Ted, 1 and 2 go together. Intel has never promised big volumes of P4, and all indications are that it's not supposed to go mainstream till the .13um process is available. There's lots of bits and pieces that fit in with this, the ones that come to mind, again, are RDRAM, big die, Albert Yu's "100s of thousands", $75 northbridge and dual channel RDRAM= expensive mobo, and the 50 new pins coming up in the next version, leaving the initial version with no upgrade path. From the figures in idc.com, Intel can no doubt produce enough to sell into the so-called workstation market, although even there they may be a bit hobbled by lack of MP support.
As for possible early DDR chipsets, it doesn't really make sense for anybody to make a new chipset when the P4 pinout is supposed to change for the real high volume part.
P4 is going to be a problem for AMD regardless, Intel may have execution problems but they certainly know how to do a PR blitz. There's not much we can do about that, but there are plenty of indications that P4 is not going to kill AMD in the short term. AMD's main problem right now seems to be Intel spinning the general PC slowdown line as the cause of its problems, I hope AMD's quarterly counters that line.
Cheers, Dan. |
