The Committee approved the minutes of the meeting held on 29th May, 1963.
Minute 5 - Data Transmission Links: Dr Howlett said that he had not yet prepared a paper on data transmission links for circulation to likely users. The Users Committee wanted first to discuss the developments which were taking place in available equipment and he proposed to write a paper for the next meeting of the Users Committee at about the end of the year.
Minute 5 - Users Committee: It was agreed that minutes of the Users Committee should continue to be circulated to the Atlas Computer Committee.
Minute 6 - Provisional Demand for Atlas time from the AEA: It was pointed out that the reduction of the AEA demand for Atlas time which had been agreed at the last meeting was to be achieved by removing the whole of the Culham demand. This would make the revised AEA demand 25, 35 and 40 hours respectively in 1965/66 and the two following years, rather than a level demand of 40 hours in each of these years as stated in the previous minutes. The new figures were included in the revised paper NI/ACC/63/7.
Minute 6 - Rate of charging for the use of Atlas: The Chairman said that in discussion outside the meeting the basis of charging had been agreed. Charges would be based on the assumption of use of the machine on three shifts with 5,250 hours per year of useful computing time.
Minute 8 - Hire of Television links: Dr Howlett said that he had made enquiries of the Post Office about the hire of time on high quality coaxial cable links and it appeared that these were on full time lease to the television companies. The Chairman asked him to continue his enquiries to see if there was free time which could be hired from the television companies.
The following particular points arising from Dr Howlett's Progress Report were discussed:
3.1 Adequacy of the main output systems: Dr Howlett was asked to look into the speed and capacity of the output well to confirm that it was not likely to be a bottle neck in the operation of the machine.
3.2 Progress with the machine: Dr Howlett said that the next progress meeting with the manufacturers would be in January. it was pointed out that this would be a most critical progress meeting and the Chairman asked Dr Howlett to probe very fully into every detail of the firm's progress.
3.3 Joint appointments with the universities: Dr Howlett said that the response to the advertisement for the joint appointment with Trinity College, Oxford had been excellent and included about 60 interesting applications and at least one from a man of established international reputation. It might be desirable to make a senior or band appointment in which case the proposal would be submitted to the Institute's Personnel Committee in the normal way. the Committee expressed their approval of making this a senior or band appointment in the case of a suitable man.
Dr Howlett said that progress towards making the joint appointment with Pembroke College, Oxford was also satisfactory.
The Chairman said that the technique of making double-hatted appointments between universities and the Atlas Laboratory seemed to him to be a promising one which might be extended in due course and the Committee agreed with this view.
3.4 Users Committee: Dr Howlett said that the Users Committee minutes attached to his report were unconfirmed and Dr Wilkes said that the third paragraph of Minute 5 would need to be amended.
Dr Howlett said that now that Culham work would not be going on to the Atlas, a Culham representative had not been included in the Users Committee so there would be only one AEA representative. it was agreed, however, that the Committee could if they wished co-opt someone from Aldermaston, not as a user but as a man with useful experience of another Atlas.
Dr Howlett said that the manufacturers had set up various committees of Atlas owners and that he would like to take part in these as appropriate. The Committee agreed that this was a matter for Dr Howlett to decide.
3.5 Programming Activities: Dr Marshall confirmed that Mr A R Curtis of AERE and Mr A E Glennie of AWRE would be preparing a Mark III version of the Fortran compiler. Dr Howlett said that it remained for decision whether the Atlas Laboratory staff should undertake the preparation of a more efficient Algol compiler than that which the Atlas manufacturers were preparing. the latter would be available at the end of 1963. It was agreed that it was Dr Howlett's responsibility to see that an efficient Algol compiler was available if the universities wished to use this language.
3.6 Additions to the Installation: Dr Howlett said that the disc store, visual display unit and third IBM tape deck mentioned in his report were items which he might bring forward for consideration and if so he would make a detailed case at the time. the disc store was an item attracting growing interest and one for which he might make a case quite soon.
The Committee expressed interest in the possibilities of visual display and it was pointed out that rapid advances were being made. A visual display of sufficiently high resolution could be of very great value in many fields including molecular biology and meteorological work. It was agreed that this matter should be more fully discussed at the Users Committee.
Dr Howlett mentioned one further proposed addition; a Benson-Lehner graph plotter which belonged to the AERE and which was available for transfer. it had already proved very useful. The value was about £20,000. the Committee agreed that it was technically desirable for this machine to be transferred and asked the Secretary to see to the financial arrangements.
3.7 The Atlas Laboratory programme of work: The Chairman said that he would like a broad programme of the work of the Atlas Laboratory for the next few years to be drawn up putting into perspective the various matters discussed in Dr Howlett's report and others like them, and for example bringing out the reasons why such things as a disc store, a visual display unit and a more efficient Algol code might be required. It was agreed that the report should be written in reasonably non-technical language but technical reports on some items might be needed as supplements to it. Dr Howlett said that he would prepare the report and it was agreed that the aim should be to do so in about six months.
Dr Marshall enquired whether it was yet possible to estimate the actual charges for particular calculations. Various requirements for fairly simple calculations were developing at Harwell and he wanted to know whether it would be more economical to get these on to Atlas or to provide a separate small computer. It was agreed, however, that it was impossible to estimate charges for particular calculations at this stage. A formula of the same kind as that used at the Manchester Atlas would be employed so that the charge would depend upon how much of the machine was used, but the formula could not be worked out from information yet available.
It was agreed that the broad aim in operation should be to get the machine used to approximately its full capacity as soon as possible and when that stage was in sight to become very selective as to the problems accepted.
With regard to the demand from the Rutherford Laboratory Dr Pickavance pointed out that most of this was for university projects at the Rutherford Laboratory. He also said that the 40 hours per week shown for 1967/68 was less than he had expected the demand to be. the figure had been kept down, like the AEA figure, in order to leave room for the general requirements of universities and Government departments.
The Chairman drew attention to the fact that the Institute would not charge itself for the use of the Atlas Computer and pointed out that the Treasury might regard this as a hidden subsidy to nuclear physics.
It was agreed that a reserve of time ought always to be retained for use at the Director's discretion for unexpected jobs. Wherever possible the time should not be left unoccupied but should be used for work of low priority, which could be made to wait if necessary. in any case a good proportion of time should be left not booked up by long term commitments. Dr Howlett was asked to deal with this matter in his report on the future programme of work.
The Chairman said that he was advised by his Finance Branch that tough arguments with the Treasury about the estimates could be expected particularly where the estimates exceeded the sums shown in the Institute's five-year expenditure forecasts. These increases occurred particularly in the estimated non-capital expenditure and in the provision for new capital equipment in 1964/65. The Chairman also questioned the figure of £5M in future years which had been put in to represent the eventual replacement of the Atlas. It was agreed that this figure could only be speculative and should not be put in so definitely. On the specific point of the amount included for salaries and wages on the last page of the estimates, the Chairman said that the figure of £82,000 in 1966/67 seemed high in relation to the staff numbers and to the AEA experience at Aldermaston. He asked Dr Howlett to look into this.
After discussion of the provision for new plant in 1964/65 it was agreed that although the technical case for a disc store had not yet been considered, there was a possibility that this might be agreed upon and paid for in 1964/65. The case put to the Treasury should be on the lines that several items of equipment, a disc store, a visual display unit etc were under consideration and that some money should be provided in 1964/65 though not necessarily the whole £250,000 shown in the estimate. The Chairman asked the Secretary to arrange for a meeting with him an the AEA Finance Branch at which the brief for discussion with the Treasury would be agreed.
Dr Howlett introduced paper NI/ACC/63/9 and first apologised for having exceeded his complement for the current year by five owing to progress having been more rapid than expected. The Committee noted that this had happened and accepted the explanation on this occasion on the grounds that the machine seemed likely to be several months ahead of the guaranteed delivery date, and that the additional five people did not involve expenditure exceeding the approved estimates.
Dr Howlett said that the main part of paper NI/ACC/63/9 presented his revised estimate of the numbers required in the operations group. These had been prepared after much more detailed consideration than the provisional estimates made 18 months previously in NI/ACC/62/1 and would mean increases of 12 people for one shift working, 14 for two shift working and 18 for three shift working. The Chairman said that he was somewhat worried by these substantial proposals and would like to deal only with immediate requirements at present. After some detailed discussion Dr Howlett was asked to deal with complement requirements more fully in a paper for the next meeting. In the meantime it was agreed that the total complement for 1964/65 should be not less than the figure of 49 previously proposed.
Research Posts: The Committee agreed to an increase in the number of research posts from 6 to 7 provided that this could be accommodated within the complement already discussed. They expressed a preference for joint appointments with universities for these posts.
The Secretary was asked to try to fix the next meeting of the Committee in early May, 1964
