Benchmark Results: Temperature, Noise, And Acoustic Efficiency
Because liquid cooling systems are designed to be mounted in a case, we closed our side panels before taking thermal and noise readings.
Voltage regulator cooling is always a big concern whenever fans are moved away from the processor. Fortunately, Asus’ Maximus IV Gene is designed specifically for this type of build. As a result, we only see important differences in CPU temperature, where Asus' “Auto” fan control mode isn’t adequate for our overclock. Thermal throttling forced us to give up on automatic controls for the H₂O 620, so we tested it only at its maximum fan speed.
With a wide dual-fan radiator, the H100 easily tops our thermal charts. Meanwhile, Evga’s SuperClock tower-style heatsink and single fan prove the adequacy of cooling without pumps.
Automatic modes require a hot processor for evaluation, so isolating the cooler noise could only be achieved at maximum speed settings by connecting the fans directly to a silent 12 V power supply. Full system noise is far more pertinent however, and this is where hot temperatures give way to quiet measurements with the Kühler H₂O 920.
We calculated the average temperature and noise level for all coolers and developed a cooling-to-noise-ratio chart by simple division, which made the average 100%. Subtracting one from the result gives us a 0% average, where the charted results show gains or losses compared to the average of all coolers in this test.
Bold claims by Corsair and CoolIT about their new cold plate design are proven true in our Accoustic Efficiency calculations, which partly cancel out the benefits of super-fast performance- or super-slow silence-oriented fans. The Hydro H100 tops the chart, followed closely by its shorter/fatter H80 sibling.