Path: utzoo!utgpu!news-server.csri.toronto.edu!mailrus!uwm.edu!zaphod.mps.ohio-state.edu!usc!snorkelwacker!bu.edu!cs!art
From: art@cs.bu.edu (Al Thompson)
Newsgroups: comp.sys.super
Subject: Re: Supercomputer ROI
Message-ID: <57013@bu.edu.bu.edu>
Date: 9 May 90 20:46:32 GMT
References: <201@csinc.UUCP> <253@garth.UUCP> <202@csinc.UUCP> <13881@dime.cs.umass.edu> <301@garth.UUCP>
Sender: news@bu.edu.bu.edu
Reply-To: art@cs.UUCP (Al Thompson)
Distribution: na
Organization: /usr/lib/news/rn/organization
Lines: 57

In article <301@garth.UUCP> fouts@bozeman.ingr.com (Martin Fouts) writes:
[...]
|
|According to a book on Bell Labs that I have been reading lately this
|is another urban myth.  Acording to the book, DOD did very little to fund
|transistor work.  There were two problems with marketing transistors.
|One was reliability and the other was acceptance.  Once the
|reliability problems were solved, A major effort was needed to
|convince engineers to design with transistors, and to find products
|where they were truely useful...

Well, this is sort of right and sort of wrong.  It's true that Uncle did
not directly fund the development of the transistor.  But, it was the
needs of NASA for low weight, low power devices that drove the
development.  The gov't was willing to pay any price for these devices,
and so they purchased them when they were so expensive no private consumer
could afford them.  The result was that manufacturing techniques were
improved to the point where the unit price fell to acceptable consumer
levels.

|
|In fact, on careful investigation, one can find very little technology
|developed as a result of government research which actually achieved
|commercial acceptance.  I've been trying for enough years that I now
|assert that no US government funding has had sufficient impact on a
|consumer technology to pay for itself, or introduce a significant
|impact on the market.

Try the flat rivet technology Boeing uses on its new generation of
airliners, the Connection Machine, Sun Microsystems, most medical
research, duct tape, etc.

|
|Before you all flame me about the early history of computers, go back
|and carefully compare the relationship of the the military projects
|like eniac to the commercial work done by companies like sperry and
|ncr.  You will be surprised at how little impact removing the
|government from the computer industry would have really had.

I'm not so sure.  Just who bought most of the original computers?  Uncle
of course.  That single customer market had a big impact whether or not
they were directly involved.  Don't forget the ARPA net.  It's very common
for the gov't to fund things indirectly by being a big (sometimes only)
early customer.  The Connection Machine is a case in point.  It's early
support from DARPA (purchase of machines before design completion and
funding a large percentage of the early purchases) made it a viable
product.

|
|BTW, the big win for transistors was cheap pocket radios.  Twenty
|years later the big win for ICs was cheap radios.  US industry missed
|both of these obvious markets.

The first transistor radio was the Regency which appeared just before
Christmas 1954.  The radio was actually built by TI and marketed by
Regency.  It was a huge success.  That others picked it up and ran with it
in no way diminishes the contribution of TI.