Test Setup
| Test Platform: | | Motherboard: | EVGA X58 SLI | | Processor: | Intel Core i7 965 Extreme Edition | | Processor Cooling: | Thermalright Ultra-120 eXtreme-1366
Included Thermalright 120mm 1600RPM/63.7CFM
120mm ADDA AD1212MS-A73GL 2050RPM/80.5CFM
120mm VANTEC SF12025L 1500RPM/50CFM
Intel Stock 965 EE heat sink | | Thermal Paste: | Arctic Cooling MX-2 | | North Bridge Cooling: | Stock | | South Bridge Cooling: | Stock | | PWM Cooling: | Stock | | Memory: | Corsair Dominator 3x2GB PC3-12800 8-8-8 (TR3X6G1600C8D) | | Power Supply: | Ultra X-Pro 750W | | Video Card: | BFG GTX 280 OCX / BFG GTX 280 OC | | Additional Fans: | 120mm AD1212MS-A73GL 2050RPM/80.5CFM | | Hard Drives: | Seagate 7200.9 80GB SATAII 8MB cache | | OS: | Windows Vista SP1 (with all updates) | | Ambient Temperature: | 23C ~ 25C |
There are a couple items we want to discuss in our setup to start out with. As we can see from the photos, we have gone with the east/west orientation for testing to simulate exhausting the hot air out of the rear of a case. We figured this would be the most likely situation for the majority of users despite the amount of cases with top exhaust fans these days. Another item of note is the fan selection as we will be going with quite a few combinations to make up for the lack of CPU coolers being tested.
We are also obviously using an open bench setup for all of the testing. We will measure ambient temperature according to the thermal probe directly above the CPU heat sink by about eight inches. The digital thermometer used is an Extech TM200. The ambient temperature range moved in-between 23C and 25C. That should cover any details relevant to testing about the setup, let's now have a look at the testing methodology we used. Testing Methodology Our test setup is fairly straight forward with just the two heat sinks but there are a few things that we want to explain to help add context to the results. As mentioned, all testing is done on an open bench in ambient temperatures between 23C and 25C as measured from right above the system with a digital thermometer. We did not re-mount the Thermalright Ultra-120 eXtreme, we simply mounted it once and ran it under load/idle cycles for a couple of days to ensure any curing of the thermal paste had occurred. This should provide comparable results as the numbers won't rely on a different mount for each fan configuration.
To provide the system load we will run Prime95 x64 v257 on "blend" for 20 minutes. The system is then left to idle for 10 minutes after stopping Prime95. The temperatures will be recorded with Everest Ultimate at 5 second intervals. Fan speed control for all fans will be a constant 100% so they maintain an even RPM, including the provided Thermalright fan.  Speaking of fans, we have essentially twelve different configurations to run the Thermalright Ultra-120 eXtreme. This should provide a solid idea of how it performs in various situations. We will be running the heat sink in a pull, push, and push/pull orientation at both stock clocks and our overclocked settings. For the push/pull configuration we will use the included fan for pushing and the Vantec Stealth outlined in the setup section for pulling. We didn't have two Thermalright fans so we used the closest rated fan we had on hand. These three orientations will then be run through both sets of clocks with higher RPM ADDA fans. Here is a screenshot of the overclocked settings as pulled from the EVGA X58 SLI review:  For the overclocked testing, the system will be running at 3.9GHz with a vCORE of 1.35v under load. This definitely isn't the highest overclock we have seen an i7 pull on air, but it is still a substantial increase in heat load from stock and very respectable. This type of over clock will also increase temperatures for the PWM as well so we not only will be measuring the core temperatures of the CPU, but also the PWM temperatures as this motherboard requires acceptable PWM cooling in order to run stable at higher clocks. The CPU heat sink and fan setup can directly influence the PWM temps so we wanted to see how the fan configurations would affect these temperatures. The graphs we are about to look at are going to show the six different fan configurations and the stock cooler result. The Ultra-120 eXtreme results are an average of the 4 core temperatures at each 5 second interval. There should be no more questions about the results, so let's see how each configuration on the Ultra-120 eXtreme does.
| 
