Professor Frank Hall Sumner

Born in Manchester, England, the son of a steam train driver, he was educated at the North Manchester High School and at the University of Manchester, from which he graduated in 1951 with a First-Class degree in Chemistry. He was awarded a Mercer scholarship and continued as a post-graduate at Manchester, working in Theoretical Chemistry under Professor Christopher Longuet-Higgins. This involved a great deal of tedious computation, initially using a Marchant electro-mechanical desk calculator. He obtained his Ph.D. in 1954, the title of his thesis being "The application of electronic computers to molecular orbital calculations". Between 1954 and 1956 he held an NRDC Fellowship in the Computing Machine Laboratory at Manchester. He spent the next eighteen months as a Scientific Officer at AWRE Aldermaston, in lieu of National Service. After this he returned to the University of Manchester as a Research Assistant in the Department of Electrical Engineering and in due course joined the staff. Since the arrival of F. C. Williams and Tom Kilburn in January 1947, this Department had been the center of computer design at Manchester. He worked on the computer recognition of hand-written characters and, for a short while, machine learning based on simulated neural networks (perceptrons). For this work he used the Ferranti Mark I and then, from mid-1958, the Ferranti Mercury computer. In 1957 a decision had been made by Tom Kilburn's research group to design and build a fast computer with a target speed of a microsecond per instruction. The project was called MUSE, short for Musecond Engine. The aim was to make MUSE a hundred times faster than Mercury ? at the time an ambition that some thought completely unrealistic. By the end of 1958 he had joined the MUSE team, at which time Ferranti Ltd. also became involved and the project's name was changed to Atlas. He worked on the logical design of the Atlas central processing unit, where his major contribution was to devise a technique for the overlapping of instruction execution. He was also a member of the group which developed the implementation of Virtual Memory for Atlas. He simulated the Atlas one-level store on Mercury and used this to develop the Drum Transfer Learning Program ? which optimized page movements. The seminal Atlas concept of Virtual Memory is today used in most computers. He was named as a co-inventor on a relevant patent. Atlas was in service from 1963 and was rated as the world's most powerful computer. Within three years a program of detailed performance evaluations had led Tom Kilburn's group to propose a new high-speed computer called MU5. From 1966 to 1974 he was a central member of the MU5 design team. At various points he was able to spend time with other computer research groups, notably three months at the University of Chicago (in 1963), three months at CERN (in 1973) and six months at the IBM Research Laboratories in San Jos?, California (in 1977). In 1964 he became the first Admissions Tutor for a separate Department of Computer Science which had been established at the University of Manchester under Tom Kilburn, with an undergraduate degree program starting the following year. He held this post for several years. He became deeply committed to the development of Computer Science as an undergraduate discipline and was a member of the UGC Mathematics and Computer Science Committee for four years. This involvement led to numerous appointments as External Examiner for undergraduate and post-graduate degree programs at other universities. He was promoted to Professor of Computing Science in November 1967 and was subsequently appointed as the Barclays Professor of Microprocessor Applications in 1980. In 1983 he was appointed to the part-time post of Director of the University of Manchester Regional Computing Centre. Then from 1989 until 1995 he was Director of Computing Services at the University. In the wider world, he was invited to serve on many national and international committees. Within the UK, he was a member of the Computer Board from 1974 to 1978 and was Chairman of the Computing Committee of the Agriculture and Food Council for eight years. He served on several SERC committees, including the Engineering Board. He had many responsibilities within the British Computer Society and was President in 1978/79. He was a member of several European Community committees and was a technical consultant to NATO. He was involved for several years on the program committee for successive IFIP World Congresses on Computing and was Chairman for the 1980 Congress. He retired in 1995, though he continued to give courses and to collaborate with Manchester's Centre for Novel Computing for some years. In 1993 he had been active in forming the North West branch of the Computer Conservation Society and he served on the main CCS Committee from 1995 to 2000 until his health started to deteriorate. His interest in history led him to give strong support for the building of a working replica of the June 1948 'Baby' computer, known formally as the Small-Scale Experimental Machine (SSEM). He was a member of the Baby Replica Management Committee from 1995 until its completion in 1998. The replica was installed in the Manchester Museum of Science and Industry, where it remains as a working exhibit. Being tall and powerfully-built ? in 1954, he had been elected to the University's prestigious XXI Club for his excellence at basketball. He has been described by a former student as "an amiable sprawling giant". He was a committed Manchester United supporter and a season ticket-holder for fifty years. A Ferranti colleague on the Atlas project has said that "Frank was the most approachable chap one could ever imagine. This is borne out by the fact that colleagues within the Atlas project, whatever activities they were engaged in, and also many individuals in various parts of the University, all had a good word to say about him". Certainly all who knew him held him in high regard, both personally and professionally. He married Ellen Macfarlane in 1956. She passed away after 56 years ? it was said that theirs was a happy relationship. They had one son, Robert.