Oh, I see. Should've read the article.
Big fucking whoop, then. How is this a giant step forward? Weren't the PowerPC chips better anyways?
Sort of. Technically, they had better branching instruction performance, but to get the full power out of them required dedication and attention to the design docs, that kinda doesn't make sense. Most developers were used to the x86 assembly anyways, and could get better stuff out of that. Ryan Gordon said it best, the PPC was better in theory, but to get that kind of power was a huge pain in the ass, so why not half-ass it and get 85% of the power anyway? So that's what happened. I mean, you can get away with coding around the PPC like it was x86. Admittedly the G5 was a lot better, and it had kickass floating point performance, but then IBM... well... they never got above 3 GHZ. And the whole laptop thing... yeah.