Minutes of the meeting held on 4/ 2/1969 at State House, London

• Present
  • Lord Halsbury (Chair)
  • Mr L S Berman
  • Dr J V Dunworth
  • Professor T Kilburn
  • Mr F J M Laver
  • Dr T G Pickavance
  • Mr L F Rutterford
  • Professor M V Wilkes
  • Dr J Howlett
  • Mr A Miller
  • Dr J H Price
  • Mr L S Smith
• Apologies
  • Professor G A Barnard
  • Professor E S Page
  • Professor Sir Rudolf Peierls
  • Mr A I Llewelyn

1 Minutes of the Last Two Meetings

The Committee approved the Minutes of the Meeting held on 26th April 1968 and also the Minutes of the Special Meeting held on 23rd July 1968.

2 Matters Arising

Distribution of Atlas Laboratory Programs by the National Computer Centre Negotiations had been taking place between the bodies concerned, i.e. AEA Contracts Branch, the Science Research Council, the National Computing Centre and the Atlas Laboratory, and it was hoped that the final details would be settled at a meeting due to take place the following day. Dr Howlett said that the Laboratory was anxious to see ASCOP published as soon as possible.

3 Progress Report - ACC/69/1

The report was noted, particularly the high standard of performance the machine had reached. The following points were made:

1. The disc was now fully integrated into the system and there was no need for any further reference to be made to it as an engineering item. The Committee asked if cost-benefit assessment could be made at a later date.
2. Experience so far indicated that the organisational work of the multi-access project was putting only a negligible load on the Atlas central processor. Eight consoles were connected at present and four more were on order; it would be simple to increase the number to 16, but to go beyond that would need more equipment. It had never been intended to develop this into a large system. When enough genuine operational experience had been gained, the Laboratory would write up an account for publication, showing what had been done, and why, and at what cost in money and man-power. The Committee suggested that a comparison with the Cambridge project would be valuable.
3. It was agreed that no further reference need be made to the SC4020 in the Progress Report as an engineering item, but the Committee would always like to hear of interesting uses. Two and a half thousand frames of hard copy and ten thousand frames of microfilm were at present being produced each week; the expectation was that the use would build up to about four times this. The Laboratory was starting to use the SC4020 to put its archive material on microfilm.
4. The amounts of time taken by the individual universities (Table 5) showed interesting differences from the corresponding figures for 1967/68. In general, the newer or newly-expanded universities showed big increases, due doubtless to their research projects getting under way; the striking decrease for Newcastle and Durham was a consequence of the commissioning of the Newcastle 360/67.
5. The continued dominance of Fortran was noted and it was agreed that the situation was unlikely to change for several years. Professor Kilburn said that Manchester University would need to have an Atlas Autocode compiler on its 1906A-1908A system, that ICL had promised to provide this and that the Atlas Laboratory would automatically get a copy.
6. The Committee questioned the need for the periodic shut-down (every four months) for maintenance of the air-conditioning plant, having in mind the consequent loss of production from the future 1906A-1908A. After some discussion of the probable economics of installing sufficient excess capacity to make this unnecessary the Committee agreed that such shut-downs were probably unavoidable.

4 Estimates 1969/70 and Forward look 1970/71-1974/75 - ACC/69/2

Mr Miller said that, whilst the Laboratory's financial needs were dominated by the purchase of the new computer, of which the total cost was now reasonably firm, the actual amounts to be paid in the successive years were uncertain because the terms of the contract with ICL were not yet settled. Another large item was the extension to the building to take the new machine; the cost of this also was uncertain, because ICL had not settled the lay-out of the 1908A nor the method of cooling. It was essential to make financial submissions now, to fit in with the programme for the Council as a whole. The figures given in the paper were based on the best information available at the time of writing. The Committee questioned the need for the Stores building (£20,000) and the Sub-Station (£22,000). Dr Howlett said that the first was for general Laboratory stores but particularly for the greatly increased amounts of cards, paper tape, printer paper, etc., which the new machine would consume; the second was needed because the present electricity supply was already loaded almost to the limit, and would not support the new machine.

The Committee approved the Forward Look, with the reservations that:

1. a more accurate estimate of the cost and rate of spend on the new computer and the new building should be submitted when these could be determined;
2. the cases for the Car Park, Stores building and Sub-Station should be made by Dr Howlett to the officers of the Council.

The Committee noted that there was an apparent run down in the expenditure provided for in the Forward look over the period concerned. Dr Howlett explained that this was connected with two features of the forecast, namely the peaking and run down on capital expenditure on the new computer which was of such a magnitude as to dominate the rest of the figures, and, secondly, what was standard practice at the SRC in drawing up the Forward Look, namely that no contingency allowance was made for capital expenditure in respect of items which could not be specifically foreseen. These were customarily provided for out of the overall contingency reserve kept by the SRC and which was drawn on as the roll forward of the Forward look became specific with the passage of time.

The Committee were satisfied with these explanations.

5 New Computer for the Atlas Laboratory - ACC/69/3

The Committee took note of the position now reached in the negotiations with ICL for purchase of the 1906A-1908A as outlined in the paper. Mr Rutterford said that there seemed to be differences in the schemes for progress payments which ICL were suggesting to the SRC on the one hand and to the Computer Board on the other. Dr Howlett undertook to keep the Computer Board informed on this point.

The first progress meeting with ICL was scheduled for 4th March 1969, at the west Gorton factory; the Laboratory would ask specifically for information on the state and expected progress of the hardware and software for the 1906A and 1908A machines generally. Dr Howlett undertook to circulate reports to the Committee on the outcome of the progress meeting, and to keep the Chairman informed of all significant developments. The Chairman would decide if anything required action by the Committee, either by correspondence or at a special meeting.

On the question of remote input/output stations the Committee agreed to leave these as items in the Forward look, but said that they would want to see a more detailed case for the purchase of these by the laboratory (rather than purchase by the users) before deciding whether or not to support the proposal.

6 Use of the Univac 1108 Computer at the National Engineering Laboratory, East Kilbride - ACC/69/4

The paper had been written in response to an offer from the Ministry of Technology to make substantial amounts of time available on this machine, at very low cost. Managing the use and distribution of this time would involve the Laboratory in extra work but it was felt that, given the link, this could be fitted in with the other activities and it was not intended to ask for extra staff. In the nature of things it had not been possible to foresee the need for the expenditure which would be incurred and to provide for this in the Estimates; if the offer was to be taken up in the way described in the paper, an extra allocation would have to be made to the Atlas Laboratory. The Committee was being asked to give a decision on whether or not the proposal should be supported, so that the UST Board could advise the Council on any financial implications.

In discussion the Committee made these points:

1. The service would be for a limited period only, as the Ministry of Technology were planning to take up the full time of the machine for their own work in 1971. There would be problems in phasing-out the university use.
2. The Scottish universities alone did not have enough work to use up more than a moderate pat of the time offered. However, since the paper was written a suggestion had been made to the SRC that the Kelvin laboratory of Glasgow University (which is housed on the NEL site) should use the 1108 for all its batch-processing work, which might take up to about two hours a day.
3. Whilst it was attractive to consider restricting the use of the machine to large projects, it was virtually certain that the actual load would be very mixed: there never turn out to be enough base-load projects. This would increase the Atlas laboratory's difficulties in managing the service.
4. Some universities - e.g. Edinburgh - had been buying time on the 1108 at the price then ruling of £150 an hour, and were finding it hard to meet the cost. It was not clear how they would feel about paying for time themselves if the cost were only £40 an hour - or even less, if that could be negotiated.
5. At Board level, the proposal would have to compete for money with whatever other projects were being submitted.

The Committee left it to Dr Howlett to reconsider the matter.

7 Interactive Visual Input/Output System - ACC/69/5

Introducing the paper Dr Howlett asked if the Committee would discuss mainly the general question of the need for equipment of the kind described, because since the paper was written Elliott Automation had put in a new, and higher estimate of cost. It would be necessary to look again at the relative merits of this and other manufacturers' equipment, to see if the technical advantages justified the new price. The Laboratory would invite all relevant manufacturers to submit quotations.

The Committee noted that this would be an on-line device, available to people working in the Laboratory - which included, of course, visiting users. It was agreed that on a general principle the Laboratory should be provided with advanced technical equipment, so that university research workers were given access to powerful techniques not available elsewhere, and that this equipment, backed by the big 1906A-1908A system, would provide exceptionally powerful facilities for work in computer graphics. The Committee felt, however, though not unanimously, that the real value of work in this field was still not certain, apart from applications to engineering design which were being funded by the Ministry of Technology. They agreed to leave the provision of £75,000 in the Forward look for 1970/71 but asked Dr Howlett to go into the possible applications in more detail and to resubmit the proposal, with firm costs, at the next meeting.

8 Extra-Mural Research Contracts with Universities - ACC/69/6

The Committee resolved that the Director of the Laboratory should decide on the award of research contracts with universities, taking advice as he felt necessary. They would require an annual statement listing the projects supported and the cost of each.

9 Industrial Fellowships in the Atlas Laboratory - ACC/69/7

The Committee thought the principle behind the proposal was very good and invited Dr Howlett to do what he could to put it into practice.

10 Additions to the S-2 Multi-Access System - ACC/68/7

This item was included simply to ensure that the records of Committee business were complete. The paper had been dealt with in correspondence, the UST Board had endorsed the Committee's support and the equipment was on order.

11 Date of Next Meeting

The date of the next meeting will be fixed later by correspondence, but in any case will be held sufficiently early for the Committee's recommendations to go before the UST Board meeting on 1st July 1969.