|1942-46||War Work on Radar at TRE Malvern Member of F.C. Williams' electronic problem-solving group at TRE.|
|1947-48||The Williams-Kilburn Tube and the Baby 1947 First effective storage device for electronic computers (2048 bits on a CRT) -- 1948 (June) "Baby", first working electronic stored-program computer ("ESPC")|
|1948-51||The Manchester and Ferranti Mark 1s 1948 Oct Basic design of Manchester Mark 1 -- 1949 Apr Intermediary Version working (first full-sized ESPC) -- Oct Full Manchester Mark 1 working (first ESPC with address registers and two-level store) -- 1951 Feb First Ferranti Mark 1 delivered (first production ESPC).|
|1951 ...||The Computing Service Tom Kilburn takes over leadership of computer development from F.C. Williams. Computing Service started for wide range of organisations (becomes an independent service late 1960s).|
|1951-55||MEG and the Transistor Computer 1953 First? transistorised computer working -- 1954 First? floating-point computer, MEG -- 1955 Expanded TC version (turned into production MV950 in 1956). MEG becomes production Ferranti Mercury 1957.|
|1956-64||MUSE/Atlas 1956 MUSE project started; aim 1,000 times faster than Mark 1s. Massive innovative effort moving from early '50s computer state to mid '60s. Effort required on 3 fronts: hardware techniques, operating systems and multi compiler production -- 1959 Ferranti joins project, renamed Atlas -- 1962 Atlas starts working -- 1963 Software fully working. Arguably most powerful and sophisticated computer in the world. In general the innovation matches simultaneous leading-edge projects in US, but is carried out largely independent of them. Main unique features Virtual Memory and (Brooker's) Compiler Compiler.|
|1964 ...||The Department of Computer Science First UK (European?) Computer Science department.|
|1966-74||MU5 New machine with architecture dedicated to high-level language requirements. c. 1970 Makes major contribution to new ICL 2900 series architecture. Working 1972, with major O/S and compiler production system (under Derrick Morris) fully operational 1973.|
|1975-81||Later Years C.S. department grows and diversifies; (o)100 students/year.|
|Honours and Awards (E.g.) 1965 Fellow of the Royal Society; 1973 C.B.E. -- plus many from Universites and Computer organisations.|
Tom Kilburn went to Cambridge University in the war to study Mathematics, and graduated in 1942 with 1st class honours. On his call-up, he was sent to take a six-week crash course in Electronics, and then to join F.C. Williams' group at the Telecommunications Research Establishment (TRE) at Malvern. Williams had already made a major contribution to the development of airborne radar, and now ran a small central group solving electronic problems for other groups within and beyond TRE. By the end of the war Tom was an accomplished member of the group.
In December 1946 F.C. Williams left TRE for the University of Manchester, to take a chair in Electro-technics (later Electrical Engineering). Tom Kilburn was seconded to Manchester by TRE at the same time to work full time under Prof. Williams on a prospective storage system for use in electronic computers. The lack of an effective storage mechanism, that could easily store both numbers and instructions in a form that could be processed at electronic speeds, was holding up the development of electronic computers worldwide. In 1947, Tom Kilburn, under the supervision of Prof. Williams, and with the assistance of Geoff Tootill, developed a storage system using a standard Cathode Ray Tube that could hold 2048 bits with random access. This overtook the work on other storage systems in the US and UK, and so when the same team developed a small computer to test the system, the SSEM, or "Baby", this was the first stored-program electronic computer to be built and to work, on June 21st 1948.
The "Baby" immediately demonstrated that the storage system was suitable for computers, and that the stored-program electronic computer was viable.
The Manchester Mark 1 was completed by October 1949, and the first Ferranti Mark 1 was delivered in February 1951. An intermediary version of the Manchester Mark 1 was available for use in April 1949, with the peripheral equipment still incomplete. Again Tom Kilburn led the design and building of the Manchester Mark 1. The intermediary version became available at much the same time as the full-sized EDSAC at Cambridge University (May 1949) and was running long programs over the summer of 1949. These two computers were the first two full-sized electronic stored-program computers working in the world. The Manchester Mark 1 introduced the first two major extensions to the classical von Neumann architecture described in 1945: address registers (albeit in primitive form) and fast random-access secondary storage (the "drum", the ancestor of today's hard disc). The address register meant that it was no longer necessary to program instructions to alter each instruction that needed to access a store location whose address was not known when the program was loaded (e.g. an array element). It was some time before another computer was working with a two-level store (apart from the Ferranti Mark 1).
Meanwhile the CRT storage system, which became known as the Williams Tube became one of the two storage systems that supported (most) computers worldwide until "core store" became widely used in the mid-fifties. The first Ferranti Mark 1 was delivered in February 1951, the world's first production computer (just ahead of the UNIVAC). In all nine were sold publicly, three of them abroad.
1951 saw some big changes. Once the first Ferranti Mark 1 was working smoothly effort on it was switched to the software side. The early software for the Manchester Mark 1 and then the Ferranti Mark 1 had been developed under the leadership of Alan Turing. But in 1951 R.A. (Tony) Brooker took over. He quickly made major advances in the quality of the software provision, culminating in the Mark 1 Autocode in 1954. A large computing service was built up, used by a number of universities, industrial firms and government organisations. Also in 1951 F.C. Williams reduced his active involvement in computer development, to concentrate on Electrical Engineering matters, leaving Tom Kilburn, now a Senior Lecturer, to run the Computer Group.
The engineering members of the Computer Group now set to work to design and build two new computers. MEG was basically a version of the Mark 1 with floating point arithmetic added, but with improved techniques it was 30 times faster. MEG first ran in 1954 and was one of the world's first floating point computers. Ferranti built a production version of it, the Mercury, using a core store for RAM instead of Williams-Kilburn Tubes. 19 were built, the first delivered in 1957, and they were a major resource for the scientific computing in the UK. In 1958 the department acquired a Mercury to take over the Computing Service from the Mark 1.
The Transistor Computer was an experiment in the use of transistors. It had no RAM, with program and data being stored on a drum. A small version was working by 1953, possibly the first working transistor computer, and an expanded version by 1955. Six production versions were built by Metropolitan Vickers, the first delivered in 1956.
In 1956 Tom Kilburn and his team started to look at the design of a machine that would be far larger and, with transistors and core store now available, much faster. It was called MUSE (for microSEcond) and aimed for a speed of 1 million instructions a second. This was 1,000 times faster than the Mark 1 still running the computer service. The innovation required to achieve this speed, and then to deal effectively with the implications of it, was massive -- a long list of new features that bridged the gap between the basic designs of the early '50s and the sophisticated mainframes of the middle '60s, most obviously multiprogramming. Although there were two similar projects in the US with a similar timescale (LARC and the IBM STRETCH) there was little help to be gained from their progress, and the detail of the implementation of the new ideas was carried out independently of them. And of course the massive improvement in power necessitated an explosion in software requirements as well -- a large operating system and a large coordinated compiler suite (given that the international languages FORTRAN, ALGOL and COBOL had to be considered as well as the local Mercury Autocode and Atlas Autocode). Both for funding and manpower reasons the department needed help, and this was provided by Ferranti joining the project in 1959, with the machine renamed ATLAS. Ferranti made particular contributions on the Operating System side and manufacturing. Tony Brooker capped his (existing) Mercury and (new) Atlas Autocodes (high level languages) with the revolutionary Compiler Compiler, a language for writing compilers in, with built in BNF definition and recognition facilities.
Atlas first ran in December 1962, with a limited Supervisor, with another year passing before the Supervisor reached full power. At the time it was arguably the most powerful and sophisticated computer in the world. The new ideas and their practical implementation had a significant influence on computer thinking worldwide, with Virtual Memory and the Compiler Compiler probably the most important ideas particular to Manchester. Three full-sized Atlases were built, and three scaled-down versions. As with the Ferranti Mark 1 and Ferranti Atlas, they provided a major resource for the UK scientific community.
In 1966, with Atlas and the new department and the undergraduate syllabus under control, Tom Kilburn and his team started work on their next machine, MU5. With four years of experience of Atlas acquired, and with the majority of programs being written in high-level languages, the main focus of MU5 was to provide an architecture geared to the efficient running of such programs. Amongst the highlights of the architecture were an associative name store (to reduce code optimisation), segmentation (to enable code to be shared between active processes), and the "descriptor" word; this meant that the code required to access a data structure element had only to give the basic operation plus the address of a location holding a description of the structure (e.g. the array bounds and array element bit size) -- the hardware would do the rest.
There was no agreement this time for the machine to be manufactured, but the large Science Research Council grant for MU5 set up a continuing cooperation with the Ferranti Atlas team, which had been taken over by ICT, and then ICL. However the MU5 architecture made a major contribution to the design of ICL's new 2900 series, and so has had a visible presence long after the department's machine was retired. Tony Brooker left before the MU5 project got going, so it was Profesor Derrick Morris who led the software side. This of course made a major contribution to the design of the architecture, and the software team also created a large innovative operating system and compiler generating system, both of which were designed to work across a connected family of disparate machines (ICL1905E, MU5 and PDP11/10 in practice).
MU5 first ran in 1972, and by 1974 the MU5 complex was running as a complete system, with operating systems and compilers running as planned, and providing a Computing Service for the staff and students of the department.
MU5 was the last in the progression of main-frame computers designed at the University of Manchester, evolving continuously from the Baby. It was now the policy of the department to give up the tradition of having a single major research project (i.e. a new machine!) taking up most of the effort of most of the staff.
Tom Kilburn retired in 1981, handing over to his second-in-command ever since the early 1950s, Professor D.B.G. Edwards. He handed over a flourishing department, now diversified over a large number of areas of Computer Science, with around 30 staff and annual intake of over 100 students.
|1965||Fellow of the Royal Society|
|1976||FEng, founder member of the Fellowship of Engineering|
|1978||Royal Medal of the Royal Society|
|1982||Computer Pioneer Award, IEEE Computer Society|
|1983||Eckert-Mauchly Award, ACM & IEEE Computer Society|
His final honour was to be made a Fellow by the The Computer Museum History Center and his last professional act was in November 2000, the week before going into hospital, to record an acceptance speech in front of the working replica of the Baby at the Museum of Science and Industry in Manchester.