Relay-Version: version B 2.10 5/3/83; site utzoo.UUCP
Posting-Version: version B 2.10 beta 3/9/83; site mot.UUCP
Path: utzoo!watmath!clyde!burl!ulysses!allegra!mit-eddie!godot!harvard!seismo!ut-sally!oakhill!mot!fred
From: fred@mot.UUCP (Fred Christiansen)
Newsgroups: net.followup
Subject: Caltech's Cosmic Cube Summary
Message-ID: <125@mot.UUCP>
Date: Wed, 20-Mar-85 13:09:22 EST
Article-I.D.: mot.125
Posted: Wed Mar 20 13:09:22 1985
Date-Received: Fri, 22-Mar-85 02:36:20 EST
Organization: Motorola Microsystems, Phoenix AZ
Lines: 67

[]
There was suprising interest in the results of my inquiry, so here's the
summary, albeit 2 months after original posting:

Subject: Re: Caltech's Cosmic Cube
-------------------------------------------------------------------------------
Date: Sun, 6 Jan 85 19:43:02 pst
From: oakhill!ut-sally!seismo!hao!hplabs!christy

You can contact Geoffrey Fox @ CalTech or Chuck Seitz
                citvax!gcf                Chuck@cit20.arpa
- Peter Christy
-------------------------------------------------------------------------------
From: oakhill!ut-sally!seismo!mcnc!ncsu!rjm
Date: Mon, 7 Jan 85 09:08:06 est

	Caltech's dean of computing (a British physics professor)
described the work at the banquet at a conference in Pasadena last
October (CAPAIDM - Computer Architectures for Pattern Analysis and
Image Database Management - by IEEE Computer Society).  It is a cube
(8x8) connection for 64 processors.  Charles Seitz from Caltech's
computer department is the head of the project.  Howard Seigel from
Purdue didn't think too much of the architecture, but the government
seems to be funding it heavily and they were getting 2-3 speedup.
I have much more information somewhere probably, so if you want more
send me mail and I'll try to find it.  They were planning to go to 1024
nodes soon.  The physics guy was somewhat of a jerk and didn't know much
about computer architectures, just cosmic physics problems (star's
gravity effects, etc.) and chemistry problems.

				robert moorhead
				north carolina state university
-------------------------------------------------------------------------------
Date: Tue, 8 Jan 85 15:00:18 pst
From: sftig!ihnp4!rlgvax!ccicpg!ifpp

A friend of a friend...( you know how it goes) is working on this project
at Cal Tech High Energy Physics. His first name is Charles, and I'm sure he
will be able to supply info/refs. USENET address: from trwrb!cithep!nncp5!cfd.

P.S. I believe that generation two of this project will use 68000's.

                                Pete Carpenter        (714) 770-5367
                                {ihnp4!rlgvax or pertec}!ccicpg!pete
-------------------------------------------------------------------------------
Date: Wed, 9 Jan 85 10:00:32 pst
From: sftig!allegra!decwrl!spar!kissell

Check out "Concurrent Processing for Scientific Calculations" by Geoffery
Fox, which was presented at the 1984 Spring Compcon in San Francisco and
which appears in the digest of papers for that conference, which is available
from the IEEE Computer Society Press.  Pretty short, but it's the only thing
I've seen in print.  The presentation was more comprehensive.

Kevin D. Kissell
Fairchild Advanced Processor Development
uucp: {ihnp4 decvax}!decwrl!\
                             >spar!kissell
    {ucbvax sdcrdcf}!hplabs!/
-------------------------------------------------------------------------------
An unknown sender from Los Alamos mailed a fuzzy copy of an article re this
in May 1984 Physics Today.   --- Fred Christiansen
-------------------------------------------------------------------------------
-- 
<< Generic disclaimer >>
Fred Christiansen, Motorola Microsystems         {ihnp4,allegra}!sftig!mot!fred
{ihnp4,seismo}!ut-sally!oakhill!mot!fred         {ihnp4,amdahl}!drivax!mot!fred