Is Intel’s adaptation of x86-64 technology (and no matter how Intel tries to spin it, that’s what it is, outside of items like SSE3 and hyperthreading which Intel implemented long before this move) a victory for AMD?
Depends on how you define “victory.”
Is it a defeat? Depends on how you defined “victory.”
Actually, Intel’s decision aborts the possibility of both victory AND defeat for AMD.
Do I speak in riddles? Let me explain.
If you viewed x86-64 as a means by which AMD was going to at least beat up if not conquer Intel, CT technology certainly ends that delusion.
It was a delusion from the very beginning because the second x86-64 looked like an actual or even potential threat to Intel, Intel could always say, “Me, too,” and if you look at Intel history, when push comes to shove, they will always eventually say “Me, too.”
If that’s how you viewed it, Intel’s announcement not only is a defeat, it ends the “war of conquest.”
It is ludicrous to call that a victory. That is like buying yourself a gun to kill somebody, and claiming a victory when the other guy gets himself a gun, too. That’s ludicrous, you wanted him dead, not armed and presumed dangerous.
On the other hand, if you viewed this in “Kill or be killed” terms, while Intel’s announcement will prevent AMD from killing with x86-64, it also prevents AMD from getting killed by it.
Intel’s announcement ought to ensure that x86-64 doesn’t turn out to be another 3DNow. It prevents AMD from staking all on x86-64 as the reason to buy a Hammer, only to have it flop.
Intel’s decision essentially takes x86-64 out of the ballgame, war, whatever you’d like to call it.
But They Got Scared Into It!
Sorry, guys, that’s an argument for losers. When you’re out to do something to somebody, and that somebody takes steps to prevent you from doing that something, that’s not a win. A moral victory is not much of a consolation prize in an amoral battlefield.
But This Kills Itanium!
No, it doesn’t, not at all. It may end up changing Itanium, or perhaps delay any Intel plans to bring it to the desktop (which wasn’t going to happen for at least another three years, anyway).
It took forever, but Itanium is starting to sell. Intel ended up selling about 100K in 2003 (with almost all those sales were in the last quarter). That’s roughly equivalent to Opteron sales.
Itanium can hang out in the higher-end server market for a few years and keep developing. There’s no reason to think Intel will pull the plug on them.
Indeed, if Intel decides to make CT universal in its processors (and more on this later), it could make x86-64 worthless to AMD (in the sense of being able to charge a premium for it). For instance, if Intel comes out with Prescott Celerons with x86-64, how can AMD possibly take out x86-64 from its “Paris” line of chips this fall? See what I mean?
It’s by no means certain Intel is going to do that. It’s possible CT technology will be restricted to the high-end just like Extreme Editions.
Either way, Intel’s move will make it as least likely, at most inevitable, that we’ll see much more affordable Hammers, at least for the 90nm generation, and maybe sooner.
Are You Going To Call This An Intel Victory, Now?
On the whole, I would have to give the edge in this skirmish to Intel, simply because it’s a preemptive move which will prevent many Intel owners from looking elsewhere and gives Intel the opportunity to play pricing games with AMD again.
However, this is just one small facet in the whole war, and the only reason why Intel shot AMD in the arm is to keep the world from noticing that they napalmed their own leg off with PressHot.
AMD may have to give up its delusions of conquest, but especially given PressHot’s heat problems, Hammer still will be a very, very competitive chip, for other reasons. Hammer is not a one-horse cart.
All AMD really lost by the announcement was delusions of grandeur. It still can hold its own and then some.
Saying this is a big defeat for AMD is just as delusional as saying the opposite.