There are two further needs for computing machinery which should be mentioned here, although nothing can be settled in this paper.
The first concerns the Rutherford Laboratory, the new National Institute for Nuclear Research, which is being set up on the Harwell site. The 7GeV proton synchrotron, due to begin working in 1962, will produce a very large amount of information whose processing is estimated to be equivalent to between one and two shifts on a 704, that is around 3000 hours per year. The Institute will have other computing needs also, amounting to perhaps another 1000 hours. There is no obligation on the Authority to take any responsibility for meeting this work, but it would seem unnecessary to have two separate computing organisations on the same physical site and we feel that an attempt should be made to arrange for co-operation between the Institute and the Harwell computing group. A suitable machine for this seems to be the Ferranti ORION, available in mid-1961; this is equivalent approximately to a 709, and costs about £400,000. If one could be bought jointly by the Authority and the National Institute and installed in the Harwell Computing Group, then both sides would benefit - the Institute by avoiding the need to set up a computing organisation, the Authority by getting experience of a powerful and novel machine and getting extra computing power at low cost; also the Harwell group would gain considerably from contact with the mathematical problems of the Institute.
The second is the day-to-day demand at the new CTR site at Culham; this will be similar to the corresponding work at Harwell and would be served equally well by a Ferranti Mercury. The need will not become at all serious until well into 1961; a possible way to meet it would be to transfer the Harwell MERCURY to Culham if the suggestion of installing an ORION at Harwell were adopted, for joint use with the National Institute.
These questions need discussion with the Directors of the two new organisations.
