Contact us Heritage collections Image license terms
HOME ACL Associates Technology Literature Applications Society Software revisited
Further reading □ Overview28/09/72ACC 72/5ACC 72/616/03/7303/07/7319/11/7307/03/7423/05/7416/10/7420/03/75
ACD C&A INF CCD CISD Archives Contact us Heritage archives Image license terms

Search

   
ACLLiteratureCommittee MinutesReconstituted ACC :: SRC Reconstituted ACC Minutes
ACLLiteratureCommittee MinutesReconstituted ACC :: SRC Reconstituted ACC Minutes
ACL ACD C&A INF CCD CISD Archives
Further reading

Overview
28/09/72
ACC 72/5
ACC 72/6
16/03/73
03/07/73
19/11/73
07/03/74
23/05/74
16/10/74
20/03/75

Minutes of the meeting held on 28/ 9/1972 at State House, London

1 Preliminary

1. Professor Bondi extended a welcome to the members of the reconstituted Atlas Computer Committee at their first meeting. Formal introduction of the members was carried out by Dr Howlett.

2 Introduction

2. At the Chairman's invitation, Dr Howlett spoke briefly about the role of the Atlas Computer Laboratory as a source of computational support for scientific research work, of its active part in the scientific life of the community, of the necessity for close contact with users in the Universities, of the considerable advantages of a computing centre without formal ties to a particular University or other institution, and of the freedom obtained thereby to serve the scientific community at large. There were prospects of sufficient computing power becoming available to stimulate some exiting new contributions to the advancement of scientific knowledge and understanding. He emphasized his hope that the new committee would feel fully identified with the Laboratory and that it would furnish him with the kind of advice and guidance based on a broader awareness of the needs of science than was possible purely within the Laboratory itself.

3. The Chairman said that the main task of the Committee lay in ensuring that the particular capabilities and strengths of the Laboratory were properly exploited for the benefit of the scientific community and indeed of the whole country.

3 Terms of Reference

4. The Committee asked for clarification of their terms of reference. These were stated to be:
To advise the Science Board on the operation of the Atlas Computer Laboratory and, in agreement with the Boards, on priorities in the work of the Laboratory.
It was held that the terms were narrowly drawn, but the presence of assessors from the Computer Board and NERC allowed for flow of information. The changing role of the Atlas Computer Laboratory in providing guaranteed support for users with especially large computational projects with approval from the various SRC specialist committees might lead to conflicts in resource: allocation. An Allocations Procedures Working Group had been formed to resolve these, and to advise the ACC of areas of special difficulty.

4 Work in Progress or Planned - ACC/72/1

5. Dr Howlett introduced his paper which gave a general outline of the work of the Laboratory and its organisation. He had aimed to give the Committee a picture of both current and projected work, of the main users of the ICL 1906A computer and of the time available to the Laboratory on the IBM 370/195 computer in the neighbouring Rutherford Laboratory. He had also included as an appendix the statistics on the use of the Atlas Laboratory facilities up to the end of March 1972, those for Atlas taking the; same form as given to the previous Committee.

6. The Chairman sought clarification on the likely demands of the had been working towards a coherent computer policy for some months. The Council had studied and reported on computational needs, and subsequently had set up a working group which had invited Dr Howlett to comment on their report and on the needs as he saw them. Dr Howlett had recommended amongst other things that ACL should provide some facilities for NERC on a centralised footing. SRC and NERC had now agreed to enter a partnership for the provision of centralized computing facilities to NERC and NERC were very pleased with the arrangement. It was a little early to be precise about the scale of demand, but it had been decided to establish at ACL a small number of NERC staff to provide not on1v the necessary liaison but also advice and expertise to NERC on the facilities and their development. Mr Seddon said. that he would circulate some short notes about the arrangement. The Chairman welcomed the statement and looked forward to a successful collaboration between SRC and NERC.

7. In response to an enquiry by Dr Thomas, Dr Potter explained that a working party under the chairmanship of Dr Francis, who had since recently retired, had looked into the SRC's future demands for computational facilities. It had also looked briefly at the needs of the other research councils. It had not made any firm recommendations and it merely identified probable demands together with possible growths in them; and the shortfall of facilities to deal with them. Dr Howlett said that the main dialogue had been with NERC and the Social Science Research Council for whom a token earmark of 5% of ACL facilities had been made. Dr Potter agreed to ask if copies of Dr Francis' report could be made available to the Committee.

8. Dr Howlett confirmed that the ICL 1906A computer was being used on work consistent with the new role of the Laboratory. Among major uses had been the data processing program suite for the Ariel 4 space satellite which was an SRC sponsored project. Professor Burke was making considerable use of the machine (over a link, in fact) and so too were a number of quantum chemists.

9. The Committee expressed concern that the impending closure of Atlas would lead to a sharp rise in demand for local University and Regional Computing Centre facilities. This in turn might accelerate the demand for guaranteed facilities at ACL by the larger users ousted from the University and Regional Computing Centres. The Committee emphasized the importance that information be sent out soon reminding users of the closure, of Atlas. The Committee was also disturbed that the ICL 1906A had taken so long to work up to its full potential and asked the Director of ACL to investigate whether more time were available on the IBM 370/195 to meet the demand in a period of shortfall.

5 Delegation of Committee Work

10. The Chairman said that the Committee needed to reduce its administrative business to a minimum if it was to work constructively on the scientific aspects of the Laboratory. The Committee had a duty to recommend the expenditure of money on capital projects connected with the Laboratory, but it could be unduly overburdened with such business. He recommended that he as Chairman should be empowered to recommend expenditure up to £25,000. The Committee accepted this on the understanding that the proposed expenditure was not part of a large capital project.

11. Above this figure and below a figure of £100,000, the Chairman suggested a drill in which the request would be circulated to all members of the committee. If neither he nor the Director received any comments within two weeks, the Committee' s approval would be assumed. Should he or the Director receive written comments which were not of real difference, the proposal revised in the light of the comments would be recirculated for approval. Again if no further comment were obtained within two weeks, approval would be understood. Whenever there were real differences, the matter would be discussed by the Committee at a meeting. The Committee insisted that proposals dealt with in this way would be covered by financial provision in the Estimates or Forward Look. It was normal for the Committee to see both the Estimates and the Forward Look. Proposals sent to the Science Board by the Committee would carry its approval. The Committee agreed on this procedure.

12. Dr Thomas said that the proposal to obtain five GE2050 terminals might be an example of a proposal of the kind just envisaged. It had a central part in the development of ACL and he felt that the ACC should have sight of what was proposed. The Committee accepted that the proposal had been put to the Science Board between the time of disbandment of the former committee and the formation of the present one. There were complex issues and the Director of ACL was asked to produce a paper showing the technical case for the terminals, the processes through which the proposal had passed, and the possible courses of action open to the present ACC on the proposal.

6 Enhancements to the ICL 1906A System - ACC/72/4

13. The Chairman invited the Committee's approval of the proposed enhancements. The Committee understood the need for extra exchangeable disc storage. The Committee felt that an EDS 60 unit would be a more suitable purchase, but it recognised that the shorter delivery time of two additional EDS 30 spindles was a particular advantage. The Committee approved this expenditure of £15,000 on two EDS 30 disc drives.

14. The Committee questioned the need for additional magnetic tapes for the 1906A, but after an explanation of the way the tape system was organised and on being assured of the advantage of another tape control unit, approved the expenditure of £35,000 on two magnetic tape drives and control.

15. The Committee asked if the proposed software enhancements in any way lessened the likely benefits of the extra high speed magnetic drum and channel. Mr Fossey said that it was true that efficiency enhancements to the operating system GEORGE 4 were planned by ICL. It was not possible to be precise about the overall effect of these, but present peak rates in peripheral activity occurred frequently enough to indicate that such improvements as were introduced by the software changes would not outweigh these gained by having more drum storage. The Committee concurred with the proposal to purchase extra high speed drum storage and channels for the 1906A at a cost of £54,500.

7 Enhancement to the PDP15 Interactive Graphics System - ACC/72/3

16. Dr Howlett said that the aim of the proposal was generally increasing the power and flexibility of the system. Mr Taylor asked if it was intended further to enhance the system and Dr Howlett said that whilst he felt it would be unwise to make a categorical statement he could say that at present no further additions to the PDP15 were planned beyond those proposed. It was necessary to do more processing in the PDP15, and at this stage only one display station planned. The Committee recommended the expenditure of £35,000 on the proposed enhancements.

8 New Microfilm Recorder - ACC/72/2

17. Dr Howlett said that the SD4020 had been an excellent machine. The present difficulty was ensuring its stability with the conflicting requirements for hardcopy and microfilm output. The machine had a relatively coarse resolution and had negligible to non-existent facility for grey scale and colour work. No upgrading of the SD4020 was available, and the machine itself was constructed of now obsolete hardware. Modern microfilm recorders incorporated small computers thus giving more flexibility. Of the machines surveyed the Information International Inc FR-80 stood out as the best device both for performance and quality of output. There was as yet only slight indication of the need within the scientific community for the precision and quality of this machine. He sought the advice of the committee on two questions:

  1. was it reasonable to provide in the UK a machine. of this quality?, and
  2. Should ACL in providing large scale computing facilities also make available a specialist facility of this nature?

18. Professor Bransden said that he was disappointed not to see even a rough forecast of likely demand. Dr Howlett said that the SD4020 was being increasingly used by people with very large projects on the 370/195. The Committee felt that this was a most important observation and were greatly interested to know that the general trend towards the use of graphical presentation was shared by this group of users.

19. The Committee felt that the cost of refurbishing the SD4020 should be examined. The Director was asked to prepare a new paper for approval by the Committee through correspondence. It should cover two specific points:

  1. the cost of refurbishing the SD4020
  2. the basis on which Ferranti could still supply a microfilm recorder and the extent to which it would meet projected requirements.

The paper should also include some indications of the likely usage by research workers with large scale computational projects.

9 DTI View of Laboratory Role - ACC/72/6

20. Mr Taylor first asked for two corrections to be made on page 1. In the final sentence of paragraph (1) the phrase "with only modifications" should read "with only minor modifications". In the first sentence of paragraph (2), the phrase "on one fact" should read "on one facet".

21. Mr Taylor said that the paper was being discussed with DES, but DTI wished the ACC to have an opportunity to consider it. Since DTI were represented on various committees and panels as assessor, they took a keen interest in computers especially with regard to policy. The development of computing in the UK needed continual review to ensure that the overall provision was properly controlled. It was consistent with this that excess capacity on existing computers was made available outside the immediate installation, as for example at the Rutherford Laboratory. DTI had seen a need for a professional unit to help in deciding what computers were needed and where work was to be done. DTI envisaged this as a fitting role for ACL. This was not a policing duty, but rather a focus for developments in the provision of computer facilities. He believed that the role was a satisfactory and challenging one for the technical staff of ACL.

22. The Chairman said that he invited discussion on the paper since the SRC was involved and the Chairman of Council might ask for the views of the Committee. He also said that it was a general problem in the organisation of government that the executive needed a formal relationship with the sources of its advice, which to be valuable had to be based on experience. Monopoly in both executive powers and advice had benefits as well as dangers. For computers, the obtaining of advice based on real experience without recourse to the power of monopoly was the problem facing DTI and their proposal delicately avoided the extremes. From this standpoint ACL was indeed well placed. Yet it was far from clear that the Computer Board and DTI needed additional supportive advice, or that ACL was a good source for it, especially as it would almost certainly create an atmosphere of antipathy among the very people which the Laboratory served.

23. Mr Rutterford said that the Computer Board had been aware of the problem for a long time. Some mechanism was clearly needed for assessing the demand for computing in the University world, for determining a strategy to meet it, and for gauging the suitability or otherwise of particular computer proposals. The Computer Board might not be properly set up to carry out this work. Mr Taylor commented that the Computer Board and ACL were both providing similar services, the former by provision of computers to Universities and the latter by provision of computing facilities in a specialist way for science. DTI wished for better co-ordination.

24. The Committee was concerned that ACL would have primacy of advice in DTI's proposed role without always having the relevant technical experience. The Universities accepted that advice would be sought and raised no objection to the choice of expert opinion. They were invariably unhappy about an ex officio source of such advice. Mr Taylor said that DTI had not considered this aspect of their proposal, but the proposal was intended to marshal the advice of ACL for advantage of both Computer Board and DTI. The Chairman said that the Director's terms of reference could be adjusted to do just this and no problem would exist.

25. The Chairman said that Government supported both University research and the use of science. The United Kingdom had been notably less successful in exploiting its scientific achievements than other countries, as for instance the industrial advantages obtained by France from its space project. The Science Research Council had some responsibility for this. In the computer field, the British Computer industry could no longer afford the heavy burden of building very big fast computers, but there was still scope for considerable growth in software capability within the UK. He thought that ACL should not act in the way envisaged by DTI but rather as an intermediary in software sponsorship. In this way a link between ACL and the software industry would be forged to the advantage of and as an attraction to the Universities. Professor Burke had expressed his view that ACL had gained considerable experience of software compatibility across machine ranges and that the Laboratory was well placed to contribute significantly on software topics.

26. The Chairman summed up the Committee's views as follows:

  1. ACL had given useful advice to the Universities and Regional Computing Centres based upon its experience;
  2. there was a fear that the DTI proposal would reduce the availability of this advice;
  3. other bodies outside the University world needed access to the advice;
  4. the new role for ACL laid down by SRC in no way limited the availability of advice from ACL, but ACL continued to have no primacy in advice.

Finally, the Chairman invited Mr Taylor to consider whether ACL could usefully take some part in the sponsorship of software schemes. This would establish a pilot relationship between DTI and ACL.

10 The New Computer for the Atlas Laboratory - ACC/72/5

27. The Chairman said that the paper presented very real problems. He was inclined because of uncertainty about the scope of the demand for time and resources to wait until the new policy for ACL had established itself and to postpone the item until the New Year. He sought the Committee's views.

28. Dr Howlett agreed that the forecast of demand was still very imprecise, but he hoped that the Committee could advise him particularly of the long term scientific needs. Professor Burke stated that with the two Regional Computing Centre 7600 computers about to start service, demand was not capable of prediction, but he thought that large jobs would soon be pushed off those machines by the many smaller jobs which in the University world took precedence over large projects. Dr Thomas observed that the Regional Computing Centres had no provision for any growth in large scale scientific computation. Professor MacLellan said that present indications in engineering were that more sophisticated and time consuming calculations were being tackled and that a significant increase in capability would be required in the next year or two. The Committee felt that the Science Board might be able to discover demand possibly through the Francis working party or its follow up.

29. The Committee examined what had to be done to assess demand. The simplest method was to obtain access to as much spare capacity as a stimulus for generating demand. Dr Thomas commended an investigation of possible spare capacity on the Meteorological Office IBM 370/195 and on the CDC 7600 computers in the London and Manchester Regional Computing Centres. Dr Howlett was asked to investigate the possibility of more time on the IBM 370/195 in the Rutherford Laboratory and to explore the possible use of other machines for which the provision of funds would be crucial. Dr Howlett made the comment that it was a very difficult task indeed to organise a computing service on a distant machine which was under the control of another body as the Laboratory's experience with the NEL 1108 had shown. The Committee also felt that it was essential to make the best use of the available resources and Mr Rutterford agreed to take up with the Computer Board and the Regional Centres the cost of using the CDC 7600 computers at the weekend. It would be necessary to explore the communications network in this context and the Committee set up a sub-panel consisting of Dr Thomas, Mr Davies, Mr Rutterford, and a representative from ACL to examine the shape and kind of communications network.

30. Mr Taylor was concerned that the evaluation put before the Committee had ruled out the ICL contender as of insufficient power without specifying very precisely the nature of the computing demand to be satisfied. Dr Howlett pointed out that certain scientific problems required individually or collectively the greater power of the American machine, but the need to carry out any particular scientific calculation could not be established by the Laboratory without the guidance of proper scientific expertise available only to the specialist panels of the SRC. The Chairman pointed out that the problem for the Committee was deciding if ACL was a Laboratory having need of the biggest computer.

31. Dr Thomas drew attention to another aspect of placing an order for an American vector arithmetic computer, since ICL had expressed an interest in the manufacture of a special hardware facility providing vector processing, possibly for attachment to the ICL 1906A or to the company's projected new range of computers. He felt that ACTP should put the matter of an exercise on vector arithmetic to DTI. Mr Davies said that if ICL wished to pursue this again, the firm should approach ACTP. Dr Howlett agreed to organise this on advice from Dr Thomas, Mr Taylor, and Mr Davies.

⇑ Top of page
© Chilton Computing and UKRI Science and Technology Facilities Council webmaster@chilton-computing.org.uk
Our thanks to UKRI Science and Technology Facilities Council for hosting this site