Within our testing, we are aware that thermal pastes can present a difference but we’re also savvy to the fact that the difference could be less than a single degree. To ensure our tests are accurate and fair, we will take the following steps:

Each paste will be applied, tested and removed 3 times to try and average out both the amount of paste applied and the way the cooler is installed which will inevitably vary slightly between each run.

We will apply thermal pastes to manufacturers guidelines where available, otherwise reverting to the standard “rice grain” in the centre of the CPU.
The machine will be powered down for 10 minutes between each test to allow everything to reach ambient temperature.

We will aim to keep the room temperature at 20°C but will keep a very close eye on ambient temperatures with a thermometer that is accurate to 0.1°C. Our results will then be a delta to ambient (Average core temperature minus ambient room temperature). We will have an office fan circulating air in the room to prevent any kind of stagnation of the air.
Tests will be conducted straight away, leaving the paste no time to cure or change.

Our tests will last precisely 20 minutes, at which point the CPU temperature will be monitored with RealTemp for 60 seconds and then averaged out across all 8 cores.
We will also have a fan circulating air around the room to help stabilise and maintain an ambient temperature instead of allowing hot air to stagnate around our testing area.

Test System


Thermal Interface Materials Tested
Noctua NT-H1
Noctua NT-H2
Arctic MX-4
Corsair TM30
Thermal Grizzly Conductonaut
Thermal Grizzly Kryonaut