VIA Hardware has posted their analysis of the Aberdeen white paper on AMD's Model numbers.

We take a look at how Aberdeen went about this white paper, a look at AMD's model number rating system, and our own analysis on the subject. This report has been controversial since it was issued, so we took the time to break it down, examine its findings, and let you know if its conclusions held water.

This is the inherent problem of benchmarking a system. Under one configuration a P4 or Athlon has one set of performance scores, swap out some patches or updates, and its scores change dramatically. Which do you assume your users have? How many people patch their computers? Given that most don’t, should you test with the base version (and give an answer that reflects what most people will see) or patch (and give an answer that properly reflects what the computer is capable of?)

Read more