Dave B There, doesn't it feel better when you can really contribute to the thread and say what you really mean? <G>
A lot of statements on this thread, (long and short), are based too much on emotion and too little on reality.
The ramp of P4 is a lot more complicated than building 1 million P4's. Demand play's an important role, and with a new product there's a lot of guess work by the box makers. I think, (and I believe Intel has also stated), the start-up demand will be primarily in the Work Station space, and 100% of that market is ~500 KU's/Qtr. I don't think anyone expects to capture 100% share. (Ironically, with RDRAM rapidly growing in that segment, P4 may not create a significant new market for RDRAM in WS's).
In addition to demand forecast for P4 systems, there are many issues beyond fabricating the P4.
You need mother boards, which I expect will be 100% Intel at the major box makers.
Will Intel build a million mother boards to support the launch?
You need the i850 chip set. You need new coolers, which from what I've seen, are quite exotic. You need QA testing at many levels. On start-up what level of sample testing is used for the various components? (It could be as high as 100%). Is Intel shipping tested platforms? What level of testing is required at the platform level?
I doubt if RDRAM availability has any impact on the ramp.
It appears the Intel road map shows some significant changes/additions to the P4 and chip set in 2H'01. What impact will that have on the ramp?
New product launches are much more complicated than building CPU's.
JMHO's |
