Intel will have nothing but big, hot, incompatible Itanium and obsolete 32-bit P4/banias processors for at least a year after AMD is offering a consistent, compatible, 64-bit platform for everything from notebooks to enterprise SMP servers.
Going to have to disagree with you there. While the K8 is 32-bit compatible, to unlock it's 64-bitiness requires a new OS, recompiled programs, etc. A lot of stuff is already ported to Itanium. As to the features of Prescott, your guess is as good as mine at this point, so I wouldn't count your chickens before they're hatched.
You might want to consider that Intel's "a little more of the same-old, same-old", strategy will probably not let them maintain $150 ASPs and 80%+ market share - and all that that implies.
AMD's been making that same threat for years. They meet with success for a while, then they have a patch of spotty execution and drop back down to a 10-12% provider and Intel's ASP goes up. *shrug* It's a cyclic business and the same story has repeated itself several times. AMD from 1998-2002 demonstrated that cycle perfectly. It will be interesting to see if AMD reverses the trend in 2003.
Intel dodged a bullet when AMD had 3 month delay in the Hammers' release - but another one is coming.
Don't get me wrong, I think the K8 series is going to be a very competitive chip, but given that it won't appear in any volume until Prescott and Madison are BOTH out, that's going to limit its impact. If it had been released to compete purely with the Northwood and Itanium 2 (McKinley), yeah, it would have really had AMD sitting pretty, but it wasn't. You have to compare it with the chips it will be going head to head with, and in that comparison, the hammer doesn't look nearly as strong. |
