Working Party on Combined Use of Expensive Research Equipment (CURE)

Minutes: Computer Section Only, CURE (60) 2

28/11/1960

• Present
  • F F Turnbull (Chair)
  • L H Curzon, Ministry of Aviation
  • Mr R C Griffiths, H M Treasury
  • J A Hudson, Ministry of Education
  • Sir Keith Murray, University Grants Committee
  • Dr P G Pickavance, NIRNS
  • Professor F A Vick, AERE
  • Mr J D Whittaker, Medical research Council

Also Present

• Mr E R Coplestone, University Grants Committee
• Dr J Corner, AWRE
• Mr M G F Hall, H M Treasury
• Mr C Joliffe, DSIR
• Dr D C Martin, Royal Society
• Dr R C Sutcliffe
• Mr J S Sawyer

Secretary: Mr G J Spence

1 LARGE COMPUTERS

The Working Party had before them the following papers:

1. A general paper on requirements for Atlas by Sir William Penney (CURE (60) 2);
2. A note of a discussion by Sir William Penney on American views of computer development (CURE (60) 4);
3. A note by the Chairman (CURE (60) 5) summarising the results of considerations, by an ad hoc group of experts, of the case for an Atlas Computer at Cambridge and London Universities;
4. A note by Sir Graham Sutton (CURE (60) 6) on Meteorological Office requirements for a large computer.

The CHAIRMAN said that the enquiries of the Working Party had revealed a requirement for a computer of the Atlas class at AEA, NIRNS, the Meteorological Office and the Universities of Cambridge and London which amounted in total to a minimum 85-100 hours per week in 1964-4, at which date Atlas was expected to be available. In addition there would certainly be appreciable demands from other Universities. It was known that there were likely to be additional Atlas facilities at Manchester University (with a moderate word store) but that these would be at least in part devoted to computer research and development. No regular requirement for Atlas facilities had been revealed with other Government Departments or elsewhere. This load was comparable with a maximum capacity of an Atlas, worked, intensively on a three-shift basis, of about 150 hours per week. at the same time, it seemed clear from the investigations that the demand was likely to rise steadily.

SIR KEITH MURRAY said that the recommendations of the expert group on the Cambridge and London requirements were acceptable to his Committee. He felt that before a decision was announced, the views of specialists in Cambridge, London and Manchester Universities should be heard, as recommended in CURE (60) 5. Previous experience with computer requirements for Universities had indicated that demand rose more quickly than had been expected, and requirements for high-speed computers from Universities other than Cambridge and London could be expected.

MR CURZON confirmed that requirements for an Atlas in Ministry of Aviation establishments would be occasional only.

PROFESSOR VICK said that the cost of hiring a STRETCH for a second shift would be greater than purchasing an Atlas.

In a discussion on where the Atlas computer might be sited, the following points were made:

1. Effective arrangements were necessary for the control and management of the computer to ensure that time and facilities could be guaranteed to different users in an agreed proportion; that work which could be done on smaller computers was refused; that spare time could be sold to potential users in industry and elsewhere; and that guidance on use and programming facilities could be available for new users.
2. From the manufacturers' standpoint it was desirable that the order for the first production model should come from an authoritative body with a high reputation in the use of computers.
3. The operation and maintenance of an Atlas required skilled teams which were at present available only in a few centres in the United Kingdom.
4. There were clear advantages in siting the computer in the National Institute of Nuclear Science; arrangements for combined use which would be acceptable to Universities could be made on the lines of precedents for other equipment, and the Institute would be equipped to carry out acceptance tests on the computer and to advise subsequently on its operation.
5. It was desirable that the order for the Atlas should be placed by the AEA in order to encourage other important overseas buyers such as CERN to buy Atlas machines; but, subject to consideration of the terms of the Institute's Charter, it was appropriate for the cost to be borne on the NIRNS subhead and that control of operation should be by the Institute. there were precedents for not charging Universities and Colleges of Advanced Technology for use of the equipment. in return, Universities and Colleges would in general be responsible for the management of their programmes of work on the computer.

THE WORKING PARTY:

1. Agreed that the case for the purchase now, for combined use, of one Atlas for delivery in 1963-4 had been made;
2. Recommended that the computer should be sited at the National Institute for Research in Nuclear Science with arrangements for control on the lines indicated in the discussion;
3. Invited the NIRNS, in consultation with the AEA, and other users, to formulate a scheme for the purchase, acceptance and operation of an Atlas computer;
4. Recommended that, in view of the growing demand for high-speed computers, the position should be reviewed in two years, in order to establish whether there will be a case for ordering a second computer to be sited in a University.
5. Invited the Secretariat to circulate a draft report on large computers embodying the conclusions of the Working Party.