If you are experiencing noticeably higher temperatures, your PC is shutting down suddenly when under stress, or you are getting loud grinding or clicking noises from the cooler the cooler this could be a sign of a potential hardware failure.  These failures can result in reduced overall performance brought on by overheating.

What is considered overheating?

A PC is considered as overheating when the CPU cooler is unable to keep the PC within a safe operating temperature.  PC components do get noticeably hot when used, reaching upwards of 100°C and higher depending on the specs of the part when performance is needed.

When a system overheats, the CPU will begin a process known as thermal throttling where voltage and performance are reduced to bring the system back to a safe operating temperature.  This can be triggered for a few reasons, including but not limited to:

• The CPU cooler does not have sufficient cooling capacity (i.e. using a smaller size radiator such as a Kraken 120 or M22 on a high-performance part such as an AMD Ryzen™ 9 or Intel® Core™ Ultra 9 Processor.
• Air is not flowing properly through the radiator due to dust or debris.
• A component of the cooler has failed, such as the fan(s) or pump.

If thermal throttling does not resolve the issue in a timely manner, the system may trigger Over-temperature Protection (OTP) to protect the CPU and other components from potential damage.

What are safe operating temperatures?

Safe operating temperatures will vary based on the processor, however the best measurement for this is based on the Thermal Junction Maximum (TJMAXX) of the CPU.

Data above provided by AMD and Intel from official product pages as of 11 November 2025.

How do I check if my PC is overheating?

A good way to check if your PC is overheating is to check the idle temperature in the UEFI/BIOS, as this will have the CPU under minimal load.  For most motherboards, the UEFI/BIOS can be accessed by pressing the Delete key at the startup splash.

Example:  System idle temperature is 31°C indicating the system is properly cooled.

Normal idle temperatures will vary between configurations, but a common sign of an overheating CPU is a much higher than normal (70°C+) idle temperature on startup.  If you see a higher than normal startup temperature, there is likely a hardware issue with the cooler.

How do I check if my cooler has failed?

When checking for cooler issues, there are three major points where an issue can occur:

Checking for Pump Issue

The job of the pump is to circulate fluid from the CPU via the cold plate to the radiator where the fluid is cooled and circulated back to the cold plate to begin the cycle again.  Pump related issues are fairly straight forward, taking only a few minutes to troubleshoot:

• Is the Pump RPM being reported properly?
• Is there any loud noises coming from the Pump area?
• Is the cold plate making proper contact with the CPU?

Checking Pump RPM

Pump RPM is reported to the CPU via the 3-pin Pump Sense or 3-pin Pump Power cable depending on the model of your cooler.  Typically this header will be connected either to CPU_FAN (for coolers with integrated fan controls like the Kraken Elite, Kraken Plus, and Kraken (2023) or to AIO_PUMP/CPU_OPT (for coolers without integrated fan controls, like the Kraken Core and Kraken 120).  

To check this reading, ensure the applicable cable is connected to the appropriate header and open the UEFI/BIOS by pressing Delete at the startup splash screen.

Example:  The Kraken's 3-pin Pump Sense is connected to the CPU_FAN header on the motherboard, reporting 3200 RPM.

If the header shows 0 RPM or N/A, this could indicate a potential power issue to the pump, so be sure to check the following connections based on your Kraken:

If the Pump RPM is still reporting N/A or 0 RPM, this could indicate a failed or seized pump which we'll cover in the next section.

Noisy or Seized Pump

A less common issue is excessive noise from the pump or in some cases a pump seizing.  The steps for resolving these issues is the same and requires ensuring that the Pump Cap is properly installed and that there is no air in the cap itself.  Air in the cap can result in the pump colliding with the outer housing and producing a loud clicking or grinding noise.  To resolve these issues, do the following:

• Ensure that the pump cap is below the highest point in the radiator.
• If the pump cap is properly installed, remove the cooler from the system and shake the pump head below the radiator vigorously for 1-2 minutes.
  • For more information, see our guide here:  My Kraken is making a bubbling/rattling noise, what do I do?
• Reinstall the pump and set it to 100% speed in the CAM software (CAM-enabled Krakens only)

Example:  When using a Front Mount (left) or Top Mount (Right), the highest point in the radiator is above the Pump Cap to ensure air does not enter the cap.

If the noise or overheating persists, the cooler may need to be replaced.

Poor CPU Contact

Lastly a common issue at the pump cap can come from improper installation of the CPU cooler.  As all NZXT Kraken coolers come with pre-applied thermal paste, the best way to check the contact is to unscrew the Pump Cap from the CPU and check that the thermal paste has spread from the cold plate of the cap to the surface of the CPU.

If the thermal compound is intact, it's like the cooler was not properly installed.  Be sure to check that the right backplate and standoffs were used for your particular motherboard socket.  For more information, check the user's manual included with your cooler.

Checking for Fan Issues

Fan issues on an AIO are going to be extremely straightforward, as fans will either be working or they wont.  If your fans are not spinning then this could indicate that the fans are not receiving proper power or that the motors have seized.

If the fans are properly connected, power off the PC and disconnect the affected fan(s).  With the fan disconnected, try spinning the fan to confirm if it freely spins.  A properly working fan should spin freely and come to a gentle slow before stopping.  If the fans cannot freely spin and come to a sudden stop, especially if you feel resistance when trying to spin the fans this likely means the fan has seized and needs to be replaced.

Checking for Radiator Issues

The last type of issue will be with the radiator and cooling loop as a whole.  To start we recommend ensuring that the radiator fins are clean of all dust and debris.  This can be done by using compressed air and a brush to clear any dust on the radiator itself.

After confirming the radiator is clear of dust, power on the PC and feel the tubes to see if there is a noticeable difference in heat between the two.  While a minor difference is normal, a major difference in heat could be a strong sign of a blockage or other hardware issue.  Be sure to check the steps in the Pump section for clearing a noisy or seized pump, however if this does not resolve the issue a replacement will likely be needed.

Quick Reference