To measure the CPU's temperature we used SpeedFan v4.21, which provided us with real time monitoring and created a graph of the temperature over time. After letting the processor cool down in idle mode, we fired up Prime95 and ran the in-place large FFT test to heat it up to its maximum level. This is the graph of the temperature with the original AMD fansink:
|Starting from 38C...|
|and the reported fan speed...|
As we can see from the above graph, the reported temperature with the AMD stock fansink is 38°C in idle mode and as the torture test commences, the CPU temperature rises up to 51°C.
After installing the Freezer 64 we waited with the CPU in"idle" to cool down and then started the FFT torture test once more...
|Starting from 34°C in idle mode...|
|the RPM graph under the Freezer 64|
As we can see from the above graphs, the temperature in idle mode with the Freezer 64 installed is 34°C, which is 4°C down from the stock AMD cooler and when the Prime 95 FFT torture test starts, the max temperature reported is 45°C which is 6°C lower. Also note the Freezer 64's RPMs, which is the reason it is such a silent cooler...
The default settings for our AMD 3500+ were 11 x 200MHz = 2200GHz. From the BIOS, setting the CPU Voltage and FSB ratio, we managed to overclock are processor. We settled on setting the voltage to 1.6V and the FSB to 233MHz. We ended up with an 11 x 233 = 2563 AMD OC processor.
At that speed, the system was able to boot and even when running Prime95 which is very good at identifying errors, it reported only correct calculations.
Let's see the graph showing the temperature of the CPU starting from idle and during the Prime95 torture test.
|an total show off of the "Freezer" 64|
The idle temperature now starts from 37°C due to the increased Voltage for the CPU. As the torture test ran, we ended up with a stable 51°C.
Great performance from the "Freezer 64" CPU cooler in all cases, both with normal and overclocked processors. It's sure to impress overclocking enthusiasts. Thumbs up for Arctic Cooling.