Maurice Wilkes

Creator of EDSAC

• Born: June 26, 1913
• Place of Birth: Dudley, Worcestershire, England
• Died: November 29, 2010
• Place of Death: Cambridge, Cambridgeshire, England

• Primary Company/Organization: Cambridge Computer Laboratory

Introduction

Maurice Wilkes was one of developers of the Early Delay Storage Automatic Calculator (EDSAC), which was the successor to the J. Presper Eckert with John William Mauchly's first general-purpose digital electronic computer, the Electronic Numerical Integrator and Computer (ENIAC). EDSAC was the first practical stored-program computer, the first capable of running realistic programs and achieving useful outputs. Wilks also invented microprogramming and wrote an early standard textbook in computer science. Today he is considered the father of British computing and a major figure in the international computing field.

Early Life

Sir Maurice Vincent Wilkes was born to a switchboard operator for the Earl of Dudley's private telephone network. His father encouraged Wilkes's interest in electronics. At King Edward VI's Grammar School in Stourbridge, Wilkes built a radio transmitter, which his father allowed him to use at home. Wilkes suffered from recurring bouts of asthma, but by his teens he was vigorously studying mathematics and science, increasing his knowledge of amateur radio from a subscription to Wireless World.

When he went to St. John's College, Cambridge, in 1931, he studied mathematics but was a spare-time student of electronics and attended engineering lectures. He held the license for the students' amateur radio station and continued building radios on his vacations so he could contact ham operators throughout the world. After taking a first in mathematics, he earned a Ph.D. While at Cambridge, Wilkes was familiar with Alan Turing (they were exact contemporaries although at different colleges), whom he later studiously avoided describing as someone he disliked by saying that “you don't regard your contemporaries as great men”; Turing was, according to Wilkes, “not in any sense a team leader; he didn't know how to get things done.” Wilkes also worked on the propagation of radio waves in the ionosphere and became interested in tidal motion in the atmosphere. He would publish his first book, Oscillations of the Earth's Atmosphere, in 1949.

Life's Work

Wilkes's attended a lecture by mathematician Douglas Hartree on (human) computing and the differential analyzer, the wheel-and-disk apparatus that solved differential equations. He built a model using parts from Meccano (a maker of toy construction sets). In 1937, Wilkes became a university “demonstrator” (essentially an associate professor) in the newly established Mathematical Laboratory (colloquially called the Maths Lab and ultimately transformed into the Computer Laboratory), located in the former anatomy school. His job was to acquire a differential analyzer to provide calculating services to university researchers. He had to suspend this academic position in 1939, however, when he was called to serve in World War II. During the war, he worked on radar with Robert Watson-Watt and John D. Cockroft and later on submarine tracking and missile and aircraft design.

After the war, Wilkes returned to head the Cambridge Maths Lab. There, his primary assignment was to explore the potential of calculating machines. In 1946, he was invited to attend a course on theory and techniques for design of electronic digital computers at the Moore School of Electrical Engineering at the University of Pennsylvania, where much of the cutting-edge work in computers was being conducted at the time: Moore was home to the J. Presper Eckert and John Mauchly's Electronic Numerical Integrator and Computer (ENIAC). Although postwar transportation was unreliable and his arrival in Philadelphia delayed, Wilkes attended the latter portion of the two-month course and heard John von Neumann's report on the Electronic Discrete Variable Automatic Computer (EDVAC), which proposed delay lines as a means of storing data in computer memory. Returning to Cambridge, Wilkes began working on what would become ENIAC's successor, and in tribute to EDVAC, which was not yet built, called it the Early Delay Storage Automatic Calculator (EDSAC). Von Neumann and others had noted that computers of the future had to have memory capable of storing not only data but also programs, sets of instructions for using the data. The future machine envisioned by von Neumann would allow users to change programs written in binary numbers without having to rewire and rejigger, a tedious and time-consuming process that took days. Three groups of scientists were working simultaneously on solutions to von Neumann's challenge: a team at Manchester University, Eckert and Mauchly in Pennsylvania, and Wilkes at Cambridge.

EDSAC debuted on May 6, 1949, computing a table of square numbers. EDSAC was a room-sized machine that used three thousand vacuum tubes arrayed on twelve racks and whose memory, like that of Eckert and Mauchly's EDVAC, was constructed of sealed tubes of mercury (mercury delay lines) that represented data as ripples in the mercury. EDSAC used thirty-two tubes of mercury, each of which had thirty-two seventeen-bit words. Delayed pulses generated by an electrically charged quartz crystal created this memory. An electrical signal became a sound wave passing through a long tube of mercury at 1,450 meters per second, and the signal was capable of traveling back and forth along the tube. Combined, these charged tubes formed an internal memory of 1,024 words. In this machine, the first load of the program was by paper tape but later uses of the program were from memory.

EDSAC performed 650 to 700 operations per second. Before EDSAC, computers such as the Moore School's ENIAC could handle only one type of problem at a time and had to be reprogrammed over several days of resetting thousands of switches and rerouting miles of cable. Although Eckert and Mauchly's computer employed a stored program before EDSAC, when it became operational, EDSAC was the first electronic stored-program computer to go into regular service (Manchester's had been the first electronic stored-program computer, but it was a small prototype). EDSAC may have been slow by today's standards, but it radically decreased the time needed to do computations, and as a practical advance, was dramatic. Moreover, its status as a practical tool for students, faculty, and researchers encouraged the creation of programs and means of using EDSAC that advanced computer programming.

EDSAC quickly, from early 1950, provided the university with an advanced computing service, the first regular computing service in the world, and Wilkes and staff developed programs and collected them into a library. R. A. Fisher's paper on genetics was the first scientific paper to use computer calculations, and the calculations were enabled by EDSAC. Another highly successful British computer, the LEO, created by Joe Lyons, was a “son” of EDSAC. In 1951, Wilkes and two graduate students, David Wheeler and Stanley Gill, published The Preparation of Programs for an Electronic Digital Computer (1951), the first book on computer programming, in which they described this use of a library of subroutines in programming.

Wilkes was also one of the first to realize that programming required such exacting detail that the future would inevitably include a lot of debugging to find errors that kept the programs from running properly. He thus spent a considerable amount of his time fixing mistakes in his programs during EDSAC's lifetime. That life ended when it became apparent that delay lines were on the verge of obsolescence and the future lay with magnetic storage. EDSAC was scrapped, its parts cannibalized, and its paper tapes used as streamers for children's parties.

EDSAC II debuted in 1958 with both magnetic storage and the first microprogrammed control unit. It used two magnetic core memories instead of delay lines, and it was faster and easier to program. The first “bit-slice” machine, its processor was constructed of interchangeable modules. Wilkes also realized that stored programs in the computer could represent control signal sequences, a concept he named microprogramming. Microprogramming is a method for implementing the control logic of a computer's central processing unit, essentially managing the construction of programs via access to low-level microinstructions from a control store of microinstructions, each of which provides appropriate control signals and sequencing information.

In the 1960s, Wilkes began to plan the next generation of EDSAC, but by then the computer industry was growing. Wilkes had a small government grant to spend on upgrading the Maths Lab's computer hardware. Wilkes considered machines from two companies, IBM and Ferranti. Ferranti's Atlas proved to be the best and most cost-effective choice, and it arrived (as the Titan) in 1963. EDSAC II was closed in 1965 during an emotional ceremony.

In 1965, Wilkes received a chair as professor of computer technology (his own name for the title). In the same year, he published a paper on cache memory, the first of its kind, and also introduced time sharing, which had been created at Dartmouth College and developed at the Massachusetts Institute of Technology—thus introducing workstations and inputting via teletype machines rather than punch cards or paper tape (he would later publish Time Sharing Computer Systems in 1975). By 1970, the Maths Lab was divided into a computer services division and a computer science/research division.

Between 1965 and his retirement in 1980, Wilkes was professor of computing technology at Cambridge. He led what had become Cambridge University's Computer Laboratory to its status as a leading research center in England. He advised British computer companies and was a founder of the British Computer Society (over which he presided between 1957 and 1960). In 1974, he was among those who collaborated on the experimental Cambridge Ring, an early local-area network. The ring was a series of computers linked by digital communications for the purpose of sharing a printer, and commercial development was done by others.

After retirement, Wilkes moved to the United States and became a research consultant in multimedia conferencing and network systems at Digital Equipment Corporation (DEC). He also served at MIT as adjunct professor of electrical engineering and computer science from 1981 to 1985. He returned to Cambridge as a consultant researcher with funding from AT&T, Oracle, and Olivetti at various times and remained active into his nineties. In 2002, after AT&T's acquisition and shut-down of Olivetti, he returned to the Computer Lab at Cambridge.

Personal Life

In 1947, Wilkes married Nina Twyman. They had two daughters and a son. He published Memoirs of a Computer Pioneer in 1985. After returning to Cambridge in 2002, he enjoyed discussions with students and colleagues and regular academic social occasions. Upon the sixtieth anniversary of EDSAC, on May 6, 2009, Wilkes gave a speech during a celebration of the event in which he noted the achievement of the EDSAC project as having opened computing to a broad range of students and researchers—highlighting his overarching focus and interests in the practical applications of computing and the ways it contributed to the success of great scientists, including Nobel laureates. His comments underscored the often little recognized contribution of Wilkes himself: not only as a computer scientist but also as a manager, teacher, and facilitator of research. He died at the age of ninety-seven, predeceased by his wife.

Wilkes was the recipient of many honors and awards over the course of his career. He became a Fellow of the Royal Society in 1956, a foreign honorary member of the American Academy of Arts and Sciences in 1974, a Fellow the Royal Academy of Engineering in 1976, and foreign associate of the American National Academy of Engineering in 1977. He received the second A. M. Turing Award from the Association for Computing Machinery (ACM) in 1967, the Faraday Medal from the Institution of Electrical Engineers in 1981, and the Harry Goode Memorial Award of the American Federation of Information Processing Societies in 1968. He was knighted in 2000.

Bibliography

“Campbell-Kelly, Martin." A.M. Turing Award, 2019, amturing.acm.org/award‗winners/wilkes‗1001395.cfm.  Accessed 14 Oct. 2024.

Hartley, David. "Sir Maurice Wilkes: The Man and His Machine." Computer History Museum. 11 May 2011, computerhistory.org/events/sir-maurice-wilkes-man-his/. Accessed 14 Oct. 2024.

Hoffmann, Leah. “Maurice Wilkes: The Last Pioneer.” Communications of the ACM 54.2 (2011): 19. Print.

Quested, Tony. “Cambridge Technology Cluster Thriving Thanks to University Dynamism.” Business Weekly 24 Feb. 2012, www2.mrc-lmb.cam.ac.uk/archive/articles/Cambridge‗technology‗cluster.pdf. Accessed 14 Oct. 2024.

Sabbagh, Dan. “Maurice Wilkes.” The Computer Bulletin 41.3 (1999): 18–19. Print.

Schofield, Jack. “Sir Maurice Wilkes Obituary." Guardian, 30 Nov. 2010, www.theguardian.com/technology/2010/nov/30/sir-maurice-wilkes-obituary.  Accessed 14 Oct. 2024.

Warford, J. Stanley. Computer Systems. Sudbury: Jones and Bartlett, 2009. Print.

Wilkes, Maurice. Memoirs of a Computer Pioneer. Cambridge: MIT, 1995. Print.