A 1998 scenario
Lets assume that 100 million PC's/servers do ship in 1998.
Conservatively assume that AMD, Cyrix/NSM, IBM, Motorola, IDT, etc. get only 20% of these. 80 million systems for Intel.
The question becomes, how many of these systems are PII based?
This will be largely a function of what the market/manufacturers demand.
Socket 7 doesn't appear to be going away. (Note that a number of suppliers have already announced new socket 7 chip-sets with AGP). P55C like systems are hot sellers, and the learning curve is extremely steep. IMO companies like Compaq and IBM want the cloners to survive. Look what they've already accomplished for their businesses. AMD may have a lot of profit pressure, (which translates directly to Intels margins), but they aren't going away. If Intel must continue to supply P55C, what is the percentage of their business, and what impact does this have on overall margins? How far is Intel willing to cut PII in order to convert sockets to Slot 1?
As an investor, these are the questions in which I'm most interested in the answers to.
If Intel is going to generate $20 Billion in CPU revenues on 80 million chips next year they need an ASP of $250. This will be very product mix dependent. I'd be interested in objective analysys. |
