<!doctype html public "-//w3c//dtd html 4.0 transitional//en">
<html>
Rakesh Bhandari wrote:
<blockquote TYPE=CITE>Enrique, I noted at least the following pts (response
based on a 3 minute
<br>conversation with a senior microprocessor engineer whom I shall not
name
<br>since I may not have understood him properly--rushed conversation):
<p>1. high end users already served; move to cheaper $500 computers which
<br>won't allow for big mark-ups henceforth--though of course there may
be a
<br>new bout of superprofits for software vendors as pc's proliferate.
<p>2. Intel now facing competition, though from mostly American companies
(via
<br>excepted)/ Profit margins will be no more exceptional in the microchip
<br>business than the potato chip business--"the MPU business won't be
worth
<br>competing for"! _____________________________
<p>True enough. Intel in litigation with Via now over dumping. Shows that
<br>catch-up has been a truly important development in the last two-three
<br>years. I'll need to look into this foundary/fabless design company
div of
<br>labor.
<p>Points 1 and 2 seem however to ignore the next round of "high value
added"
<br>product: system on a chip integration by which total chip count will
be
<br>reduced. Should be a burgeoning market in telecom (hardware and software)
<br>and appliance intelligence--big future market not just internet via
pc.
<br> </blockquote>
Perhaps. He seems to be conceding that Intel's current model (make fat
profits by getting business and consumers to replace their computers with
high-end, high-margin CPUs every two or three years) is gone. Systems on
a chip may or may not be a replacement for that. I am skeptical - at best
it means Intel will capture an increasing share of the decreasing PC cost.
Appliance intelligence strikes me as a very iffy proposition: a solution
in search of a problem.
<p>The main issue remains: Intel's source of monopolistic rents, well-heeled
customers who demand top-of-the-line performance that only Intel can deliver,
is gone. It remains to be seen whether they can replace it.
<br>
<blockquote TYPE=CITE>
<p>3. microprocessor industry can't easily be cartelised.
<p>Been a bout of joint ventures though including across national boundaries.
<br> </blockquote>
Yes, but as long as there are two or more suppliers of essentially identical
products at near-zero marginal costs, I don't see how cartelization can
work very well.
<br>
<blockquote TYPE=CITE>
<br>4. a new broader question of my own: do any technology monopolies enjoy
<br>superprofits?</blockquote>
Can't answer that.
<blockquote TYPE=CITE>
<p>yrs, rakesh</blockquote>
<pre>--
Enrique Diaz-Alvarez Office # (607) 255 5034
Electrical Engineering Home # (607) 272 4808
112 Phillips Hall Fax # (607) 255 4565
Cornell University <A HREF="mailto:enrique@ee.cornell.edu">mailto:enrique@ee.cornell.edu</A>
Ithaca, NY 14853 <A HREF="http://peta.ee.cornell.edu/~enrique">http://peta.ee.cornell.edu/~enrique</A></pre>
</html>