Conference on Automatic Computing Machines (first Australian computer conference). Pearcey, Beard.

Conference on Automatic Computing Machines (first Australian computer conference).

Publisher Information: Melbourne: 1952. [Pearcey and Beard.] Conference on Automatic Computing Machines. Proceedings of conference on automatic computing machines held in the Department of Electrical Engineering, University of Sydney, August 1951. Reproduced typescript. Melbourne: Commonwealth Scientific and Industrial Research Organization, 1952. Original tan printed wrappers, cloth backstrip, light wear and creasing, one corner chipped. Boxed. 220pp. 12 plate leaves, text illustrations. 248 x 198 mm. Provenance: Signature of N. F. Astbury of the New South Wales University of Technology on the front wrapper. Astbury was one of the participants in the conference. Book Id: 39064

First edition. Proceedings of the first Australian conference on electronic digital computing as well as analog devices. Trevor Pearcey, designer of CSIRO Mk 1 (later known as CSIRAC) and Maston Beard, the machine's chief engineer, held the conference to demonstrate CSIRO Mk 1, the first electronic digital computer built in Australia. Their research group at the Sydney-based Radiophysics Laboratory of the Council for Scientific and Industrial Research started to build the machine in 1948. It ran its first program in November 1949, becoming the fourth stored-program computer to operate, after the Manchester "Baby", EDSAC, and BINAC (which only ran for a short time). CSIRO Mk 1's main memory consisted of twenty-four acoustic mercury delay lines, each with the capacity to store sixteen twenty-bit words; this was later upgraded to 32 words to give the machine a total of 768 words of storage (a magnetic drum storage system was added in 1954). With two thousand vacuum tubes it could process about one thousand multplications per minute.

In 1955 CSIRO Mk 1 was moved to the University of Melbourne, where it was renamed CSIRAC; it became operational again in 1956. The machine was retired in 1964.

CSIRAC was notable for the originality of its architecture, which was not directly derived from ENIAC, EDVAC, the Manchester "Baby," or Turing's Pilot ACE, since Pearcey did not have close contact with the groups developing those machines. However, from the brief references in his 1948 report, it is clear that Pearcey did read von Neumann's First Draft of a Report on the EDVAC, and that he was influenced in his conception of CSIRAC by von Neumann's theoretical ideas.

Considering that Pearcey was educated in England, and that the Cambridge EDSAC was being developed at the same time as CSIRAC, we may also assume that Pearcey and Beard had some contact with the Cambridge computing group. This might also explain why Douglas Hartree from Cambridge was the only non-Australian representative at this conference.

"The CSIRAC was in operation before the date of the first Computing Conference in Australia held in Sydney in June 1951, but the actual date of its first simple program run, late in 1949, was not properly recorded. It was first publicly demonstrated during the 1951 conference, which was attended by Hartree. From then until mid-1956 it was steadily improved and gave good service at the Division of Radiophysics; this was in spite of serious delays of six to nine months in its development caused by power shortages in Sydney that led to frequent prolonged power blackouts and subsequent fault-finding delays" (Beard and Pearcey 1984, 112).

The first Australian computer conference proceedings is mimeographed from crude double-spaced typed text. Trevor Pearcey contributed four papers:

1. Digital calculating machines used by C.S.I.R.O. (pp. 42-50) This may be the first published description of the CSIRAC, which was then known as the "Radiophysics Mark 1 Automatic Computer."

2. Programming for the C.S.I.R.O. digital machine. (pp. 81-92).This explains the differences in programming between the CSIRAC and the EDSAC, the programming for which Douglas Hartree described in another paper in the conference. Pearcey wrote, "Certain differences between the approach to programming for this machine and that for the EDSAC are due to the different methods adopted for coding commands in the two machines."

3. Programming for punched card machines (pp. 107-126).

4. The functional design of an automatic computer (pp. 127-142). A theoretical discussion.

Of the remaining nine papers in the conference proceedings, four were contributed by Douglas Hartree. Other papers chiefly concerned analog machines, such as the C.S.I.R.O. Differential Analyzer, or analog to digital conversions.

The bibliography of references on pages 210-11 of this work is notable in its omissions. References are given to the early works on the Harvard Mark I, which Pearcey visited in 1945. There are also several references to the works of Maurice Wilkes with whom Douglas Hartree, the only non-Australian speaker at the conference, worked in Cambridge. The list of about 150 registrants at the conference indicates that all attendees were from Australia. When we checked OCLC cited eleven copies (nine in the United States, one in Australia, and one in England). Origins of Cyberspace 832.


Price: $2,500.00

See all items by ,