphics driver issue or a background process stealing CPU time, which usually cause a steady, predictable slowdown. Thermal throttling creates a sawtooth pattern in performance: fast, then slow, then fast again. If you graph your frame rate or CPU usage over time using a tool like FrameView or the built-in performance monitor, you will see this cycling clearly.

What to check before assuming throttling

Before you open your case, rule out simpler causes. Check your Task Manager (Windows) or Activity Monitor (Mac) to see if another program is using CPU or GPU resources. A background antivirus scan, Windows Update, or a browser with 50 tabs open can cause slowdowns that look like throttling but are not.

Also verify that your power settings are not capping performance. On Windows, open Settings > System > Power & Battery > Power Plan and make sure you are not on "Power Saver" mode, which artificially limits CPU speed. Check your BIOS settings as well—some motherboards have power-saving features enabled by default that reduce clock speed to save energy.

Finally, confirm that your drivers are current. Outdated GPU drivers in particular can cause performance issues that mimic throttling. Update your graphics driver from NVIDIA, AMD, or Intel before you assume the problem is heat.

How to fix thermal throttling

Once you have confirmed throttling is happening, the fix usually involves one of three steps. First, clean the dust out of your case, fans, and heatsink using compressed air. Dust is the most common culprit—it insulates heat and prevents air from flowing. Power off your PC, unplug it, and use short bursts of compressed air to blow dust out of the CPU cooler, GPU cooler, and case fans. This alone fixes throttling in most cases.

Second, check whether your thermal paste has dried out. If your PC is more than three or four years old, the paste between your CPU and heatsink may have hardened and lost contact. You will need to remove the cooler, clean off the old paste with isopropyl alcohol, and explore a fresh layer. This is a straightforward 15-minute job if you are comfortable opening your case; if not, a local repair shop can do it for $30 to $60.

Third, verify that all your fans are spinning. Listen to your PC while it is under load—you should hear the fans speed up. If a fan is silent, it may have failed and needs replacement. A replacement case fan costs $15 to $30; a CPU cooler replacement runs $40 to $150 depending on the model.

Frequently Asked Questions

Can thermal throttling damage my CPU or GPU?

No. Throttling is a protection mechanism—it exists to prevent damage. Your hardware will not be harmed by throttling itself, but the underlying cause (overheating) can shorten the lifespan of your components if left unfixed for years. Fixing the cooling problem is the right move, but throttling is not an emergency.

Does thermal throttling happen on laptops?

Yes, and it is more common on laptops because they have less space for cooling and less airflow. Laptop throttling is often normal under heavy load, but if it happens during light tasks, the air vents may be blocked by dust or the internal fans may need cleaning. You can use the same monitoring tools to check temperatures on a laptop.

What is the difference between thermal throttling and power throttling?

Thermal throttling is triggered by heat. Power throttling is triggered by your power plan settings and is designed to save energy. Power throttling can look like thermal throttling, but it happens at lower temperatures and is reversible by changing your power plan to "High Performance" instead of "Balanced" or "Power Saver."

If I see high temperatures but no performance drop, is throttling still happening?

Possibly. Some systems throttle at lower temperatures than others, and some throttle gradually rather than all at once. If your temperatures are consistently above 85°C, your cooling system is working harder than it should be, and cleaning or reapplying thermal paste is still a good idea even if you have not noticed a performance hit yet.

Can I fix thermal throttling by undervolting?

Undervolting (reducing the voltage sent to your CPU or GPU) can lower temperatures and reduce throttling, but it is a workaround, not a fix. It works, but it requires software configuration and carries a small risk if done incorrectly. Cleaning your cooler or reapplying thermal paste is safer and usually solves the problem without any configuration.