I suggest that PRICE be included in the benchmarking of all CPU's. Perhaps we can take whatever # generated by a benchmark and divide it by the cost of the CPU itself so that we can have performance per $ as a nice scale to go by. If we're testing Frames per second, fine... make it FPS/$ so that we can see what we're paying for. I don't think that just because Intel has taken "the performance lead" means it is the superior processor for average users. Who exactly is going to pay the ultra-high price for a top of the line P4 anyway? Price has always been a good industry standard for what is best for consumers, and I seldom see it included along with any benchmarks. For instance, say I'm willing to pay $200 for a processor... what can AMD and Intel offer me at that price, and which one is better? Of course both AMD and Intel can roll out chips that are faster, but in limited quantities because they know they'll be higher priced! The "who's faster" argument isn't as important as "who's faster for the money" these days because even the slowest computers selling today are good enough for average users.