🚀 go-pugleaf

ian@syacus.acus.oz.au (Ian Joyner) writes:
>(The B5000 also had one of the first commercial implementations of
>virtual memory, 10 years before IBM invented it!).

Well I was naturally picked up for being over brief here. I was not meaning
to imply that Burroughs had invented virtual memory, just that they had
a commercially available product 10 years before IBM's first virtual
machine.

In fact the pioneers were at Manchester University who built the Atlas
computer.  I have a description of how this technology was directly
transferred from the Atlas to the B5000.  This is an excerpt from August
1977 Creative Computing, a letter to the editor from Bill Lonergan who
was then a Principal at Xerox, but who was Manager of the Burroughs
Product Planning Group Jan 1959 to April 1961.  He was writing to set
the record straight on an article by David Bulman on stack computers:

"Virual memory: In May 1960 UCLA conducted a two week seminar entitled,
"Using and Exploiting Giant Computers." The program covered the IBM
STRETCH computer, the Univac LARC, the Ferranti ATLAS 1 and Orion
computers, the Bendix g-20, and a few other machines. The list of
attendess shows that 14 people attended from IBM and seven from Univac.
We sent Paul King and two others from Burroughs. Paul and I have often
mused that the 14 people from IBM were so wrapped up in STRETCH that
they failed to grasp the significance of what the late Stan Gill was
saying about the virtual memory organizations of the ATLAS 1. Paul
King did understand its significance and returned to Pasadena greatly
excited about the concept. After a relatively brief period of review
and discussion about how best to incorporate it, a segmented virtual
memory was defined into the B5000 system (its project name in Product 
Planning at the time was the 4000 system). The credit for the first
use of a virtual memory in a U.S. machine clearly lies with Paul King
not Bob Barton.

"(It is worth noting at this point that the conceptual notion of a
virtual machine had by this time already been a topic of much discussion
around the Burroughs Pasadena facility. I believe the notion originated
with Ted Glaser earlier in the 1950s. Ted was in the Pasadena engineering
group from 1956 to mid-1959.)

"Several other B5000 design features can be traced to the May 1960 UCLA
seminar. The idea of separate, modular input-output controllers can be
traced to the LARC and the single number form can be traced to the G-20."

-- 
Ian Joyner  ACUS (Australian Centre for Unisys Software)    ian@syacus.acus.oz
"Where is the man with all the great directions?...You can't imagine it,
 how hard it is to grow, Can you imagine the order of the universe?" ABWH
Disclaimer:Opinions and comments are personal.