Path: utzoo!utgpu!news-server.csri.toronto.edu!cs.utexas.edu!usc!ucsd!ames!amelia!wilbur.nas.nasa.gov!eugene
From: eugene@wilbur.nas.nasa.gov (Eugene N. Miya)
Newsgroups: comp.sys.super
Subject: Re: The ETA Saga
Message-ID: <5971@amelia.nas.nasa.gov>
Date: 4 May 90 17:00:19 GMT
References: <198@csinc.UUCP< <181@nachos.SSESCO.com> <1990Apr21.022632.12639@murdoch.acc.Virginia.EDU> <226@garth.UUCP> <5797@amelia.nas.nasa.gov> <232@garth.UUCP> <10199@batcomputer.tn.cornell.edu>
Sender: news@amelia.nas.nasa.gov
Reply-To: eugene@wilbur.nas.nasa.gov (Eugene N. Miya)
Organization: NASA Ames Research Center, Moffett Field, CA
Lines: 36

In article <10199@batcomputer.tn.cornell.edu>
kahn@tcgould.tn.cornell.edu (Shahin Kahn) writes:
>SSI is not out of business.

>I know what you mean, but E&S is also not out of business.

>You really should be careful who you declare out of business on the net
>with such a wide audience.

THANK YOU for noting the editing mistake.  It's okay, happens all the
time ;).  Not the first time one's been misquoted and Marty can handle
himself.  He wasn't declaring them out of business.  In fact, ames gives
SSI a mail gateway to the Internet.  The explanation for E&S was
actually probably necessary.  Only the people in graphics realize they
sell $10M flight simulator computer image generators.  We have one of
their CT-5A units in another part of Ames.  These kind of are supers.

>The best supercomputer is yet to be built. (perhaps a
>perpetually true statement.)

>And in my opinion, the market *is* large enough to accomodate several
>companies.  Certainly the 'global market' is.

I would tend to disagree with you.  I don't think the market is large,
there's more political hot air than necessary.  The problem as I see it
is that we need a next stage of integration.  I think that as the
semiconductor technology begins to top out it's going to become
important to produce modules (which include software) that non-CS types
can put together and make their own application specific "supercomputers."
We are seeing more and more functionality in a given chip, it's when
it makes the software leap (things like routing in this sense are dumb),
such a stage will come to pass.  The two companies I've seen thinking
about this are DEC and independently Rockwell.  There will be others.

--e. nobuo miya, NASA Ames Research Center, eugene@orville.nas.nasa.gov
  {uunet,mailrus,other gateways}!ames!eugene