While we still suspect some minor errata is part of the reason for the delay, it’s possible marketing played a role, too:
“Intel has already notified its partners that it will push back the launch of the three CPUs to February or March next year, depending on AMD’s schedule for triple-core and the upcoming Phenom CPUs.”
Waiting until AMD launches sets up a review showdown, one Intel obviously feels quite confident it will win, which will give AMD more bad publicity.
Intel may also be pushing some of AMD’s well-used buttons to provoke a little foot-shooting.
AMD likes to preempt Intel announcements when they can, so maybe they’ll have a “surprise launch” before they’re ready to supply. More bad publicity.
It’s also likely this will also press AMD’s “Cut the price” button, which I’m sure the stock market will love.
Intel can do all these things because it has the upper hand. A delay in some Penryn parts is not the same thing as delays in K10 parts because Intel has plenty of almost-as-good, still superior Kentsfields for sale, and it can afford to delay the new stuff and clear out the old stuff because AMD can’t do anything about it.
With the upper hand, Intel can jerk AMD around like a puppet on a string, but only because it has that upper hand. (To a lesser degree, this is now also true for nVidia.)
This is why it is good to be king. This is why it is not good to be under the king. This is why AMD’s sudden interest in “value” is not something they would ever do voluntarily; you lose a lot of control (and money) over your own destiny.
Nor is that good for us, outside of getting some cheapish parts for a while. You see, for us to benefit from competition between Intel and/or nVidia and AMD, AMD has to compete. If they withdraw from the competition and settle for making second-rate stuff, all the kings have to do is price a middle- to low-end part competitively, and jack up the price for anything above that. Meanwhile, the lower-priced R&D-lite stuff will get (relatively) worse and worse.
One reader nicely called this the “Die as slowly as possible and pray” strategy. That’s pretty good, but maybe calling this AMD’s Zombie Strategy puts it a bit more succinctly. Cutting R&D and fabbing renovation might well keep them alive longer, but just how do you bring the zombie out of it?