• C#ASP.NET Developers
• Pre-Sales Solutions Architect -...
• BUSINESS SUPPORT OFFICER
• IT Systems Administrator
• Sales Manager Australia -...

Paying homage to SILLIAC

Peter Davidson, Information Age

18/10/2006 21:51:13

They're showing their age a bit, and for some their step has slowed. But they came in celebration, alert and full of reminiscence and joy of reunion, all 300 of them, to remember SILLIAC.

Sydney University, Tuesday, September 11, 2006: the 50th birthday of the first computer built in an Australian university.

They gathered first in the Stade Lecture Theatre for the Pioneers Reunion where old amities could be renewed, some of them for the first time for 50 years.

Then came the anniversary dinner in the MacLaurin Hall, a lavish affair to honour those pioneering individuals involved with SILLIAC's operation from 1956 to 1968.

The war stories were rerun, new claims made - good humour and genuine affection for each other and what they had achieved in the creation flowed as it should.

SILLIAC, the object of their devotion and zeal, was not there, long ago broken up and its bits scattered. Some remain in museums, others cosseted by private individuals who know what it was and what it took to create it - the rest gone to the computing scrapyards formed in step with the first computers.

Towering over the celebrations was the man who started it all, Emeritus Professor Harry Messel, now 85 and still very much in command.

John Butcher, the Adolph Basser Research Student in 1956 (and SILLIAC's first), was there too. Later to head the mathematics department at Auckland University, the anniversary event brought him and Messel together for the first time in half a century.

Another day-one research student there was John Bennett who joined Maurice Wilkes at Cambridge in 1947 to work on EDSAC. He joined the SILLIAC team in 1956, and with Trevor Pearcey and others worked to form the Australian Computer Society, becoming its foundation president.

The other names so closely linked with engineering Australia's leap into computer technology were there as well: Barry de Ferranti, Peter Aplin, Jenny Edwards, SILLIAC operators June Crawford and Elizabeth Johnston and a host of others.

The celebration was part of wider commemoration under the aegis of the Science Foundation for Physics at Sydney University.

Industry figure Max Burnet, who has organised and hosted some major international ICT spectaculars in his time, hailed the event: "I have to say that the events were positively brilliant.

"To get 300 attendees to a computer history event was pretty amazing; restores one's faith that people are interested in IT history. What a lift Harry Messel gives to an event!"

It was all recollection and renewal for so many people, and Sydney's daily papers for days before and after the anniversary were dotted with stories of personal connection with SILLIAC.

In its heyday, SILLIAC toiled all day, every day and more than 2000 people used it; judging by the anecdotes the event unearthed, that number was understated by orders of magnitude.

Getting history straight But as with events varnished over the years with the revisions of wishful thinking, the history of SILLIAC as it was related in media drew its share of dissent and disagreement from those with a different recollection.

The history of CSIRAC and its creation not far from SILLIAC's birthplace has prompted some to argue about the Science Foundation's description of the computer as "the first computer built in an Australian university".

Melbourne-based Peter Thorne, deeply involved with CSIRAC during his time at Melbourne University and in ICT generally, reports "a flood of e-mails, some heated, demanding that the record be set straight.

"Some of it stems from the niceties of Sydney University's language being blurred in various media reports, and perhaps the university could claim greater credit for CSIRAC than it does.

"In he end it doesn't matter; CSIRAC, SILLIAC, UTECOM and others were all significant in establishing information technology all those years ago, and all should be recognised within a national context.

"Besides, John Deane has written the history of SILLIAC from meticulous research, and I would accept it as the real story."

A 30-year veteran of CSIRO Radiophysics, Deane has spent eight years trawling the world for SILLIAC's early documentation and interviewing as many as he could of those remotely involved with its creation, operation and the data it produced.

His book, SILLIAC, vacuum tube supercomputer, is not his first, having already produced a detailed history of CSIRAC, inspired by the discovery that the division he was (then) working in had been that pioneering machine's birthplace.

Other works comprise a wide range of monographs including a major work on the Manchester Mark 1. He is also the secretary of the Australian Computer Museum Society and an extraordinarily generous source of information and material.

A scientist first and author second, with typical self-effacement he begins his book with an apology:

"Exception has been taken to referring to the family of computers which includes SILLIAC as a supercomputer. While this is quite understandable for machines which approach the speed of a modern pocket calculator, I believe they represent the supercomputer architecture of the first computer generation.

"This apology is for readers who would prefer to reserve the term for Britain's Ferranti ATLAS, the USA's CDC6600, Russia's BESM-6 or other worthies."

A labour of love, like most of Deane's computer history activities, the book was published by the Science Foundation with support from Alcatel which absorbed Standard Telephones and Cable (STC), SILLIAC's builder.

Launched at the anniversary dinner, it was hailed by Harry Messel who said "at long last the record will be straight".

Max Burnet, also significantly involved with the museum, says that Deane knows as much about the technology as anybody: "He is a specialist in instruction sets, and during lengthy interviews with John Bennett particularly, pointed out that there were flaws in those for SILLIAC.

"To his credit, he agreed but answered that because they worked they were left alone."

His book is comprehensive and reflects his technical prowess in the detail he offers in a number of appendices, including one which lists nearly 100 "SILLIAC folk" who worked in, on or with the machine, their roles and period of service.

There are photographs, diagrams, schematics and tables - and complete instructions for its programming and maintenance.

But in and around all that it tells the story of how it all began:

When SILLIAC was fired up for the first time on July 4, 1956, it had already amassed a wealth of pioneering history.

By its 50th anniversary, the volume of lore and legend surrounding the university's engineering triumph had become epic.

As well it might; the story of how an Australian version of the University of Illinois Automatic Computer (ILLIAC) was chosen, built, funded and used is significant in Australia's history, and is a milestone in the development of computing worldwide.

SILLIAC's gestation and birth began with a 29-year-old senior lecturer in theoretical physics being importuned in 1952 at Sydney Airport by staff from Sydney's University's Department of Physics.

They had tried unsuccessfully to recruit a head of department and offered the young academic a job. His name was Harry Messel, and he was on his way to an overseas lecture tour.

He agreed to take the job with an ambit claim of conditions - which to his surprise the university accepted and Dr Messel started work.

Among his requests was a trebling of permanent staff to 21 and funding for cosmic ray research as an entry into nuclear physics.

The first among the group of eminent physicists recruited by Messel was Dr John Blatt of the University of Illinois. He knew that a computer was an essential tool for theoretical physics, and better still, also knew a lot about one called ILLIAC.

At that time, the only computer in the southern hemisphere was CSIRAC, designed by Trevor Pearcey and built not far from Sydney Uni. It operated until 1964 and remains intact, if not working, in Museum Victoria (see last issue of Information Age).

Like its (few) contemporaries, CSIRAC was a serial machine and unable to offer the speed that Blatt required, and in any case, its CSIRO work came first.

Blatt approached ILLIAC's engineer Ralph Meagher, also well-known to Messel, and secured the blueprints for ILLIAC.

How Messel secured private sector funding to get things going is the stuff of a separate book: Knowing someone that knew someone (etc), Messel met jeweller Adolph Basser, who, enthused by his vision in nuclear physics, joined the Science Foundation.

Basser also donated 50,000 pounds to Messel's projects in 1954 (about $2m today), and lore tells that it was winnings from his horse Delta's victory in the previous year's Melbourne Cup.

Deane's real story is far more complex, but Basser (later Sir Adolph) gave his name to the computing laboratory and subsequently a further 50,000.

Funded, the project needed engineers and Messel recruited Brian Swire, who like John Bennett, Maurice Wilkes and other eminent engineers of their day, had wartime experience in radar.

He was working at the Aeronautical Research Laboratories in Melbourne using CSIRAC, and became senior engineer.

(Later he was to die by SILLIAC's hand, electrocuted during an otherwise routine engineering operation.)

His assistant, John Algie, joined the burgeoning group from CSIRO and stayed until the end of 1954 to be replaced by Barry de Ferranti, who played a leading role in the anniversary celebrations and particularly the futurists' symposium held in conjunction with it.

De Ferranti is vice president and Honorary Life Governor, Foundation for Information Technology at Sydney University.

An atypical benchmark In what would have been a welcome standard for the ICT industry today, SIILIAC was completed seven months earlier than scheduled at a total cost of 75,000.

It ran its "leapfrog" self-diagnostic test successfully on July 4, 1965, and was declared fit for work by Swire, de Ferranti and engineer/programmer Peter Aplin.

Until finally decommissioned in 1968, SILLIAC undertook a huge range of tasks, largely through the evangelistic work of John Bennett who proclaimed its capabilities and benefits to business, proving its value to both public and private sectors.

Bennett had been given Chair of Physics (Electronic Computing) in 1961 and strove to encourage students into an information technology career and generally to nurture the principles and progress of the discipline.

He still does.

SILLIAC on DVD The history of SILLIAC's creation is told on a DVD produced for the anniversary by the Science Foundation. Narrated by scientist/broadcaster Adam Spencer, the highly polished DVD has interviews with Messel, Aplin, Peter Poole and many others.

Like the book, it is an essential item in any library of Australian history. Written, produced and directed by Alison Muir of the Science Foundation Physics, both can be requested by e-mail from a.muir@physics.usyd.edu.au