May have figured out the discrepancy in Arcticooler reviews

[Strafe]

UPDATED WITH LINKS

Well, maybe not but it is something to think about...

I took a look at the Arcticooler review on HardOCP (http://www.hardocp.com/reviews/cooling/arctic_ca/index.html)and noticed that Kyle's readings were with the CPU at (most likely) full load. Then I looked at 2Cool's review (http://www.2cooltek.com/socket_test.html) and it looks like his temp readings were taken at idle. But the interesting thing was when I read a post on this forum (http://forums.anandtech.com/messageview.cfm?catid=37&threadid=233690) of bmg's personal experience with an Arcticooler. He compared it to his Alpha I believe. With both the Alpha and the Arcticooler the idle temps were the same, it was not until a full load was applied that the Arcticooler kept the CPU cooler than the Alpha.

 

[2Cool]

Heh... I do not even use a CPU to test the heatsink, rather a peltier that supplies a continuous ~30 watt heat load. Sorry, but you cannot get consistent results on a system, as even a TSR or background program can throw the whole thing off. Most review sites run QIII or whatever in loop, then stop and take a temp reading. They get a temp at near idle reading, kind of useless. You really have to have a consistent temp load to be able to test them, exactly as the manufacturers rate their products.