Correspondence 1960

16.08.60 M I Michaels, Office of Minister for Science, to A E Drake, UKAEA

Immediately on my return from leave yesterday, a Mr Swann of Ferranti's Computer Department, came along to see me following up a talk which the Minister had had with Mr Sebastian de Ferranti. Turnbull wrote to you about the Authority's requirements for an Atlas computer, to which Hitchman replied:

Ferranti's were somewhat disturbed by the Authority's recent announcement of the arrangements for hiring Stretch. Swann said that, from the point of view of other potential purchasers of Atlas, it would seem that the Authority preferred hiring the American machine to buying an Atlas.

He was aware that the Authority had in mind the possibility of buying the Atlas some time in the future. Swann indicated that he thought Atlas would be ready in about three years, and that, while they fully realised that they could give no guarantee about the performance of the Atlas, they would hope that the Authority would be willing to place a provisional contract, which would include clauses to protect the Authority against the uncertainties of performance and of the Authority's own requirements. I pointed out that, whatever the protective clauses, the Authority would have made a provisional commitment, which might leave them open, if something were to go wrong with Atlas, to criticism by the PAC.

Ferranti's seem to attach so much importance to the pyschological value which would result from an order placed by the Authority that I thought I had better convey their suggestion to you. There are obvious difficulties in this suggestion, and perhaps we can have a word to see what might be done to help Ferranti's.

01.11.60 Jack Howlett to Sir William Penney

I have just been reading the draft paper we are going to discuss at the CPC tomorrow and I am trying to take stock of my personal views of the Research Group's case. I talked this over with Dr Lomer and we both thought you might like to have a note of these before the meeting.

1. It seems to be generally agreed that the Research Group has plenty of computing in the ATLAS class; most of the demands now visible come from Fission Reactor and CTR projects.
2. The computer could be sited at Harwell, Winfrith or Culham.
3. Bearing in mind the fact that ATLAS, at the cost of about £2m, could be a major scientific asset:
   1. Winfrith is not a good site: it is physically a long way from all other scientific centres and is essentially a one-discipline (reactor technology) station.
   2. Culham is not a good site as it too is a one-discipline station; also, it is intended for the present at least, to remain fairly small.
4. Harwell has the following advantages:
   1. It is a very large scientific centre with wide interests and an international reputation.
   2. It is on the same physical site as the National Institute which fast will improve its already close contacts with University and International science.
   3. It is presumably intended to be equipped so as to stay in the forefront of scientific work: an outstanding computer would hope to attract first-class people.
   4. It has already done a lot in the way of spreading the use of automatic computation into all fields of this work, so the climate for an ATLAS would be favourable.
   5. It is close enough to Culham to supply the large-scale needs of the CTR work without any difficulty.
5. We must bear in mind the fact that no new machine can be obtained before the end of 1962 and whatever is bought will then have to serve for 2 to 3 years at least; so what we decide now will settle Harwell's computing power until 1965 and probably later.