You bring up good points Paul. However, I believe that your "stated opinions" of how intc will fare may be somewhat optimistic. First off, most people were counting on intc to shift entirely to PII by now (w/ the ??? high margins). I recall that when I started looking into x86 makers (about one year ago) that estimates were for intc to be making $4.80 for '98. Do you recall, how many times since then the numbers have been revised?
If you heard the last cc, I think you would agree that although intc speculated about possible improvement, they also left leeway to fall back on if things didn't pan out. Since you know I am a NSM fan, I will bring them up also. Nothing has been seen yet from their new fab. However, this output will be on top of what is coming from IBM. Their die sizes seem competitive and since their emphasis on a slot?/socket7/super? solution is known to be only as a partner-to-be (publically ??) they have the insulation of attracting vendors specifically wanting an integrated solution (not to mention that nice price benefit!). Since the 2nd half is (traditionally) expected to be the better of the two, we should expect to see some intense competition between all three x86'ers with IDT & maybe other jumping in to boot. Yes, Xeons & fast PII's will help, but no one seems to be showing evidence that supports this point. In addition, I noticed that fairly soon baseline systems will be 300 class. Since consumers won't be seeing 2XX labeled machines in the stores, starting with 300's & going to 450's, one would think that the price spread should not be too large. Since 300's go for less than $1000, it would be reasonable to see it at $700 or less by Christmas??? How much would 450's have to sell for for them not to "collect dust" on the shelves? |
