Popular watercooling options compared — Joe
SUMMARY: Watercooling Cases/Kits deliver great performance at low noise.
We tested a number of watercooling cases and kits using the CPU Die Simulator. Comparing kits and cases are system specific – by that I mean you can not infer from these tests that the Swiftech waterblock is better than the Koolance waterblock.
You CAN conclude that the components in kit/case A perform better than the kit/case B based on test C/Ws. However, especially with a kit, performance in a consumer’s case can vary from these results. For example, if you place the radiator in a case with restricted airflow, you will not achieve the performance levels shown below.
Note also that changing components will impact results – if you substitute an inferior (or better) waterpump for an EHEIM, you will not see these results.
Last, and most important, actual motherboard results may be substantially better than CPU Die Simulator results.
The Die Simulator tests CPU cooling solutions “in utero” – that is, it excludes any secondary heatpath effects which are found on motherboards. It is a pure test of cooling performance.
Secondary motherboard heatpath effects, overall, usually result in better performance observed by consumers. The ratio between Watercooling Die Simulator results and motherboard results, so far, averages about 60%. For air cooling, the ratio averages about 80%.
I think this is due to watercooling’s superior capacity to absorb heat – it’s taking more waste heat off the motherboard than air cooling solutions, resulting in enhanced performance vs air cooling.
However, because motherboard, CPU and case results are so variable, relative rankings are based on the CPU Die Simulator, not on a specific motherboard/CPU/case combination.
Each system was tested in an open case environment – no case fans were used during the test and the CPU Die Simulator was not enclosed in a case. Each system was limited to its waterpump, radiator, radiator fans and waterblock. Airflow through radiators was unrestricted.
Please view each test report for details.
|Swiftech Q-Power, 77.8w|
|maxXxpert MXL-MIRO, 77.3w|
|Koolance, Fan HIGH, 76.9w|
|Koolance, Fan LOW, 77.3w|
C/W = Delta / CPU Watts
Interpreting C/W: For every watt the CPU radiates, each system will cool the core by the (C/W x watts) plus ambient temp. For example, at an ambient temp of 25 C, a C/W of 0.25 with a CPU radiating 50 watts means that the CPU core temp will be 50 x 0.25 = 12.5 C over ambient temp, or 37.5 C. The lower the C/W, the better.
From a noise/performance standpoint, any of these cases/kits are superior to aggressive air cooling solutions, although the better air cooling solutions may yield similar cooling results. However, there is more “overhead” with watercooling, so that peltier solutions are more viable with water compared to air cooling.
These tests are conducted at a constant heat load. In actual use, users will see little or no difference among closely ranked systems. A C/W difference of 0.02 translates to an observed temperature difference of 2 C at 100 watts. For a CPU radiating half that – 1 C.
Motherboard CPU sensors are notoriously inaccurate and are not test instruments – small differences such as this will most likely not show up on-screen.