Testing Methodology
Testing Methodology
For this review we wanted to give you readers the most complete tests possible.. So, we decided to test the NF-P12 in both air and water cooling applications against some of its closest competitors. While the list of fans we tested it against does not represent ALL of the Noctua’s competitors, these are the fans we could get our hands on in short order.
Fans Used
Noctua NF-P12 (1300 PRM)
Silverstone FN-121 (1200 RPM)
Ultra Kaze DFS123812L-1000 (1000 RPM)
Yate Loon D12SL-12 (1350 RPM)
Air Cooling Testing Methodology
For the air cooling tests the following system was run:
System Used
Processor: Intel Core 2 Quad Q6600 (B3) @ 3.5Ghz @ 1.494V (in BIOS)
Memory: 4GB Corsair Dominator DDR3 @ 1556Mhz (7-7-7-20)
Motherboard: Asus Blitz Extreme
Graphics Card: EVGA 8800GT Superclocked
Power Supply: Silverstone DA850
Disk Drive: Pioneer DVD Writer
Hard Drive: Seagate Barracuda 320GB SATAII
CPU Heatsink: Thermalright Ultra Extreme (single and dual fan)
OS: Windows Vista Ultimate x64
Temperature Logging Program: Core Temp
CPU Load Program: Prime 95 (Small FFTs)
Ambient Temperature Measurements: Extech EX430 with Type-K Thermometer
Room Temperature: 21.3*C
Relative Humidity: 52-56%
In these tests we are aiming to push the most heat possible through the air cooler in order to accentuate the differences between the fans and that is why the vCore was bumped up so much. This is also why you will see some pretty wild temperatures coming from the Q6600 in all of the air cooling results.
As you will see in the benchmarks, we ran tests using push and push / pull fan configurations in order to give you the best idea of how your setup will react to these fans. Due to the lack of an extra 120mm fan mounting bracket used for the push-pull configurations on the Thermalright Ultra Extreme heatsink, all fans were mounted using small pieces of Velcro Ultra Mate in each corner. All other fans on the test bench were disabled.
Temperature Control
In order to better control ambient temperatures, an open test bench was used for all air cooling tests. All of the ambient temperatures were taken using a Type-K thermometer about 7” in front of the intake fan as shown in this picture (note the black circle):
Between each test, the system was shut down for 2 hours in order for the room temperature and the system temperature to normalize. The next test was then begun and the process was repeated for each fan configuration.
Temperature Testing & Results Analysis
All temperatures were logged with Core Temp over a 15 minute period for both Idle and Load results. The processor was allowed to display a stable Windows desktop for 15 minutes followed immediately by 4 instances of Prime 95 (small FFTs) being launched to load the processor for another 15 minutes.
The temperatures were then averaged out to give the final temperatures you see in the charts. The overclock on the processor is stable so every 30 minute test (idle + load) was completed.
Water Cooling Testing Methodology
In the water cooling tests we once again tested with push and push/pull configurations with a single fan rad. For these tests, all of the fans were shipped to our resident water cooling reviewer and he had a go at them so the testing methodology and system change slightly.
System Hardware:
Processor: Intel Core 2 Duo Q6600 G0 @ 3.6GHz 1.5Vcore (in BIOS) 9x400
Memory: 2x1GB Mushkin XP2 PC2-6400 @ 1.9V DDR2-800 5-5-5-15
Motherboard: Gigabyte GA-965P-DS3 rev.1.0
Graphics Card: Geforce MX 440 PCI
Power Supply: OCZ GameXStream 600W
Hard Drive: Samsung 80GB SATA2 HDD
Disk Drive: Lite-On IDE DVD-ROM
OS: Microsoft Windows XP MCE 2005
H20 Cooling Hardware
D-TEK FuZion (unnozzled, but with gasket)
Swiftech MCP655 vario @ 5 (D5)
Clearflex ½” tubing
Swiftech MCRES-MICRO
Hardware Labs Black Ice Xtreme single 120mm rad
Pure distilled water
Temperature Logging Program: Core Temp
CPU Load Program: Prime 95 (Small FFTs)
Ambient/Water Temperature Measurements: HH506RA multilogger thermometer with one type-T probe & one type-K probe
Room Temperature: Varies (see below)
Here is a quick look at out test setup
Our testing methodology for this part of the review is a little bit different than the air cooling section. Rather than reporting straight CPU temperatures, we will be reporting the “delta T” between the water and the ambient air in the room. The Type-T probe was used to measure water temperatures while the Type-K probe was used to measure ambient temperatures. In order for our readers to better understand why we are testing things the way we have, we’ve created this short explanation of what a “delta” is.
What is a “Delta”?
Delta simply refers to a difference (in this case in temperature between) two things. To give a quick example, all other things being equal, a room with a 20C ambient temperature (the temperature of the air) would have water temperatures that are 5 degrees lower than a room with a 25C ambient temperature. Using delta T allows us to compensate very easily for small fluctuations in room temperature.
Delta T also gives our readers an easy way to determine what their results would be if they had the same setup as we do. For example, if our results with one of the fans was a delta of 15 degrees between air temperature and water temperature, then no matter what the temperature of a room, you could, with reasonable accuracy, deduce what temperature your CPU would run at with our test setup (at the same overclock settings of course).
So, to put it plainly: the Delta T we are using is the difference between the room and water temperature.
More about the Methodology
Now on to the testing methodology: The CPU remained loaded for the duration of the evening that we performed our testing. It was never turned off, and was running 4 instances of Prime95 small FFT for the entire test period. Each fan was hot-swapped onto the running computer, then water temperatures were given 20 minutes to reach an equilibrium (meaning they were not changing anymore), then temperatures were logged over 5 minutes to get our readings. After that, the next fan was tested. Temperatures were recorded using Core Temp plus the Type-T probes reading water and ambient temperatures.
So all of that said, how do you interpret the results? The fan that yields the lowest delta T on our test radiator is simply put, the best fan for high restriction applications. There are few (PC) scenarios that this reviewer can think of that are more restrictive on a fan than a Black Ice Xtreme, so that radiator was chosen for this review.
