This is simply a series of factual statements for the record.
The two SRC General Notices, 21/75 and 30/75, giving the decisions on regrouping and management of the Atlas Laboratory respectively, are reproduced as Appendix A.
A Computer Regrouping Co-ordinating Committee has been set up with membership and terms of reference and approved by the Council, as follows:
Terms of Reference
The report of the Working Group on Engineering Computing Requirements was given to the Atlas Committee as Paper ACC 75/5 (11 March 1975). The relevant Council Minute is reproduced as Appendix B.
A Technical Committee, chaired by Professor Rosenbrock (who was Chairman of the original Working Group) has been set up to make explicit proposals for the interactive facility which it has been agreed should be set up. The intention is that proposals shall be put to the Engineering Board and Council in October 1975.
The report of this Panel - now becoming known as the Burke Report - was given to the Committee as ACC 75/4 (13 March 1975). It was discussed at the Science Board on 14 May 1975, and the relevant Minute is reproduced as Appendix C.
The Panel was set up by the Science Board to deal with the short term and urgent problem concerning the allocation or Atlas resources with particular reference to the facilities available on the 360/195 at the Rutherford Laboratory. It was asked to advise the Board on what steps should be taken by the subject committees in their handling of new requests for facilities, and to consider whether to recommend that more time should be made available on the 360/195 or a similar computer. The Panel met on 4 July ]975, and will report its findings shortly.
1. At its meeting on 16 April, after extensive consultation with the various interests concerned, including Staff Side, the Council took a number of decisions about the regrouping of activities in Establishments.
2. Support for High Energy Physics will be concentrated at the Rutherford Laboratory. Work at the Daresbury Laboratory in preparation for CERN experiments which are already approved, such as the e γ and muon programmes, will continue there, but work in support of all new proposals will be the responsibility of Rutherford Laboratory. Any HEP work undertaken at Daresbury after the end of 1978 (ie one year after the closure of NINA) will be carried out only at the request of, and under the policy direction of, the] Director of the Rutherford Laboratory. Computing in support of High Energy Physics will also be concentrated progressively at the Rutherford Laboratory. The consequences of this decision for the staff at Daresbury are being examined both locally and in London Office.
3. A substantial part of the computing now carried out at the Atlas Computer Laboratory to advance a wide range of sciences will be transferred to the Daresbury Laboratory where it will support a growing variety of work outside the field of high energy physics. The transfer will begin in 1976 and will take several years to complete.
4. The Council have decided to set up the Interactive Computing Facility recommended by the Engineering Board. The part of the Atlas Computer Laboratory remaining at Chilton will form the base for the new facility. The Council are also considering with the Department of Industry the establishment at Chilton of a national computing campus to which the interactive facility would be the initial SRC contribution. A further announcement will be made later.
5. After the retirement of Dr Howlett later this year, Dr Stafford will become Site Director responsible for all the Council's activities at Chilton. A new Division Head will be appointed to be responsible to Dr Stafford for the activities of the Atlas Computer Laboratory remaining on the site.
6. The Council have set up a committee under the Chairman of Council, consisting of representatives of each Board and of the Laboratories concerned and the Director, Administration, to supervise and co-ordinate the implementation of the decision for the regrouping of computing and to ensure the preparation of a concerted plan for the development and procurement of the facilities that will be needed over the next five years.
7. The specific plans for redeployment of staff will be worked out progressively in consultation with the staff and trade union sides. Information will be given about these plans as soon as possible and, in accordance with the normal SRC practice, staff affected will be given good notice of any changes. Subject to the over-riding needs of the work, it is the intention that wherever possible every effort will be made to ensure that any necessary movement of staff will be made on a voluntary basis and full account will be taken of the personal circumstances of the staff involved.
S F Edwards
London Office
Chairman
2 May 1975
Dr J Howlett, CBE, who has been the Director of the Atlas Computer Laboratory since its inception, will retire on 29 August. Responsibility for the Laboratory will be assumed by Dr G H Stafford as Site Director for all the Council's activities at Chilton. The appointment of a new Division Head to be responsible to Dr Stafford for the activities of the Atlas Computer Laboratory remaining at Chilton will be announced later when more progress has been made with the plans for an interactive working facility and a national computing campus. Meanwhile Dr Stafford has appointed his Deputy Director, Dr G Manning, to take personal charge of these activities.
J B Visser, Director Administration
8 July 1975
Minute 5
5.1 Mr Ferguson explained that the Working Group on Engineering Computing Requirements had been established because the Engineering Board needed to provide interactive computing facilities systematically. The recommendations of paragraph 2 (SRC 3-75) were summarised and it was noted that the Working Group had concluded that the proposed central facility would be more acceptable if sited at an SRC establishment rather than at a university centre. The financial details of section 5 of the Report were provisional. The Group recognised that free-standing computers would. continue to be required in some areas. The Board and the Computer Review Panel had welcomed the Report and there was now an urgent need for detailed proposals.
5.2 In discussion the following were the main points:
5.3 The Council:
Minute 8
8. Computational requirements in science - report of a study SB 37/75
The Chairman introduced Professor P G Burke who had chaired the ad hoc panel set up to carry out the study.
Professor Burke said that while the brief of the panel had been to take a long term view of national computing needs, their deliberations had led them also to make recommendations on the Council's computing services in the short term. The Panel was convinced that there was a continuing need for the role of the Atlas Computer Laboratory, under circumstances which are now very different from those obtaining when the Laboratory had first been set up. The Panel saw major needs for computing resources arising from the nuclear structure facility, from the high powered laser laboratory, and especially from the new synchrotron radiation source, as well as from the University activities which were discussed in the text of the report. He drew attention to the table which set out the present and projected future requirements of major fields of science, referring in particular to quantum chemistry, the largest individual user. He pointed out that it takes five years to write a major computer package for use in this field; the projection of future usage or computing time therefore corresponds to the exploitation of areas in which extensive preliminary work has already been carried out. In a similar way, the activities of the School of Theoretical Astronomy at Oxford were expected to generate a substantial demand for computing time in three or four years.
Professor Burke referred to the important role. that vector(array) processors would play in the future. Vector processors by handling arrays of numbers rather than digits can have a capacity of from ten to twenty times that of a conventional machine at the same capital cost. He considered that the report may have been optimistic in its estimate of the time that would elapse before a vector processor became available; he would not now expect this to be the case before the end of the present decade. Any Computer system installed by the Council in the future must be able to be linked to a large number of active groups throughout the country - perhaps in time as many as fifty. more immediate requirement, however, was for the enhancement of the existing resources at the Rutherford Laboratory to meet the demands of the near future.
The Chairman thanked Professor Burke for his presentation. It was clear that the identification of the computer hardware appropriate to future needs would need to be examined very carefully. In the immediate context, however, he would ask the Board to comment on the current use of the Atlas facilities and on the projections of future demands estimated by the Panel.
Professor Cadogan said that the Chemistry Committee's decisions on applications from quantum chemists for computing time were coloured by the fact that once computing capacity had been established the notional costs associated with the use of ACL facilities were not a real charge on Council's or the Committee's funds. If the requests received by the Committee had been in direct financial competition with other requests for funds, then, he considered, much of it might have been rejected. He would not advise Council to rely on the demands of quantum chemistry to furnish part of the case for new computing resources. The Chairman agreed that the development of computing packages by IBM, for instance, had stimulated research programmes that had little effect; as yet, on mainstream chemical thought; nevertheless computational and theoretical chemistry would have a major role to play in the future in support of experimental work. Professor Coles considered that the applications of computing to physics were in general, rather nearer to the mainstream of current developments than Professor Cadogan considered to be the case with chemistry: for instance, his Committee was giving support to a "band structure facility" which is expected to involve much computation. He agreed, however, that his Committee was not considering applications for computing time in strictly realistic financial terms. Dr Fender agreed that part of the time used on the Atlas facilities was taken because suitable packages were available, rather than because of a realistic demand for new schemes; nevertheless, a substantial increase in genuine demand is likely to arise in solid state studies. Overall the expansion would be considerable, though he felt that the predictions in the Panel report were rather too large.
Chairman of Council said that an increased demand for computing time might stem from concentration on special areas in the future. He drew attention to the different respective roles played by theory and practice in different fields. The subject of colloid science was an example of a field where theoretical support was at present lacking; the Council might play a useful role in reorienting theorists to it from fields such as quantum chemistry which had been described as over-subscribed.
Chairman of Council emphasised the importance of SRC's computer hardware policy to the future of the next generation. of machines in this country, since the initiative for the introduction of New Range computers was unlikely to come from the Computer Board. He would like to see the Panel report referred to the Computing Science Committee. Dr James agreed that the case for expansion of SRC's computing resources might be better based on the needs of computing science, rather than of the Physical Sciences as set out in the Panel report. Professor Burke had said that the increase in the Council's computer capacity proposed by the Board corresponded only to about 10% of the total computing capacity in the UK and that there was room for more collaboration between the UGC and the Council in the provision of computing. Dr James felt that since the quality of much of the work done on university computers was generally acknowledged to. be very low, there was room for a rationalisation of the use of existing capacity before the Council embarked on a major programme of expansion of its own resources.
In reply to a question from Sir Norman Lindop, Professor Burke said that the interactive facility, to be established under the aegis of the Engineering Board would be of interest in several areas of science. One example was work on molecular dynamics. It was agreed that more detailed study of this matter was needed.
The Board agreed that before it would make any authoritative comment on the conclusions and recommendations of the Panel, it was essential that they consider both their current use of computing resources and the predictions by the Panel of their likely future demands on a strictly objective financial basis, comparing the real rather than the marginal cost of computing with their expenditure in other fields of science. It agreed that detailed discussion of the Panel's choice of proposed machine would be inappropriate at this stage.
The Board