Connections in the History of Australian Computing - History of Computing: Learning from the Past - page 5

AGWAC 8 . Presenters came from London, Cambridge, Leeds and Manchester Univer-
sities, the UK National Physical Laboratory, Ferranti, EMI, Elliotts and English
Electric. Also Michigan University, the US Cape Canaveral testing ground, three
Australian Universities, CSIRO and, of course, WRE. A few of the visitors were An-
drew Booth, Stanley Gill, Tom Kilburn, and Maurice Wilkes. Locals including Trevor
Pearcey, Murray Allen, John Bennett, Brian Swire and John Ovenstone renewed
acquaintances and a great deal of information was transferred [24].
“The conference was a great success, both technically and socially” [25]
WRE hired an IBM 7090 from 1960, and in late 1962 WREDAC was scrapped [26].
6 Brian Swire, John Bennett and SILLIAC
In 1952 the University of Sydney's Physics Department got a new Head, Dr Harry
Messel, and a new budget. Messel appointed Dr John Blatt from the University of
Illinois. Dr Blatt had programmed Illinois' ILLIAC , one of a series of copies of John
von Neumann's IAS Computer , and easily made the case that Physics needed a com-
puter. Australia's only computer was on the Sydney Uni site, but it was being kept
very busy with CSIRO work. Prof Messel campaigned for funding and in 1954 Ad-
olph Basser donated his horse's Melbourne Cup winnings. Dr Blatt arranged to get
circuit details and construction samples from Illinois while Prof Messel arranged for
staff to build their computer and to program it [27].
Construction of “Sydney's ILLIAC”, or SILLIAC , was directed by Brian Swire.
Brian had worked as a radar engineer at Radiophysics then moved to the CSIRO
Aeronautical Research Laboratory, and he attended the 1951 conference. He took the
von Neumann/ ILLIAC design and reworked it for maximum reliability [28].
When they advertised for a software expert who could also teach programming, the
best applicant was an Australian working for Ferranti UK. John Bennett had experi-
ence with CSIR's radar team and David Myers before heading to the UK. He was
Maurice Wilkes first PhD student and he had helped build Cambridge University's
first computer EDSAC , which was largely based on John von Neumann's EDVAC
design. John Bennett's work for Ferranti involved reworking the instruction set of
their first computer, logic design (including NIMROD , the first games console), soft-
ware development and customer relations [29].
SILLIAC's circuitry was constructed by Sydney electronics firm Standard Tele-
phones and Cables then assembled and tested by Brian's small team from mid 1955.
John started programming courses, which included a helping of numerical analysis
concepts, and assembled the operating software based on Illinois' experience. The
first successful run in July 1956 gave Sydney a brief lead in quantum theory 9 and her-
alded a decade of intense work for the University, CSIRO and business [30].
Late in its life SILLIAC was interconnected with other University computers as an
input/output server in what we would now call a local area network. Hardware and
software were developed by Chris Wallace. Chris later spent some time with English
Electric and contributed to their KDF9 team [31].
8 The Australian Guided Weapons Analogue Computer was used to model missile behavior.
9 Specifically, the mathematical description of helium superfluidity.