Performance
Test setup and Testing Methodology
The system used for testing is listed in the table below. Ambient temperatures were kept at 24 degrees Celsius +/- 1 degree. The thermal paste used was CoolerMaster thermal fusion (for testing consistency). For the overclocked results I upped the voltage to 1.30v and attained a clock speed of 4.2GHz. The cool and quite mode was disabled to prevent the CPU idling at lower voltages to attain accurate temperatures at idle. The fan speed was set to auto. The idle temperatures were recorded after 10 minutes of idle and max temperatures were recorded after a 5 minute torture test using Prime95. The software used to monitor temperatures was HWMonitor courtesy of CPUID.
In the graph above we see the idle temperatures for both stock and overclock. Now the NH-U9S is towards the bottom but remember these are idle temperatures and the fan on the Noctua NH-U9S is merely spinning 1200 rpm and is hardly noticeable at that RPM.
Now you can see that when we turned up the heat the cooling capabilites started to show as the NH-U9S moved up the ladder and when you had the second fan installed the NH-U9S blew the doors off the competition! With dual-fans I seen an increase of 9% cooling performance. LOOK, WHEN SETUP WITH DUAL-FANS, IT HAD A COOLER TEMPERATURE READING THAN THE NH-D15! You may ask well why did it have a hotter temperature on stock settings? That is because the Voltage difference the PWM it is getting from the board and thus the fans will spun a few hundred RPM faster. The dB level of the Noctua NH-U9S was barely noticeable compared to the ambient room sound. The fans spun up to 1803 RPM at full load on overclock and at idle the fans spun around 1149 RPM.
One thing Noctua offers with this cooler is a set of low-noise adapters. The purpose of the low-noise adapter cables are to reduce the voltage sent to the fans, in turn limiting the RPM at which they will spin. The low-noise adapters will prevent the fans from spinning over 1200 RPM. This accessory is a nice added bonus from Noctua. As you can see from the results below we did lose performance in the cooling area but we did achieve dBs at lowers levels.
At idle, the difference showed to be a couple degrees. At idle the fans spun around 900 RPM. Under load with the adapters the fans pegged out at 1203 RPM and thus limited the cooling capabilities.
With the low-noise adapters under load the NH-U9S struggles to compete. Really the reduction to the fans are not needed because the fans with out the low noise adapters are hardly noticeable. The amount of performance lost due to lower fan speeds does not justify the lower decibel level achieved. The 92mm fans where meant to run at high RPM and those RPM really are need to keep this cooler competitive.
There we have it! The NH-U9S brought to us by Noctua! Let us cover a few final thoughts and conclusion of our review.
