Jump Over Left Menu
AEA Atomic Energy Executive Meeting: AEX(60)
Extract from the Minutes
The Executive had before them a paper by Sir William Penney summarising the use of computers in the Authority and making proposals for meeting the needs foreseen over the next five years. Sir William Penney said that the Computer Policy Committee had been examining future requirements in the Authority against the background of a growing use of these machines and evidence that the prospects were not encouraging of using computers at a distance by means of data links. Their starting point was that the cost to the Authority of computing was currently running at £1.5M and 180 man-years a year. Referring to the recommendation that the 7090 machine at Aldermaston should be purchased, Sir William Penney said that it might be suggested that it would be more advantageous to buy a CDC 1604 computer, which was a new American machine, of which twelve were now in operation and which cost less than half the 7090. the Authority, however, was committed to using the 7090 on weapons research for at least the next one-and-a-half years and the hire charges were such that the machine could be bought outright for a further net cost to the Authority of £200,000. It was therefore a better financial proposition to purchase the 7090 than to switch in due course to the CDC 1604. Moreover, the 7090 was familiar to the staff operating it and was known to be a reliable machine.
Turning to the proposal to purchase a KFDF 9 for Winfrith, Sir William Penney said that this machine was six times more powerful than the 704 and would be able to handle all the work arising over the next five years on single shift operation. Culham might use an existing Mercury computer and would also make use of an Atlas if one were available, but it was not thought necessary to provide a new machine for that site.
The major problem to be settled arose on the Atlas computer. This machine would be nearly as powerful as the Stretch but would cost only about one-half or one-third as much. Mr Ferranti had asserted that his Company would cease development of the Atlas if no order was received by the end of the year and it would be wise to take this statement seriously. But while a number of institutions wished to make use of an Atlas computer none of them could establish a case for acquiring one for its own exclusive use. Against this background Sir William Penney had put forward proposals for a collaborative programme which were now under study by a working party set up by the Minister for Science. He assessed that the Authority would need 24 hours a week on an Atlas computer, and the National Institute had indicated a requirement for 8 to 10 hours a week and the Meteorological office for 10 hours a week. These requirements represented one shift. In addition the Ministry of Aviation and DSOIR had requirements and would probably be prepared to join in a collaborative programme and it was possible that the universities would also co-operate.
In discussion the following points were made:
- The proposals in the paper might be criticised on the grounds that they did not make full use of the capacity of the Stretch computer which was to be installed at AWRE. On the other hand, to hire this computer for extra time would be more expensive than the proposals in the paper and, moreover, would rule out the possibility of buying a British computer
- It might be argued that the cost of the proposed KDF 9 computer would be saved if Winfrith were to rely on the Atlas computer to meet its requirements. On the other hand, the indications were that it would be difficult for Winfrith to make efficient use of a computer sited at Harwell. Moreover, Winfrith expected to be able to make full use of a KDF 9 computer. More generally, it could be said that the smaller machines, such as the KDF 9 and the 7090, would enable more effective use to be made of a large Atlas computer
- While a good case could be argued on general grounds for providing Winfrith with a computer of the capacity of the KDF 9, it was not clear that the expenditure involved was fully justified at a time when there was great pressure for economy throughout the Authority.
- More generally, it might be questioned whether all the programmes planned for computers were strictly necessary. There would be a natural tendency to make full use of whatever machines were available and it was difficult to ensure that only essential needs were provided for.
- Further support for British industry could be provided if a KDF 9 computer was purchased as soon as it became available to meet requirements at Risley, these requirements being met in the meantime by the hire instead of the outright purchase of the 7090 machine. this, however, was likely to add significantly to the cost of the proposal in the paper.
THE CHAIRMAN in summing up the discussion said that the case had been made for buying the 7090 computer and there were strong arguments in favour of a collaborative programme on an Atlas machine. The case for the KDF 9, however, was not quite so clear and there would be advantage in deferring action on this machine for a few weeks until a decision had been taken by Ministers on the Atlas.
The Executive approved the paper for submission to the Authority, with the recommendation that a decision on the KDF 9 computer should be deferred as proposed by the Chairman.