Performance

For the test setup, Prime95 was used to load the CPU, and Windows at the Desktop was used to get the idle temperature. Different clocks/voltage were used to increase the total heat output. The amount of heat output of a processor is measured in watts. All CPUs/GPUs have what's known as a TDP or Thermal Design Point which is given by the manufacturer (different manufacturers measure TDP in different ways). This TDP rating gives a bases of how much heat is being dissipated off of a CPU when it's at its default state (not overclocked). However to get the heat output of an overclocked CPU, a formula is needed. To calculate this, the formula " TDP * (OC MHz / Stock MHz) * (OC vCore / Stock vCore )² " was used.



The cooler was mounted, then removed immediately after and the contact area was inspected. The paste that comes pre-applied had fairly decent contact, but pre-applied stuff is usually put on thick. I applied my own thermal paste later then checked and contact was superb.


17 Celsius idle! Very nice indeed, and 32 Celsius on full load is very nice as well. I decided to run the cooler on silent mode as well, which is not completely silent but is very quiet in its operation. It didn't do too well if you consider the fact that it has a 300USD price tag - It idled pretty much the same as most coolers and did worse on full load. When the cooler was ran in silent mode, the TECs had less power on them, and they are the only means of cooling the water, so the water had almost nothing cooling it.


The rest of the tests were done on full power as we saw how warm the CPU got when the cooler was run on silent at stock processor settings. Simply put, the Freezone's idle temperatures are amazing. On full load the cooler did beat every other cooler tested, but not by leaps and bounds as it did at idle. One thing you do need to remember is that this cooler is rated for 170 Watts max; so we're nearing the limit of this cooler.

As stated before, the Freezone's cooling method is done via thermo-electric coolers and there are no radiators used. The hot water returning from the CPU block goes through a small reservoir and two blocks with six TECs on the blocks. If you were to take a look at a radiator used in an "conventional" water system, they can be quite large. Another key feature to water systems is the reservoir. It serves two main purposes; One is to bleed air in the system, and the other is to hold more water. Holding more water is key, because more water takes a lot more energy to heat up due to its specific heat capacity.

An analogy - Take two pots of water; Both are the same pots, both have the same amount of heat being applied, but one pot is 1/4 full, and the other is 1/2. Which pot will start to boil first? The pot with 1/4 water will because it has the same amount of energy being applied, but there's less water to be heated.

Where the Freezone's reservoir is more for bleeding air and lacks any size to hold a significant amount of water, and the two blocks over the cold TECs are relatively small, once the water starts to get hot, the system has trouble keeping up. So in short, water returns to be cooled quicker than the cooler can cool the water and the water keeps getting warmer. When there is a small load, not much heat is being radiated to the block, and the TECs have enough power to efficiently cool the water. A reservoir will give the TECs more time to cool, and because it will add more water to the system loop, it will take more to heatup, again giving the TECs more time to do their job.

More powerful TECs are not the best option either because you have to consider that when you increase their power, their heat output as well as their cooling ability goes up. When the system is on full power, the heat coming out of the unit is quite high. I think there could be a little head room for slightly more powerful TECs, but not much. It's quite a balanced system.