System and testing methodology
We tested the cooler on the following system, mounted on a bench table, with a temperature of about 25 °C:
We chose to use a Core i7 920 CPU Revision D0 because it can put stress on the heatsink properly, thanks to its 130W TDP at default.
We measured the temperatures at idle and full load with the RealTemp software. The temperature at full load were measured after 25 minutes of execution of Prime95 "InPlaceLargeFFTs" stress test (maximum stress and consumption), a benchmark which is well known for its ability to heavily stress the CPU, much more than any video game.
The temperatures are compared with those of the Noctua NH-C14, the new HR02 Macho and the new Kozuti Scythe, a low profile cooler.
With regard to the temperature measurements, they are performed with the fan speed set to the minimum and maximum speed, using the settings from BIOS. We used the Arctic Cooling MX-4 as thermal compound.
We performed three different test sessions, with the following settings:
- 2.66GHz 20 X 133, 1.20V, and QPI / DRAM set to 1.26V, the rest of the settings to AUTO (Hyper Threading ON, Intel Virtualization ON
- 3.4GHz 170 x 20, 1.25V, and QPI / DRAM set to 1.30V
- 4.0GHz 200 X 20, 1.30V, and QPI / DRAM set to 1.35V
