| Thermalright IFX-14 CPU and Back-side Heatpipe Cooler
|
Testing:
I will be comparing the IFX-14 and its back-side cooling solution to three other well known coolers. From left to right in the below left image we have the contenders; the IFX-14, the Sunbeamtech Tuniq Tower 120, the Zalman 9700 CNPS, and the stock Intel cooler . Other than the stock Intel unit, these other coolers are very popular and should serve as a good performance benchmark for Thermalright's IFX-14.
 
The above right image takes a closer look at the bases of the big heatsinks involved, and the difference in heatpipe diameter should be clear.
It is important to note that all of these coolers use different fan solutions, so although the numbers give us a good approximation of cooling ability, they don't really give us an accurate measurement of each cooler's cooling efficiency or potential. For testing purposes I have paired the IFX-14 with two Vantec Stealth 120mm fans. These are 28dBA fans that move approximately 53CFM of air at 1500RPM. The fans on the other heatsinks were the stock fans they shipped with. I ran all fans at their maximum speeds and did not use any of the motherboard's fan control features. Temperatures were measured using CoreTemp 0.95.4, which accesses the processor's on board thermal sensor. This is not the most accurate way to measure CPU temperatures, but it has the advantage of being consistent throughout the test. The actual numbers aren't as important as the differences between them. To obtain loaded temperatures I used Windows Prime95 v25.6, build 6. I chose this version specifically for it's ability to use all four cores of the Q6600. There are four different testing options with this version of Prime95, and the test I chose to run was the "In-Place Large FFTs" test which primarily loads the CPU, rather than both the CPU and the RAM.
For this test I ran each heatsink at the CPU's stock speed of 2.4GHz (1.22V, 9x266MHz) and then again overclocked to 3.0Ghz (1.35V, 9x333MHz). I measured two temperatures for each cooler at both speeds. The first temperature was taken approximately five minutes after booting into Windows with the CPU at idle. The second measurement was taken after running Prime95's large FFT's torture test for at least one hour. Ambient temperatures throughout the testing period were between 20.6C and 21.1C (69F-70F).
You can see that the IFX-14 appears to do slightly better than the Zalman 9700 and the Tuniq Tower 120. However, if you factor in the slight change in ambient temperature throughout the testing period, it essentially becomes a dead heat between the three competitors. What is most notable from this test is just how poorly the stock Intel heatsink performed. I should mention that I only ran the overclocked prime95 test for 15 minutes with the stock Intel heatsink because I saw no benefit to exposing the CPU to the extremely high temperatures for more than that.
What does this test tell us? Specifically it tells us that the IFX-14 with 2 Vantec stealth 120mm fans performed as well or slightly better than the Zalman 9700 and Tuniq Tower 120 with their respective fans. Keep in mind that the IFX-14 has the option to use as many as 3 fans of either 120mm or 140mm size. The Tuniq Tower 120 can certainly be fitted with a more powerful 120mm fan as well. Zalman's 9700 on the other hand has an integrated fan so although it's not impossible to change, it's not a trivial matter either.
The IFX-14 was the quietest of the three aftermarket heatsinks tested despite having two fans instead of one. The only fan quieter was the fan on the stock Intel heatsink. In the world of air cooled processors, quiet generally means less cooling. In this case, we were able to get good performance and still maintain acceptable noise levels.
|
|
|
|
