n o i r s . o f a C o m p u t e r P i o n e e r W i l k e s $19.95 and long-sought breakthroughs against life's simpler pleasures and trials. His account brims with assessments and anec­ dotes of such contemporaries as Turing, Memoirs of a Hartree, von Neumann, Aiken, and a Computer Pioneer dozen others—and with his impressions of by Maurice Wilkes America and Germany formed during his scientific journeys. Maurice Willces was one of the leading sci­ Maurice Wilkes retired from his post at entific explorers in the development of the ' Cambridge University in 1980, when he modern digital computer. He directed the becfame a Senior Consulting Engineer at Mathematical Laboratory (later named the Digital Equipment Corporation in Massa­ Computer Laboratory) at Cambridge Uni­ chusetts and Adjunct Professor at MIT. versity,. where he and his team built the Memoirs of a Computer Pioneer is included EDSAC, the first stored-program digital in the History of Computing series, edited corhputer to go into service. by I. Bernard Cohen and William Aspray. • Wilkes describes in nontechnical detail the growth of EDSAC and its successor, EDSAC 2, his introduction of micropro­ gramming, and the first experiments with time-sharing systems. In the 1950s, when machines were still getting larger rather than smaller, Wilkes was one of the few who foresaw a time when nonspecialists would be using computers almost univer­ sally, and he reviews his anticipatory ef­ forts to develop simple programming sys­ tems. But his book is more than a history of computing; it also recounts the allied scientific effort when he was one of those scientists and engineers ("boffins" as they were called by the RAF) who were in the thick of it, his electronics skills enlisted in the new and exciting development of radar. In this absorbing autobiography, Wilkes is as concerned with people and places as he is with computer components and pro­ gams of development. He deftly sketches his childhood in the English midlands and his student days at Cambridge, where he studied mathematical physics and his boy­ hood fascination with radio matured. He conveys the excitement of sudden insights Memoirs of a Computer Pioneer MIT Press Series in the History of Computing I. Bernard Cohen, editor; William Aspray, associate editor Editorial Board: Bernard Galler, University of Michigan, Ann Arbor, Michigan; J. A. N. Lee, Virginia Polytechnic Institute, Blacksburg, Virginia; Arthur Norberg, Charles Babbage Institute, Minneapolis, Minnesota; Brian Randell, University of Newcastle, Newcastle upon Tyne; Henry Tropp, Humboldt State College, Areata, California; Heinz Zemanek, Vienna, Austria Memories That Shaped an Industry, Emerson W. Pugh, 1984 The Computer Comes of Age, R. Moreau, 1984 Memoirs of a Computer Pioneer, Maurice V. Wilkes, 1985 Memoirs of a Computer Pioneer by Maurice \ Wilkes The MIT Press Cambridge, Massachusetts London, England © 1985 by The Massachusetts Institute of Technology All rights reserved. No part of this book may be reproduced in any form by any electronic or mechanical means (including photocopying, recording, or information storage and retrieval) without permission in writing from the publisher. This book was set in Baskerville by The MIT Press Computergraphics Department, using magnetic tapes supplied by the author, and printed and bound by Halliday Lithograph in the United States of America. Library of Congress Cataloging in Publication Data Wilkes, M. V. (Maurice Vincent) Memoirs of a computer pioneer. (MIT Press series in the history of computing) Bibliography: p. Includes index. 1. Computers—History. 2. Wilkes, M. V. (Maurice Vincent) 3. Engineers—Great Britain—Biography. I. Title. II. Series. QA76.17.W55 1985 001.64'09 85-6667 ISBN 0-262-23122-0 Contents Series Foreword vii 1 Shooldays 1 2 Undergraduate years 9 3 Post-graduate research 20 4 War 31 5 Experience on radar sites 40 6 Air Defence Experimental Establishment 54 7 Operational research 64 8 TRE 81 9 A trip to Germany 90 vi Contents 10 Post-war reconstruction 103 11 Atmospheric oscillations 111 12 The Moore School 116 13 The EDSA C 127 14 First steps in programming 143 15 Germany revisited 154 16 Computer progress in the United States 160 17 EDSA C 2 184 18 Can machines think and other topics 195 19 Computer progress 1955—80 208 Sources and acknowledgements 231 Index 233 Series Foreword Maurice V. Wilkes is known to the computer world for his pioneering breakthrough in constructing the first machine to go into service that was designed to embody the new concept of the stored program. This machine, EDSAC (Electronic Delay Storage Automatic Calculator), was constructed in the Mathematical Laboratory (later known as the Com­ puter Laboratory) of Cambridge University. It performed its first cal­ culation on 6 May 1949 and was in continuous use until 1958. Most critical historians of the computer do not consider any machine to be a “computer” unless it embodies the stored-program concept; the EDSAC is by this criterion the first fully operational computer. Ac­ cording to a statement made by Wilkes, the EDSAC “was designed according to the principles expounded by J. Presper Eckert, Jr., John W. Mauchly, and others at the summer school held in 1946 at the Moore School of Electrical Engineering” of the University of Penn­ sylvania, which Wilkes was “privileged to attend.” The EDSAC was the inspiration for one of the earliest commercial computers, the LEO, completed in 1951 by the Lyons Company, whose computer operations today form part of ICL. Wilkes had in mind three objectives from the beginning: (1) to show that a binary stored-program computer could be constructed and operated, (2) to make a start with a development of programming techniques, even then seen by him to be a subject of more than trivial content, and (3) to apply the techniques developed in a variety of application fields. Wilkes succeeded better than any other early pioneer in his second objective. According to his colleague Stanley Gill, Wilkes “led the first practical development of programming for stored-pro- gram machines including the first program library” of subroutines. The new programming techniques were described in a book published in 1951, by Wilkes, Gill, and David Wheeler, which served as the basic primer on the subject for more than a decade in Britain and the United States.1 Another area in which he was an important pioneer was a viii Series Foreword specific technique “to provide a systematic approach and an orderly approach to designing the control section of any computer system.” We are told that in this context, “the term ‘control’ is taken to mean the interpretation and execution of a machine instruction.” Wilkes is also known for later developments in machine-independent computing. In this connection he developed an elementary language for list­ processing known as WISP. He also contributed to the development and use of time-sharing systems. Fellow of the Royal Society (1956), Wilkes was the first president of the British Computer Society (1957-1960) and the first member from the United Kingdom of the Council of IFIP (1960—1963). He gave the ACM Turing Lecture in 1967 and received the AFIPS Harry Goode Award in 1968. Elected a Foreign Associate of the U.S. National Acad­ emy of Engineering in 1977, he was elected three years later to the U.S. National Academy of Sciences. After a distinguished career at Cambridge University Wilkes came to the United States in 1980 and joined Digital Equipment Corporation as a senior consulting engineer. Wilkes’s autobiography spans the early years of the development of the computer. It is a work of particular interest, furthermore, because it relates Wilkes’s experience during World War II as a “boffin,” that is, a scientist in military service. It is difficult to think of any other account that portrays this activity with the vivid qualities of the present book. This aspect is of significance, for, as Wilkes exemplifies and as others have observed, many of the early pioneers in the computer field drew heavily on their experience of electronics in wartime service. This book is part of a series devoted to the history of computers and data processing. Other volumes have dealt with, or will deal with, various aspects of the development of systems, hardware, and software, encompassing both general works and specialized monographs. Some of these may concentrate on a particular development, such as magnetic memory, or the technical history of an industrial company. This is the first work in the series of an autobiographical nature. It is certainly to be hoped that other pioneers will make their experiences available to us and that we will be able to publish additional critical and bio­ graphical accounts of those who have made the modem computer. I. Bernard Cohen, Editor William Aspray, Associate Editor1 1. See Maurice V. Wilkes, David J. Wheeler, and Stanley Gill, The Preparation of Programs for an Electronic Digital Computer, with special reference to the EDSAC and the use of a library of subroutines (Cambridge, Massachusetts: Addison-Wesley Press, Inc., 1951; reprinted by Tomash Publishers and The MIT Press, 1982).