Relay-Version: version B 2.10 5/3/83; site utzoo.UUCP
Path: utzoo!mnetor!uunet!seismo!sundc!pitstop!sun!amdcad!ames!ucbcad!ucbvax!jade!eris!mwm
From: mwm@eris.BERKELEY.EDU (Mike (My watch has windows) Meyer)
Newsgroups: comp.arch
Subject: Cray Research walkout (Was: brash micros versus the Big Iron: not yet)
Message-ID: <5270@jade.BERKELEY.EDU>
Date: Wed, 30-Sep-87 00:21:10 EDT
Article-I.D.: jade.5270
Posted: Wed Sep 30 00:21:10 1987
Date-Received: Sun, 4-Oct-87 19:38:21 EDT
References: <622@winchester.UUCP> <2083@sfsup.UUCP> <954@edge.UUCP> <7470@steinmetz.steinmetz.UUCP>
Sender: usenet@jade.BERKELEY.EDU
Reply-To: mwm@eris.BERKELEY.EDU (Mike (My watch has windows) Meyer)
Organization: Missionaria Phonibalonica
Lines: 60

In article <7470@steinmetz.steinmetz.UUCP> davidsen@crdos1.UUCP (bill davidsen) writes:
<In article <954@edge.UUCP> doug@edge.UUCP (Doug Pardee) writes:
<| ...
<|I wonder if Seymour Cray knows that he's barking up the wrong tree by
<|designing multiple-chip CPUs?  Somebody ought to tell him  :-)
<
<Actually I think someone did. I read that his chief designer for
<advanced products was "allowed to resign" when the single chip processor
<was advocated over the multichip GaS version. Said designer then left to
<form his own company.

[To avoid confusion: when I say "Cray", I mean the person; I'll say
"CRI" when I mean the company.]

What you heard and what I read (in the Wall Street Journal) differ
slightly. The story as they printed it was that Steve Chin (the
designer in question) resigned when the MP project was cancelled. The
difference of opinion appeared to be that Chin wanted to start from
sand(*) on the MP parts, whereas CRI wanted him to start with
pre-existing silicon.

This could be a case of "pre-existing" meaning a multi-chip cpu
similar to what CRI has already, and Chin wanting to do single-chip
GaS cpu, but I didn't (and still don't) read it that way.

Chin isn't "Cray's chief designer." He was CRI's *other* chief
designer, the first being Cray. The two had their own teams for
developing machines. Cray is working on the Cray 3; chin was working
on the MP. This was to be the Cray for the next decade, with "many"
(dozens?) of processors. Before the MP, Chin had designed the X/MP
(the current fastest Cray's are X/MPs; Apple's got one, SDSCC's got
one, UCB's got an old one) and the Y/MP. The Y/MP project is nearly
complete, and Chins leaving isn't supposed to affect the release date
for it. 

The parting between CRI and Chin was "amicable." Chin is being allowed
to use the technology he developed for the MP project outside of CRI,
but *not* for any of their competition. He has to build a company from
the ground up. If I recall correctly, CRI is even providing some
funding for this (thus allowing them to eat their cake, and still have
a part of the pie :-). Chin is expected to hire most of the MP design
team, while CRI is keeping some of them on their payroll until
December - even though they aren't doing anything.

As for multi-chip vs. single-chip cpus, the painful part of a
Cray-class cpu appears to be cooling the things. The cooling system
for every Cray I've ever seen has been at least as large as the cpu.
That the Cray 2 is a liquid cooled machine (yes, chips and boards are
immersed) ought to indicate how serious the problem is.  I suspect
going to multi-chip cpus lets you get the same speed with less heat.

	<mike

(*) Sand is the second-lowest level I know of to start with. To get
below that, you start with physicists :-).
--
ICUROK2C, ICUROK2.				Mike Meyer
ICUROK2C, ICWR2.				mwm@berkeley.edu
URAQT, I WANT U2.				ucbvax!mwm
OO2EZ, I WANT U2.				mwm@ucbjade.BITNET