Path: utzoo!utgpu!jarvis.csri.toronto.edu!mailrus!cs.utexas.edu!uunet!zds-ux!gerry
From: gerry@zds-ux.UUCP (Gerry Gleason)
Newsgroups: comp.arch
Subject: Re: Here are five counterexamples
Message-ID: <104@zds-ux.UUCP>
Date: 15 Jan 90 18:56:30 GMT
References: <143@daedalus.nsc.com> <34315@mips.mips.COM> <161@csinc.UUCP> <7375@drilex.UUCP>
Reply-To: gerry@zds-ux.UUCP (Gerry Gleason)
Organization: Zenith Data Systems
Lines: 56

In article <7375@drilex.UUCP> dricejb@drilex.UUCP (Craig Jackson drilex1) writes:
>In article <161@csinc.UUCP> rpeglar@csinc.UUCP (Rob Peglar x615) writes:
>>3)  The way US Congress acts, US business law will not change soon.
	[ with respect to protection from foriegn buyouts ]

>The somewhat scary thing is that you seem to wish that it does.  The sure
>way to make American industry weak is to protect it from the outside.
>In this case, the protection would be largely from capital investment, which
>is very important in the semiconductor business.

>This strays quite a ways from comp.arch towards sci.econ stuff.  (Both
>in content and tone.)  I'll try not to say more.

I agree that this is straying, so I'll make a quick comment on this, and
then head off in a direction that should be more relevant.  Whether foriegn
ownership is good or bad depends on how the foreign run companies behave.
If they gut the company after buying it, keeping only things like sales
and support networks and final assembly plants, and move all the design
and planning tasks back to their home country, or they bought it just to
eliminate the competition, it is a bad thing.  Watching after your own
interests is not protectionism.

Yesterday, Adam Smith's Money world was on the topic of Supercomputers,
and whether Japan's targeting of this market represents a serious threat
to the technological leadership of the US.  All the computer industry
people they had on thought it is a very serious threat, and that it is
possible that if this market is lost, the rest of the computer technology
markets are likely to go too.  My own oppinion is that the Japanese are
still fighting WWII in the economic arena, not very far fetch when you
consider that in both cases it is the leaders of the powerful corporations
that are/were pressing the issue.

But I continue to digress, what I want to as this forum is whether
supercomputers are that important, and whether the US is really in
danger from the Japanese strategy.  I suspect that the danger is not
in the super-computer area, but instead in basic technologies
(fabrication, lithography, etc.).  It has already been suggested that
the super-computer market is in trouble from "killer micros," and that
it will be more and more difficult for the super-computer people to
find enough perfomance to justify the cost differentials.

On the show they suggested that the closing of ETA was a very short
sighted action, that they didn't even give their product a chance
to be completely developed.  Maybe their isn't room for as many
players as have entered this market, i.e. killer micros have and/or
will reduce the size of this market even if they can't eliminate it.
The most basic question here is really whether the super-computer
market will continue to grow enough to support all the companies in
it.

My thinking is that current super-computer architectures will become
available on just a few chips (or one) within the decade, and the
leading edge for extremely compute intensive jobs will move to machines
with radically new architectures (TMI's connection machines maybe?).

Gerry Gleason