[bigben2k]

First let me thank Incoherent for an excellent post (#67). I too refuse to believe that using a heat die / simulator is not possible. To add to your comment, I often confuse the power level question, because I don't have a firm grasp on the effect (since I haven't started my testing yet), but last I checked, it had no effect, which leaves (me with) a lot of doubts/questions on the results.

Festivus (for the rest of us):
Using a CPU leaves us with an unknown Q (wattage), which is also why I choose to stay away from it. We've been over this in the past, countless times. While pHaestus uses it, and can get it to heat up to approximately the same level, it makes for a reproducible heat source, which is good for comparative testing, not analytical testing (i.e. measuring the actual performance).