CSIRO Computing History, Appendix 4: Opportunities

By September 6th, 2019

These pages attempt to give some of the history of CSIRO’s use of computing in its research, focussing mainly on the large shared systems and services.

This page/post is under construction.
Last updated: 26 Oct 2020.
Robert C. Bell

Appendix 3 — Computing History Home — Appendix 5

At the C3DIS 2019 conference, CSIRO’s CIO, Brendan Dalton, mused on the lost opportunities in CSIRO’s history, particularly in the computing field, and the possible value of these lost opportunities.  These notes are a partial response.

  1. CSIR Mk 1 (CSIRAC)
    A lot has been written about CSIRO’s decision not to proceed with developments based on CSIR Mk I (later renamed CSIRAC).  See for example, [1].
  2. The CSIRO Computing Research Section developed its own operating systems, called DAD (Drums and Display) for the CDC 3600 between 1963 and 1966.  This provided interactive access, including the use of graphical terminals, and a file store (the Document Region).  This OS was not even ported to the subsidiary CDC 3200 computers.
  3. CSIRO Division of Computing Research developed the Document Region into a hierarchical Storage Management system by 1973, with automatic recall of off-line files.  This was pioneering, but was not built upon.  Previously, there had been a few innovative resource allocation and flushing schemes.  When charging was instituted for storage, a user was charged up-front for the storage based on the product of the size and the retention period.  So, if a user specified 10 days’ retention, he or she was charged for 10 days upon creation of the document: such a policy might encourage more careful management of CSIRO’s data.  There was innovation that was re-invented, e.g. HSM and flushing algorithms
  4. In 1969, CSIRO DCR sponsored a “Conference on Picture Language Machines” [2].  Subjects to be discussed include “picture interpretation, graphical communication and question answering”, rather anticipating the growth in personal digital assistants today.
  5. Between 1966 and 1975, CSIRO DCR developed remote access and a network, based on packet switching, so that by 1975 there 50 connected nodes in Australia.  CSIRO’s packet-switched technology was contemporary with that used for the founding of DARPANET in 1969, which became the internet.
  6. In 1977, CSIRO DCR developed an interactive system by extending the capabilities of the CDC Cyber 76, which was fooled into thinking it had two infinite tape drives connected, one for reading and one for writing.  The interactive system could support over 100 users simultaneously across Australia.  This was probably the only Cyber 76 or 7600 running an interactive system.
  7. In the 1980s, CSIRO DCR developed a Micronode based on Motorola microprocessors, which was deployed to replace nodes based on PDP-8s and PDP-11s.
  8. In the 1980s, CSIRO DCR developed the Micronodes to run UNIX and become workstations, but was over-run by commercially-produced mini-, micro- and personal computers.
  9. In about 1981, CSIRO established a VLSI design facility. It may have been the basis for chips running FFTs for astronomy, and the first processors for cochlear implants (John Morrissey, personal communication, who has a sample).

Back to contents