CPU Cooler Testing Methodology

Testing was conducted at a series of temperature intervals, with the ambient room temperature measuring the same for each cooler's reading. Lavalys EVEREST rounds up temperature readings to the nearest whole number, however all ambient temperatures were recorded and accurate to one-tenth of a degree Celsius.

EVEREST Ultimate Engineer Version 4.20.1170 was utilized to measure all CPU temperatures. It should be noted that temperatures obtained through software most always reflect the thermistor readings as recorded by the BIOS. All of the units compared in the results were tested on the same motherboard using the same BIOS. These readings are not absolute or correct, but they are relative since every BIOS is programmed differently.

Test System

For each test, the computer system was powered on and left idle for ten minutes prior to recording the idle temperature with EVEREST. After idle temperatures have been recorded, two console versions of the Folding @ Home client were simultaneously run to create full load on both CPU cores. After ten minutes of full load, the temperature was again recorded. This process was identical for all cooling solutions used in this test, and was repeated for each ambient temperature interval.

Scythe Mugen Results

Up first are the results under idle load. The CPU temperature is listed vertically along the left-hand side of the chart, and the ambient room temperature is displayed for each group along the bottom. In this test battery, the OEM cooler provided by Intel with the Core 2 Duo E6600 was used as a reference, which offered a high-temperature baseline reading.

When I tested each cooler, I made certain to keep the hardware settings identical across the entire test platform. I used the same front side bus and clock speeds, which would enable me to clearly compare the performance of each product under identical conditions. As the ambient room temperature would increase, all of the coolers I have ever tested would also record higher readings (all except for the Ultra ULT33186 Chilltec TEC CPU Cooler). Since the OEM cooler supplied by Intel was the first unit to be tested, there was quite a difference between the ambient room temperature and the temperature of the cooler (41° C idle at 23° C ambient room temperature).

At idle, the Scythe Mugen 5-Heatpipe CPU Cooler SCINF-1000 was recording results around 9° C over the ambient room temperature, which is good. The average temperature of 33° C recorded at 23° C ambient was exceptional, but nowhere near the Ultra Chilltec, and a few degrees warmer than the Zaward VIVO, OCZ Vendetta, and Xigmatek HDT-S1283 which all use the HDT technology. But the heat had just started to be applied, and once I completed tests on the other coolers it would be time for real cooling under serious load.

Under full load, the Intel Core 2 Duo E6600 began to create tremendous heat. The little Intel OEM cooler, despite its inserted copper base, gave the worst results anyone could suffer recording, registering 59° C at full load while the room's ambient temperature remained 23° C. But that's what you come to expect from a bundled cooler, right? Sadly this must be the case, because I still hear of many users (even some gamers and hardware enthusiasts) which still use the stock reference cooler. I feel for those of you in this situation, I really do...

...but there are some very good reasons to evolve, or at least save up your money and upgrade. Previously, the very expensive Ultra ULT33186 Chilltec TEC CPU Cooler held the crown for cooling, if only by a very small margin. In all honesty, the Zaward Sylphee ZCJ003 CPU Cooler was the real king of the hill prior to these tests, since it is roughly half the size of everything here. current champion "Champion of Cool" is the Xigmatek HDT-S1283 Exposed Copper Heatpipe CPU Cooler, which defeated all challengers with a maximum recorded temperature of only 39° C in a 23° C ambient room temperature and at full load. But what about the Scythe Mugen?

To my disappointment all of that extra cooling surface and those five copper heatpipes were not enough to secure a top-three position in our rankings. While the Scythe Mugen was not a top finalist, it certainly was good enough to rank among the top-ten. I think this is all the proof I needed to be convinced that Heatpipe Direct Touch designs are going to be the next standard for CPU coolers. If coolers like the Zaward VIVO PCJ004 and OCZ Vendetta can outperform the Mugen with half the footprint or more, it seems evolution has made up its mind.

<< Prev - Next >>