Article 108 of comp.sys.cdc:
Path: matra.meer.net!news1.best.com!feed1.news.erols.com!news-peer.sprintlink.net!news.sprintlink.net!Sprint!news.mathworks.com!blanket.mitre.org!news.mitre.org!news
From: ?@mitre.org
Newsgroups: comp.sys.cdc
Subject: Re: Reminiscing About the Cyber 6600
Date: Tue, 20 May 1997 13:30:37 -0400
Organization: Mitre Corp
Lines: 23
Distribution: inet
Message-ID: <3381DFB5.273D@mitre.org>
Reply-To: ?@mitre.org
NNTP-Posting-Host: buick.mitre.org
Mime-Version: 1.0
Content-Type: text/plain; charset=us-ascii
Content-Transfer-Encoding: 7bit
X-Mailer: Mozilla 3.0 (Macintosh; I; PPC)
Xref: matra.meer.net comp.sys.cdc:108

I haven't watched this newsgroup in a long time, but saw the post having
to do with the Cyber 835 at Concordia and it struck a reminiscing cord.

While working for CDC and Pratt & Whitney the summer of '79, I heard
about the "virtual machine running NOS."  Later, in '81, I was the site
engineer for Raytheon in Bedford, MA, which received 835 S/N 7 (or so)
as part of the "Early Bird" program.  Pretty amazing to see all that
central memory.  Remember the BLITZ command?  We set up 1/2 the memory
as extended memory, started a bunch of programs, and typed BLITZ at the
console.  Wow!  Those programs were swapped to the psuedo-ECS in what I
remember was less than a second.

And remember the 180/825?  The sales pitch went "why is an 825 like a
jalepeno pepper?  They're both small and powerful."  The rumors around
Benchmark Labs at CDC HQ were that someone coded up some vector
instructions that fit within the unused micro control store for the
825.  The floating point rate in that mode was supposedly the same as
old RUN-FORTRAN code.  (got us back to Fortran!)  Maybe not quite as
fast as using the CYBERStack library routines, though.

Larry Fisher
The MITRE Corporation
Ex-CDC, EX-Alliant Computer Systems alumnus