|Best CPU Cooler Performance LGA1366 Q2-2009|
|Reviews - Featured Reviews: Cooling|
|Written by Olin Coles|
|Tuesday, 04 August 2009|
Page 13 of 18
Cooler Test Methodology
Benchmark Reviews has been obsessed with testing CPU coolers over the few years. We've solicited suggestions from the enthusiast community, and received guidance from some of the most technical overclockers on the planet. As a result, our testing methodology has changed with every edition of our Best CPU Cooler Performance series. Because of this, each article is really its own stand-alone product, and cannot be fairly compared to the others. This article is going to be a perfect example of that principle, although certain tenants still hold true. Benchmark Reviews continues to test CPU coolers using the stock included fan (whenever applicable), and then replaces it with a high-out fan for re-testing.
Manufacturers are not expected to enjoy this sort of comparison, since we level the playing field by replacing their included fan with a common unit which we then use for every CPU cooler we test. Manufacturers regularly include fans with their CPU cooler products, and more often than not these fans are very high RPM units which offer great airflow at the expense of an obnoxiously loud noise level. By using the same model of cooling fan throughout our testing, we can assure our results are comparable across the board. This is one of the more significant changes we have made to our test methodology, since many of the benchmark tests we have conducted in the past have compared the total package. Ultimately we're more interested in the discovering the best CPU cooler performance and we believe that you'll feel the same way.
Testing was conducted in a loosely scientific manner. Ambient room temperatures levels were held to within one degree of fluctuation measured at static point beside the test equipment with a calibrated digital thermometer. All coolers had their original manufacturer-supplied thermal material removed and replaced with a product of our choosing listed in the support equipment section below. Each product then received the same amount of Thermal Interface Material (specified below), which amounted to roughly a BB-sized drop placed onto the center of the CPU. The CPU cooler product being tested was then laid down flat onto the CPU, and compressed to the motherboard using the supplied retaining mechanism. If the mounting mechanism used only two point of force, they were tightened in alternation; standard clip-style mounting with four securing points were compressed using the cross-over method. Once installed, the system was tested for a baseline reading prior to testing.
At the start of each test, the ambient room temperature was measured to track any fluctuation throughout the testing period. EVEREST Ultimate Edition v5.02.1750 by Lavalys was then utilized to create core loads and measure each individual CPU core temperature. It's important to note that software-based temperature readings reflect the thermistor output as recorded by the BIOS. For this reason, it is critically important (for us) to use the exact same software and BIOS versions throughout the entire test cycle, or the results will be incomparable. All of the units compared in our results were tested on the same motherboard using the same BIOS and software, with only the product itself changing in each test. These readings are neither absolute nor calibrated, since every BIOS is programmed differently. Nevertheless, all results are still comparable and relative to each products in our test bed.
One unfortunate problem is that CPU's report temperatures as a whole number and not in fractions. This in turn causes the motherboard BIOS and subsequent software applications such as EVEREST to also receive whole-number reports. Thankfully, EVEREST also does offer averages in the statistics panel, which gives us more precise readings. To further compensate for this, our tests were conducted several times after complete power down thermal cycles. Conversely, the ambient room temperature levels were all recorded and accurate to one-tenth of a degree Celsius at the time of data collection.
When each cooler is tested, Benchmark Reviews makes certain to keep the hardware settings identical across the test platform. This enables us to clearly compare the performance of each product under identical conditions. While the ambient room temperature did fluctuate between 18~19°C, this would not be enough to cause a noticeable impact on our test results since only the thermal difference is reflected in the charts. For the purpose of this article, thermal difference (not the same as thermal delta) is calculated by subtracting the ambient room temperature from the recorded CPU temperature.
All of our tests are now conducted using only the vertical motherboard orientations traditional to tower computer systems. At the start of our test period, the test system is powered on and EVEREST system stability tests are started with Stress CPU and Stress FPU options selected. For a minimum of sixty minutes (one hour) EVEREST loads each CPU core to 100% usage, which drives the temperature to its highest point. Finally, once temperatures have sustained a plateau, the ending ambient room temperature and individual CPU core levels are recorded thus completing the first benchmark segment.
The second test segment involves removing the stock cooling fan (while the system is still under load) and replacing it with a high-output 120 mm Yate Loon D12SH-12 cooling fan. The system is given thirty additional minutes with EVEREST loading the CPU cores before final temperature readings are taken and recorded. Once the testing has been completed at the stock processor speed, Benchmark Reviews turns up the voltage on our Core i7-920 processor and overclocks to 3.6GHz using a 1.35V vCore. When the system restarts, we start our testing over from the beginning and allow a minimum of sixty minutes of loading before taking our readings.
The Accuracy Myth
All modern processors incorporate an internal thermal diode that can be read by the motherboards' BIOS. While this diode and the motherboard are not calibrated and therefore may not display the actual true temperature, the error is constant. This means that if the diode reports 40°C when it is actually 43°C, then it will also report 60°C when it is truly 63°C. Since the design goal of a thermal solution is to keep the CPU core within allowable temperatures, a processor's internal diode is the most valid means of comparison between different heatsinks, or thermal compounds. The diode and motherboard may be incorrect by a small margin in relation to an actual calibrated temperature sensor, but they will be consistent in their margin of error every time.