Before continuing with the temperature results a brief word on the MadShrimps testbed which comprises of the following parts:

• ASUS Rampage IV Black Edition X79 motherboard
• Intel i7-3960X stock and OC'ed at 4500MHz 1.35Vcore
• Corsair Vengeance Pro 32GB 2666C11
• Western Digital 1TB Green Caviar HDD
• eVGA GTX 780 Classified
• Corsair HX1050W Power supply
• SilentiumPc Aquarius X90 case
• Ambient air temperature is 20°C

 

 

 

 

Prime95 is our favorite CPU torture test. By selecting the Custom test and setting 12-12K, we force the CPU to go flatout. After 60 minutes we verify in RealTemp the maximum load temperature results across the 6 cores. For the idle temperatures we just allow the system to warm up during a period of 15 minutes. Just basic idling at 4500MHz, doing nothing more then just monitoring the temperatures.

 

 

The monitoring software we use is RealTemp version 3. And the output results are the average out of three runs.

We compare the Fractal Kelvin S36 versus the different Corsair, Cooler Master and Enermax models. Secondly versus several air coolers: the single fan Alpenföhn Matterhorn Pure, Thermalright True Spirit 120M, Scythe Ashura and the Zalman XPS14. Retesting the older Hydro versions ( read H50/70 ) on the socket 2011 platform might yield inaccurate results, partly due to the aging nature and abuse these AIOs have undergone throughout the years. The full blown do it yourself EK waterblocks L 360 and the more basic Larkooler Skywater 330 kit are also included in the charts for comparison.

Noise measurement for the fans is done via Corsairs Air Fan Reviewers' Guide. Thus measuring the generated FAN noise at 1 meter from the FAN's front. Fans were hooked up straight to a PSU via a MOLEX adapter. Noise tests are only conducted at 12V.

 

 

 

We will focus in this article solely on the unit itself how it performs without addition of any supplemental waterblocks in the loops. Time to take a look at the results...