Relay-Version: version B 2.10 5/3/83; site utzoo.UUCP
Path: utzoo!utgpu!water!watmath!clyde!rutgers!husc6!necntc!ames!lll-tis!ohlone!nelson
From: nelson@ohlone.UUCP
Newsgroups: comp.arch
Subject: Re: Steve Chen resigning (was Re: Cray Research walkout)
Message-ID: <365@ohlone.UUCP>
Date: Thu, 1-Oct-87 22:31:29 EDT
Article-I.D.: ohlone.365
Posted: Thu Oct  1 22:31:29 1987
Date-Received: Sat, 3-Oct-87 09:02:31 EDT
References: <622@winchester.UUCP> <2083@sfsup.UUCP> <954@edge.UUCP> <2956@ames.arpa>
Organization: Cray Research Inc., Livermore, CA
Lines: 35
Summary: last week's w.s.journal article is very poor

In article <2956@ames.arpa>, fouts@orville.nas.nasa.gov (Marty Fouts) writes:
> The Wall Street Journal article leads one to believe that Steve Chen
> wanted to start with "sand" rather than chips, but this isn't even
> that uncommon at CRI.  Seymour Cray himself started with "sand" on at
> least three machines: The Cray 1, the Cray 2, and the Cray 3, which is
> said to be a GaAs machine.

The brief article in last Friday's Journal is really very poor.  Don't
believe anything it has to say.  On the other hand, the articles that
came out just after the announcement 4 weeks ago were rather reasonable
I thought.

As has been noted many times, Seymour Cray is not a big risk taker or a
'leading edge' technologist.  He tends to build machines with proven
components bought from third parties.  He *IS* a real genius at packaging,
cooling, and design.  Steve Chen was much more agressive about developing
technology in house, rather than buying it off the shelf.  This has higher
potential payoffs, but of course is riskier and more expensive (e.g.
Trilogy).  This comparison between Seymour and Steve needs to be taken with
a grain of salt, since the Cray-3 *is* being built out of GaAs (still pretty
new/leading-edge), but it should give you the idea.

Anyway, the board of directors thought Steve's MP project was too expensive,
too risky, and taking too long.  Steve was for the most part unwilling to
compromise on the design/budget/timeline for the MP, so the project was
cancelled.

If anyone has a specific question (that has a non-company private answer),
I'll be happy to tell you what I can.  However, since I am a part of the
cancelled project, my machine may go away fairly soon.  Also, lll-lcc
has been flaky for awhile; try ihnp4 if lll-lcc bounces you.

-----------------------
Bron Nelson     {ihnp4, lll-lcc!lll-tis}!ohlone!nelson
Not the opinions of Cray Research