This was Professor Sir Rudolf Peierls's first meeting as Chairman. Before the formal proceedings started Sir Rudolf expressed his personal appreciation of Lord Halsbury's work for the Committee.
The Minutes of the last Meeting were approved.
Effect of the disc: The Committee noted the attempt to evaluate this, reported in paper ACC/69/21 (circulated at the meeting). Dr Howlett said that the evaluation was very difficult, because both the work-load was changing continuously and the general level of performance of the machine was increasing all the time. It seemed safe to say that the disc had given an increase of not less than 5% in throughput, and this would be an economic justification if Atlas were retained for another two years. The Operations Group felt that the disc, which was outstandingly reliable, had increased the general reliability of the complete installation, but the effect being necessarily small was difficult to quantify. They certainly found it simplifies the operators' work by eliminating a good deal of tape handling. A number of things were made possible which were either impossible or very inefficient when only magnetic tape was available for backing store, for example sorting of very large bodies of data; and the multi-access system was completely dependent on the disc. Professor Kilburn said that the corresponding benefits to the Manchester machine were almost certainly a good deal greater, because of its smaller core store and fewer magnetic tape decks.
Report on the multi-access system: Dr Howlett said that the Laboratory had agreed to give a full account at the Datafair meeting in 1972; the paper would have to be written before the end of 1970.
The Committee noted the various items. In connection with Item IV(iv) - the ASCOP statistical package - Dr Howlett assured the Committee that the universities would not suffer as a consequence of the agreement between the Council and the National Computing Centre on the sale and distribution of this program. The agreement included the specific statement that the Laboratory would distribute ASCOP to universities free of charge.
The Committee noted the details of the 1906A configuration given in Appendix A to the paper and agreed that the decision to increase the core store from the original 128K to 256K was sound. There was some discussion of the relative merits of renting and buying the machine; Dr Howlett said that he expected it to be in service in the Laboratory for at least four years, and that this made buying the sounder policy. In any event the Treasury had approved the policy on the basis of outright purchase. Asked about compatibility with other machines, in particular ICL's projected P.52, Dr Howlett said that the 1900s were not strictly compatible with any other machines but that one could always link disparate machines by means of a storage device, for example a disc file. (Note added: this is precisely how Atlas and the SDS Sigma-2 are linked.) The process might be inefficient but was possible.
Dr Howlett confirmed that there was no commitment to the ICL P.52.
The purpose of this paper was to inform the Committee of the design and cost of the new building and of the present state of progress. No action was required. Dr Howlett confirmed that ICL had provided full details of floor loadings, power demands, heat dissipation and cooling demands.
Mr Miller said that, whilst the total amount provided for the 1906A was firm, the actual rate of spend would be decided in the course of contract negotiations with ICL, in which the Finance Branch's plans for phasing the Council's spending as a whole would be taken into consideration.
The Committee approved the proposals of the paper.
Mr Miller said that it had been agreed that the charges should be reviewed after a year's experience, and that this experience had shown that the original forecast of use of the SC4020 was too high. It was likely that the use would increase in the future, especially when the interactive visual-display system (see ACC/69/13) was in operation, but meanwhile it was necessary to increase the charges so that they matched the actual cost. There should be a further review in a year's time.
The Committee approved the proposed new charges.
The meeting concluded with a short discussion of the function and the future of the Laboratory. The main points made were:
The Committee invited Dr Howlett to write a paper giving his views on these points and on the future policy and development of the Laboratory.
To be settled by correspondence, with preference for a date in March 1970.