OXFORD UNIVERSITY COMPUTING LABORATORY
Annual Report 1967-8
By stages the KDF9 computer reached its final "up-grading" during the year, with the addition of a PDP8 satellite machine, an 11 inch Calcomp plotter, and several remote consoles controlled by the PDP8. Various software improvements were made, taking advantage of the disc store to increase machine "productivity". The latter was also helped by an increase in the number of operators and therefore of Laboratory operated machine time from ten to sixteen hours a day.
The demand for time continued to grow so rapidly, however, that by the middle of the year the machine, operated for some 100 hours per week, was seriously overloaded. It was necessary to introduce a strict check on the rate of machine use by each individual client, and by mutual co-operation the turn-round time was reduced to the order of one day by the end of the year. Further improvements are possible, for example with full 24 hours/day Laboratory operation and extra week-end shifts. This can only be temporary, however, since it is very probable that many problems remain unsolved, both in scientific and non-scientific University departments, through lack of sufficient machine power and certainty of quick turn-round. Moreover many problems are sent to the Chilton ATLAS, the Imperial College 7090 and various other places, with no guarantee of quick turn-round and with some obvious inconvenience.
The academic staff of the Laboratory continued to give undergraduates lectures for the Honour School of Mathematics, and to graduates studying either Numerical Analysis or Theory of Programming Languages in the Diploma in Advanced Mathematics. Nine students supervised in the Laboratory were successful in the Diploma examination, and two others, El Tom and J. Williams, achieved the degree of D.Phil. At the end of the year nine Diploma students and twelve D.Phil. students were expected to be in residence at the start of the 1968-9 session.
Dr. Eldon Hansen, an expert in the relatively new research field of "Interval Analysis", spent a sabbatical nine-month period in the Laboratory, and the Director took the opportunity to organise a three-day symposium on this topic. Other lecturers were invited from Germany (2), Belgium (1), U.S.A. (1) and U.K. (2), the meeting attracted an audience of more than a hundred and Dr. Hansen is editing the proceedings of the symposium.
The "standard" short ALGOL courses are now reinforced by FORTRAN courses of similar length. One of each kind is given in each vacation, and during the year more than 350 graduates and a mall but increasing number of undergraduates took part therein. For the first time, in the summer of this year, the Laboratory staff gave a three-week introductory course in computing to some 40 undergraduates recently successful in Mathematical Moderations, and this is likely to become a permanent annual event.
Mr C. Strachey, Head of the Programming Research Unit supported by the Science Research Council, was "taken over" by the University, at the beginning of the year, with the title of Reader in Computation.
The year under review will be remembered as a turning point in the history of the Laboratory. Early in the year the Laboratory Committee asked the Heads of Science Departments, and some other interested Departments, to make an estimate of the size and scope of the future staff, computing machinery and buildings of a well-equipped University Computing Laboratory. For this purpose the Committee of Heads of Science Departments appointed a sub-committee who prepared a comprehensive paper for consideration by the General Board. The recommendations of this paper, which differed in detail but not in principle from the views of the Laboratory Committee, formed the basis of two applications by the General Board, first to the Computer Board for a new computer for Oxford, and second to the University Grants Committee for a new building for the Laboratory. The General Board also gave a special grant of £7,000 to the Laboratory for the appointment of up to four more senior staff with the particular aim of improving the Service side of the work of the Laboratory.
Members of the Computer Board visited Oxford in July, 1968, were clearly impressed by and sympathetic to Oxford’s case for a new and quick replacement for KDF9, and asked for a more detailed application for consideration in December 1968. The Computer Board and the University Grants Committee are collaborating on building problems and costs, and despite the building "shut-down" announced in July, 1968, there would appear to be every prospect that Oxford will have a new computer, and at least a building to house it, in the comparatively near future and in fact much earlier than seemed likely at this time last year.
FOX and I.B. PARKER. Chebyshev Polynomials in numerical analysis. Oxford University Press 1963.
D.F. MAYERS and F. O’BRIEN. "The calculation of atomic wave functions".
J.Phys. B. ( Proc.Phys.Soc) Vo1. 1, p.145, 1968.
D.M.R. PARK. Some semantics for data structures. From Machine Intelligence 3 (ed. D. .Michie), p.351. Edinburgh University Press, 1968.
J.S. ROLLETT and G.C. FORD. On versions of a procedure for scaling X-ray photographs. Acta Cryst. B 24, p.293, 1968.
Professor G.E. Blackman (Chairman), Dr. H. Bulmer, Professor C.A. Coulson, Dr. D. ter Haar, E.F. Jackson, Professor H.M. Powell, Dr. R. Shackleton, Dr. J. de Wet, Professor L. Fox (Secretary).