Corsair Blog

I recently did some testing with our new Hydro Series™ H60 high performance liquid CPU cooler. The H60 comes with a stock fan rated at 1700 RPM, which is designed to deliver good static pressure against the radiator without excessively high RPMs, and the accompanying noise levels. Of course, like any enthusiast I decided to test the H60's performance.

My current test system consists of an EVGA® 759 Classified motherboard, an Intel® Core™ i7 920 D0 stepping CPU, and Corsair Dominator® GT 2000C7 DDR3 memory. I did a bit of quick tweaking and managed a stable overclock of 189 x 20 on the CPU for a CPU frequency of 3.770GHz. The memory clocked in at 942MHz, 7-8-7-24 1T timings. The actual voltages under load, measured with a DMM, varied a bit from what was set in the BIOS. Measured, the CPU Vcore was 1.33v, Vdimm was 1.65v, and CPU VTT was 1.41v. The CPU did not respond well to voltage increases after this point and this was the maximum stable overclock using Prime95 over 24 hours.

I wanted to use a utility to not only test the system for stability, but to also monitor the temperatures during the testing. I decided to use the most recent beta version of the AIDA64 PC system utility. Among the many useful features, there is a system stability test. This test places a 100% load on the CPU. Hyperthreading is enabled and all 8 virtual cores are stressed.

The utility allows you to also monitor and log quite a few different system parameters including the main CPU temperature and individual CPU core temperatures. I compared the AIDA temperatures to those shown by the EVGA Eleet monitoring utility, and both showed the same temperatures within one degree during testing.

I decided to run the test for a minimum of one hour and log the results. My test area is in a basement, and my ambient temperature was a nice steady 23C. Here’s the AIDA summary for 1 hour at 100% CPU load at the settings listed above. Not bad at all for what appears to be an only moderately overclockable 920.

We took the new H60, installed it onto an OC'ed Core i7 920 CPU, and tested the cooling performance using the stock fan. Then, we compared those results to 3 distinctly different pairs of fans; the Quiet, the Loud, and the Ridiculous. Come on inside and see how they compare.

The Stock

I discussed the results with George Makris, the product manager of our cooling components. We were both curious as to how the H60 would do on this rig with a dual fan setup. So, I gathered up 3 distinctly different pairs of fans for testing: 2 x Enermax UCAP12 fans rated at 700RPM – 1700RPM, 2 x Sunon KD1212PTB1-6a fans rated at 3100RPM, and 2 x Delta QFR1212GHE fans rated at 6000RPM. I opted to test quiet, loud, and obscenely loud fans just to get a good look at a wide spectrum of fans, in addition to the single stock fan. During the testing, the pump and the fans were attached directly to the PSU to ensure a solid steady 12v power delivery.

The Quiet

My first run was with the Enermax fans. These fans have their own thermal control probe and this probe varies the fan speed dependent upon the temperature.

The probes were in open air, so the fans stayed very close to 800RPMs the entire test. They cooled acceptably well and performed virtually identical to the single stock fan. To the naked ear, the two Enermax fans are very close in noise level to the single stock Corsair fan. The CPU core temperatures, while in the same range, were slightly higher than the temperatures achieved using the single Corsair fan.

The Loud

I tried the two Sunon fans next. These fans are rated at considerably higher RPM ratings and will definitely move a lot more air than either the single stock fan or the two Enermax fans. Of course, a lot more noise is the penalty for moving a lot more air.

My test bench has a 120mm fan bracket that I attached to the side of the bench for easy access. This made changing the fans out much easier than it would be inside a case. Here's an action shot where I was changing from the Enermax fans to the Sunon fans.

It's no surprise that the two Sunon fans dropped the temperatures. But it is quite a trade-off as they are considerably louder than the Enermax fans. If this was a daily user machine sitting close to you in a home or office environment I don't think the noise level is acceptable.

The Ridiculous

The mighty Deltas are thicker than the standard 120mm fans and required longer screws to mount.

If you have never heard two 6000RPM Delta fans running, it’s somewhere in the neighborhood of a hair dryer or vacuum cleaner. They move a massive amount of air, but this is not a sound that any computer user would realistically want to sit in the room with. But, they do cool the CPU the best of the 4 tested configurations. However, the two Deltas at nearly twice the RPMs of the Sunons only managed to cool an additional 2°C over the Sunons.

I have to admit that I expected a much larger delta between the stock fan and the two Delta fans, no pun intended. Based on the testing, which is admittedly limited in scope, I do think that we can determine that the single stock H60 fan in combination with the radiator and block/pump assembly is a very efficient setup. After that, yes, you do get increased cooling with the higher speed fans but it comes at quite a cost in terms of decibel level.

My conclusion is that two low or moderate speed fans are a great option for cooling on the H60 for users that wish to upgrade over the stock fan and maintain acceptable noise levels. And for those that want ultimate cooling and can stand being in the room with a food processor, there’s always the 6000RPM Deltas!

The Video