"The next three - the TaisolCEK733, the Zalman, and the Alpha PAL6035 - are acceptable for OC use. I don't know if I would throw them out quite yet, but they would not be on my list to check out unless you can just find a hell of a deal on them."
", I personally like to test HSFs on a mainboard. Why? Because testing them in that manner most closely approximates the way they are intended to work"
Well, I don't buy the fact that the PAL6035 performs as badly in real life as it did in those "tests". Also, it doesn't matter if you set up a TEC tester a la 2cooltek or try it in a mainboard as long as you do the same for every heatsink. What's important is the accuracy of the test. As Mikewarrior has shown, using the thermistor on a T-Bird mainboard seems to be the least accurate way to test a heatsink. You can determine "real world performance" by looking at the temp differences between heatsinks that are tested alike. If I compare an Alpha to an Orb and find out that the Alpha performs 10 degrees better than an Orb on a TEC tester, than it's not farfetched to assume that, given proper circulation in the case, the Alpha will perform about 10 degrees better than an Orb in your computer. I suppose you also have to consider how rapidly an HSF might approach heatsoak. It won't do you much good putting a golden orb on that T-Bird if it only keeps it at a reasonable temp for half an hour. It was neat to see that Korean heatsink, but beyond that I gained nothing from that review.
