Aerocool Extreme Socket 478 HSF ReviewFormer staff writer -
Intel Pentium 4 @ 2533MHz
ASUS P4S533 Motherboard (SiS645DX)
2x Samsung DDR333 256MB Modules
Verre V770 Black, NO system fans
Arctic Alumina Thermal Compound
Running Windows 2000 Professional, SP4
Arctic Alumina was used because my local computer/electronics store in did not stock Ceramique yet, and they didn't have any Silver III in stock (and also because my tests with the other heatsinks already used Arctic Alumina). Expected temperatures were first determined by using Motherboard Monitor, and then verified using an IR temperature sensor, recording the highest temperature achieved in the scanning area with less than 0.5 degrees difference. For recording idle temperatures, the computer is left alone for 15 minutes. For the load temperatures, I used SiSoftware Sandra, using the Burn-In test, running in the background for 15 minutes. The other heatsinks I've compared here are the Zalman CNPS5700D-Cu, as well as the Thermaltake Volcano 7+, both at various fan speeds.
In my particular system setup, the V7+ on high setting pushed so much air through that not all of the warm air could be circulated out by the PSU fan, and in return, some of the warm air got recycled, which explains the slightly higher temperatures when compared to itself on the medium setting.
Overclocked results for the stock Intel HSF and the AeroCool Extreme were also determined. Using the stock heatsink, I was able to reach 148MHz at 1.6V, while with the AeroCool Extreme I peaked at 155MHz at 1.7V. In order to achieve these clocks I changed the FSB to RAM ratio to 1:1, so that my modules would not become the "ceiling" for overclocking. Thanks to the idle algorithm present in both Windows 2000 and Windows XP, idle temperatures are usually lower than expected, given the operating environment. This is evidently shown in the highest overclocked test, where the processor temperature jumped almost 20* degrees Celcius. While the copper heatsink is able to conduct heat quickly off the CPU, the fan is unable to exhaust it fast enough. However, given that there were no other sources of airflow, this is expected. A system equipped with proper airflow will significantly help in reducing CPU temperatures. * If memory serves me correct, this CPU should begin throttling at 60 degrees Celcius, which may explain why the temperature did not rise above, even it had been running at 59~60 degrees for over 1/2 the time during the Burn-In test.
As an HSF, this unit does its job very well, beating the other three heatsinks in the idle test, and sitting in the middle for the load test. However, due to its peciliar design, I strongly recommend against using this CPU under an Intel configuration, as there is no guarantee that the force of the clips alone can generate enough friction to keep the heatsink in place if the tower is stood upright. As an AMD heatsink, the fan and heatsink are already attached as one unit, so the supplied socket A clips can keep the unit securely in place.
With that said, if you do not plan to overclock your Athlon too heavily, but want a unit that looks flashy in blue and orange, be sure to consider the AeroCool Extreme. Watch those fingers though, there's nothing between you and the spinning blades out-of-the-box!
Relatively quiet compared to other third-party HSFs
Strong performance-to-noise ratio
Very nice blue/orange lighting combination