Dr Howlett, having formally resigned from the Committee, on his appointment as Head of the NIRNS Atlas Computer Laboratory, attended in the latter capacity.
The Committee approved the minutes of their meeting of 20 October 1961. Three points arising from the minutes were dealt with briefly:
2.1 Dr Howlett reported that the manufacture of the NIRNS Atlas had started in November 1961. The completion of the machine in the firm's works was scheduled for October 1963, after which working up to full reliability might take nine months, and installation at Harwell should begin in July 1964. the machine ought to be in full service by the Spring of 1965.
2.2 Dr Howlett said that the contract was not yet signed. Negotiations were at present concerned with guaranteed clauses. In reply to questions he confirmed that it would be a fixed price contract, and that a penalty clause for lateness was under discussion. the Committee firmly expressed their wish for a penalty clause, either in money or in the form of free computer time, and asked the Secretary to inform the Contracts Department.
2.3 Dr Kilburn said that some components of the Manchester Atlas were now being tested, and it was already clear that the speed of the arithmetical processor was equal to expectations.
3.1 The Chairman said that Dr Vick had offered to look at the building plans, with the help of the AERE Building Committee. Dr Vick outlined certain suggestions which had resulted, in particular the re-siting of the power house by feeding from a new AERE sub-station, and making the office block shorter and wider. the changes, if all proved practicable, would result in a saving of about £10,000, in addition to a reduction of £22,000 in the Project Estimate, as compared with the previous Token Estimate of £310,000 (excluding furniture and fees).
The shorter, wider building would have more rooms shared by several people, and fewer single offices. There was a good deal of discussion on this point, and the Committee wished it to be recorded that they attached importance to providing a reasonable number of single rooms for undisturbed work but that the rest could be in rooms shared between several people, in order to reduce the cost. Glass partitions (double glazed for sound-proofing) and roof lighting were suggestions made for increasing the number of single rooms.
3.2 After discussion of the complement paper (Minute 4) the Committee endorsed the proposals for the office building, on the understanding that the alterations suggested by the AERE Building Committee should be worked out in detail and adopted where appropriate. They also endorsed the proposals for the computer block, noting that these had been very thoroughly reviewed and reduced as much as possible.
The Committee discussed the proposals made by Dr Howlett in Paper NI/ACC/62/1, and agreed that the numbers shown could be accepted for planning purposes. The following points were made in the discussion:
The Chairman asked for critical comments on the need for the full 48K core store, since a reduction in this was the only practical way of saving a large sum of money. It was clear that the bigger the core store the bigger were the problems which could be solved efficiently. Sir Graham Sutton said that as far as could be seen the meteorological problems could be done with 32K store and 14 tape units. In general discussion it was pointed out that the Manchester Atlas would have a 16K core store and the London Atlas would have a 32K core store. Whilst these were adequate for many purposes a great variety of problems of different sizes could be expected and it was essential to have one Atlas in the country with a very large core store so that the largest problems could be solved with maximum efficiency. the NIRNS Atlas, with its 48K core store, would become the centre on which such large problems would be focussed; if its store was to be reduced its value for this purpose would be almost entirely lost.
The Committee agreed with these views and concluded that it is essential not to cut down the computer and that it is therefore necessary to seek the additional cost of the building, and any contingency on the plant, from the Treasury.
6.1 The Chairman said that an unexpected situation had developed in the scheme originally proposed by the Minister for Science's Office and referred to at the first meeting (Minute 1 (b)) and in Paper NI/ACC/62/2. The Minister's Office were now questioning the proposal that the time should be provided through NIRNS without charge to Universities, and suggesting that the Universities should get funds from the UGC or DSITR for the purpose. Mt Jolliffe said that DSIR could handle the provision of grants, referring any large applications to their Research Grants Committee for consideration along with other claims. He had already discussed procedure with Dr Howlett. However, DSIR had suggested that it might be simpler for NIRNS to deal with the matter, since it was so similar to the case of work on the NIRNS Atlas, later on. The Committee agreed with this latter view.
6.2 In a general discussion, the Committee took up again the question of charges for University use of the NIRNS Atlas (which had been briefly discussed at the first meeting on 7 July 1961) and of AEA computers, and it was agreed that no charge should be made to Universities. Points made in the discussion were:
6.3 Turning to the detailed arrangements proposed in Paper NI/ACC/62/2 (subject, of course, to the agreement of the Minister for Science and the Treasury that the NIRNS may finance this scheme) the Committee agreed on the following points:
6.4 The Committee agreed that Dr Pickavance should try to settle the matter on agreed lines with the Office of the Minister for Science. in case of difficulty arising, he should consult the Chairman of the Atlas Computer Committee.