Skip to main content
main-content
Top

About this book

This unique book presents the story of the pioneering manufacturing company Ferranti Ltd. – producer of the first commercially-available computers – and of the nine end-user organisations who purchased these machines with government help in the period 1951 to 1957. The text presents personal reminiscences from many of the diverse engineers, programmers and marketing staff who contributed to this important episode in the emergence of modern computers, further illustrated by numerous historical photographs. Considerable technical details are also supplied in the appendices.

Topics and features: provides the historical background to the Ferranti Mark I, including the contributions of von Neumann and Turing, and the prototype known as The Baby; describes the transfer of technologies from academia to industry and the establishment of Ferranti’s computer production resources; reviews Ferranti’s efforts to adapt their computers for sale to business and commercial markets, and to introduce competitive new products; covers the use of early Ferranti computers for defence applications in different government establishments in the UK, including GCHQ Cheltenham; discusses the installation and applications of Ferranti computers at universities in the UK, Canada, and Italy; presents the story of the purchase of a Ferranti Mark I* machine by the Amsterdam Laboratories of the Shell company; details the use of Ferranti Mark I* computers in the UK’s aerospace industry and compares this with the American scene; relates the saga of Ferranti’s journey from its initial success as the first and largest British computer manufacturer to its decline and eventual bankruptcy.

This highly readable text/reference will greatly appeal to professionals interested in the practical development of early computers, as well as to specialists in computer history seeking technical material not readily available elsewhere. The educated general reader will also find much to enjoy in the photographs and personal anecdotes that provide an accessible insight into the early days of computing.

Table of Contents

Frontmatter

Chapter 1. The Small Seeds of Innovation

Abstract
What was it like in post-war Britain, before the arrival of the modern computer? How, in particular, did organisations carry out calculations and tabulations, analyses and predictions, before high-speed electronics came to their aid? When, in 1948, a small university research group came up with ideas for a new kind of computing machine, why did the government get excited and give the Ferranti engineering company a contract to turn the university prototype into a large production computer capable of solving problems of strategic importance to the nation? In this chapter we describe the seeds of innovation that bore fruit in the Manchester area, starting with a lively interaction between a small group of academics and an unsuspecting electrical engineering company founded seventy years previously. On the way, we cover the history of the Ferranti company and introduce the Ministry of Supply as the government organisation most interested in helping Ferranti take an early lead in computer manufacture.
Simon Lavington

Chapter 2. Academic/Industrial Collaboration: From Chorlton-on-Medlock to Moston, and Back

Abstract
Although computer hardware and software is beginning to come to the fore in our story, it is more interesting to start by describing the human background: the people and the places. Who are the main players? What are their hopes and fears? How does Ferranti set about manufacturing a very large and complicated piece of equipment at its Moston factory? How does Manchester University, in the quaintly-named Chorlton-on-Medlock district, collect suitable staff together in a new, purpose-built laboratory to which the world’s first production computer, the Ferranti Mark I, can be delivered and put to use? There are teething troubles to be overcome but by the end of 1951 end-users from far and wide were coming to Manchester to use the giant machine to solve problems in science and engineering. Was it all serious stuff? No, there were light-hearted moments which, whilst not computer games as now understood, nevertheless provided some amusement.
Simon Lavington

Chapter 3. Canada Calling: Toronto Gets a Mark I

Abstract
Whilst the first computer was being put to use in Manchester, Ferranti’s Moston factory was constructing a second Mark I machine—encouraged by a Letter of Intent from the Ministry of Supply (MOS). In the event, the MOS delayed placing a firm order. Meanwhile, the Canadian National Research Council was keen to install a computer at Toronto University so the sale of the second production Ferranti Mark I was agreed. The computer was known locally as FERUT. This gives us the opportunity to recall the pioneering activity, both hardware and software, in Canada and the interesting links that developed between Toronto and Manchester. Luminaries such as Christopher Strachey and Trixie Worsley shine out. We tell the story of how they wrote the huge programs that lay behind the construction of the St Lawrence Seaway which allowed large ocean-going vessels to reach the Great Lakes. FERUT then had a long and useful life, which is described in this chapter.
Simon Lavington

Chapter 4. A Star Is Born: Ideas and Upgrades

Abstract
Returning to Moston, we describe Ferranti’s efforts to recruit a team of programmers and house them in a special prefabricated building known affectionately as the Tin Hut. What is remarkable about Ferranti’s team of programmers was its size, skill and its high proportion of females—in these respects out-shining competitor groups elsewhere in the UK. One of the wisest people in the Tin Hut was the Australian John Bennett. He suggested some desirable enhancements to the instruction set of the Ferranti Mark I. After much debate, the upgraded version known as the Mark I* (Mark One Star) was launched. Thereafter, the remaining seven members of the Mark I family to be built were of the Star variety. Meanwhile, down in London a group of engineers who had left a rival manufacturer had persuaded Ferranti Ltd. to embark upon the design of a smaller and cheaper computer to be called Pegasus. When this finally entered the market in 1956 it was to prove somewhat of a rival to the Ferranti Mark I*.
Simon Lavington

Chapter 5. Into the Market

Abstract
The Ferranti sales staff had been making efforts to sell the Mark I* computer, encouraged by the National Research Development Corporation (NRDC). This government body went as far as to underwrite the company’s manufacturing costs, then taking a royalty out of eventual sales. Notwithstanding the efforts of Ferranti and NRDC to widen the market, it was actually three of the government’s own defence establishments that took the early deliveries. Then followed the Amsterdam Laboratories of the Shell company and the Instituto Nazionale per le Applicazioni Calcolo in Rome, to each of which a Ferranti Mark I* was shipped. The aerospace companies A. V. Roe (Avro) and Armstrong Siddeley Aero Engines completed the deliveries of the seven Mark I* machines to be built. But by the mid-1950s other UK computer manufacturers had begun to sell competitor products and Ferranti faced competition. Ferranti therefore created its London Computer Centre as a more effective marketing base than Manchester and embarked upon the design of new computers. We will return to Ferranti’s fortunes in Chap. 13. Before then, it is necessary in Chaps. 612 to describe the life and times of the seven Mark I* computers. We start with the aerospace sector.
Simon Lavington

Chapter 6. The AVRO Mark I* Installation at Chadderton

Abstract
The illustrious background of A. V. Roe, the company that produced the Avro Lancaster bomber, is described. Although the company was based at Chadderton, just a couple of miles from the Moston birthplace of the Ferranti Mark I*, Avro was at first reluctant to take the plunge and order its own computer. Once the machine arrived, however, full and innovative use was made of its computing power. We describe the applications and the people who were involved, the company’s efforts in 1964 to donate its Mark I* to a museum, and the computing resources that followed the Mark I* at Chadderton and at Avro’s nearby plant at Woodford.
Simon Lavington

Chapter 7. The Mark I* at Armstrong Siddeley, Ansty, Coventry

Abstract
Armstrong Siddeley’s entry to the world of digital computers was more hesitant than Avro’s. Fortunately some of the original players and their documentation survives, which has enabled us to describe the decision-processes, the installation and maintenance efforts, the activity of the mathematicians and programmers and the applications to which the Ferranti Mark I* was put. Then in 1959 Armstrong Siddeley had the chance to purchase another Mark I*, namely the one that had previously been installed at GCHQ Cheltenham. So the two computers worked side-by-side at the Ansty premises until 1964. We end by recalling the company’s computing arrangements that followed amalgamation with Bristol Aero Engines.
Simon Lavington

Chapter 8. The Ferranti Mark I* Installation in Amsterdam

Abstract
Shell’s Amsterdam Laboratory was the first organisation in the world to purchase a computer as a commercial venture, without direct or indirect government support. Shell’s energetic and innovative computing activities are described, enlivened by the personal memories and photographs of those who were involved. The Mark I* computer was named MIRACLE by Shell. MIRACLE stood for Mokum’s Industrial Research Automatic Calculator for Laboratory and Engineering, Mokum being a Yiddish word for ‘city’ and the local name by which Amsterdam is sometimes known. Amongst the MIRACLE anecdotes is an account of the emotional final switch-off of the computer just before Christmas 1961. The novelist Gerrit Krol was present and we have his poetic observations of the assembled staff as the giant machine fell silent and the console lights went out.
Simon Lavington

Chapter 9. The Ferranti Mark I* Installation in Rome

Abstract
This computer was purchased with funds from Azienda Rilievo Alienazione Residuati (ARAR), an Italian government entity created in October 1945 for “… the recovery, custody and disposition of material remnants of war, left behind by the Allies or abandoned by the Germans in Italy or otherwise acquired”. It was installed in the Consiglio Nazionale delle Ricerche (CNR) building in central Rome, the inauguration ceremony being attended by the Italian Head of State, President Giovanni Gronchi. The Ferranti Mark I* in Rome was called FINAC, indicating that it was acquired and run by the Instituto Nazionale per le Applicazioni Calcolo (INAC). FINAC had a long and useful life—so much so that when it came to seeking a replacement computer, INAC in partnership with Olivetti built CINAC, a powerful successor machine with a special console giving the look and feel of FINAC. This allowed existing programs to be run directly through a simulator. CINAC replaced FINAC in 1967.
Simon Lavington

Chapter 10. GCHQ Cheltenham’s Mark I*

Abstract
The next three chapters tell the story of three government defence establishments that acquired Ferranti computers. First in line for delivery, though curiously not first to become actively interested, was GCHQ, the cryptanalytical successor to Bletchley Park. Indeed, GCHQ continued to use its so-called Rapid Analytical Machines such as Colossus and several in-house successors until the late 1950s because they had a specialised performance far superior to that of the early general-purpose computers. We compare the progress of British and American code-cracking organisations in acquiring and using general-purpose computers and the relationship between GCHQ and Ferranti Ltd. The Mark I* finally installed at GCHQ’s Cheltenham headquarters in September 1953 was called CLEOPATRA. Within the limits of recently-released classified documents, we describe the installation, modification and use of CLEOPATRA.
Simon Lavington

Chapter 11. The Mark I* at the Armaments Research Development Establishment, Fort Halstead

Abstract
Fort Halstead in Kent was the first defence establishment to become interested in acquiring a computer, primarily because the UK’s early research into nuclear weapons was carried out there. Ministry of Supply (MOS) priorities were then re-adjusted so that GCHQ got the first Mark I* and the nuclear weapons calculations were transferred temporarily to the Ferranti Mark I at Manchester University. Fort Halstead got its Mark I*, called AMOS, in the summer of 1954. We describe the applications of AMOS over its long and useful life—a life so useful that Fort Halstead decided to build a successor computer in-house that provided a programmers’ interface that had the look and feel of AMOS. The result was COSMOS, about 50 times faster than AMOS. During January 1967 all work was smoothly transferred to COSMOS and AMOS was switched off.
Simon Lavington

Chapter 12. The Mark I* at the Atomic Weapons Research Establishment, Aldermaston

Abstract
It took some time to convert a former RAF aerodrome at Aldermaston, Berkshire, into a top-secret research establishment, so that Aldermaston did not get its own Mark I* computer until early 1955. Providing more of the latest computing equipment at Aldermaston had, by 1957, become a high priority so that, of all the UK’s sites, Aldermaston led the way with the early installation of top-of-the-range IBM computers whose performance out-classed the Ferranti Mark I* and anything that other British suppliers could match. We describe the successive upgrades, placing them in the context of available British computers, up to the installation of Ferranti’s own supercomputer, Atlas, towards the end of 1964. It was in the provision of software, particularly high-performance Fortran compilers, that Alick Glenny and colleagues at Aldermaston made significant advances in the decade 1957–67. Alick was also the ‘father’ of Autocodes, as we describe in Chap. 13.
Simon Lavington

Chapter 13. What Came Next?

Abstract
We pick up Chap. 2’s story of the first Ferranti Mark I at Manchester University. In 1954 this machine was moved to a large new Department of Electrical Engineering, to which end-users such as the Meteorological Office came to run their programs. Meanwhile, collaboration between the University and Ferranti continued in the design of new high-performance computers and their software—notably Autocodes. The high point in this collaboration was the Ferranti Atlas which, in 1962, briefly became the most powerful computer in the world. We describe the subsequent fortunes of Ferranti Ltd. and the eventual demise of the company’s computer manufacturing endeavours. We end with a review of surviving artefacts from the nine Mark I and Mark I* computers, indicating the museums where bit can be found today. Finally, there are plans for software simulators.
Simon Lavington

Chapter 14. Appendix A. Baby’s Conception: The Back Story

Abstract
What ideas, and what people, lay behind the design of the small Manchester University computer called Baby that first ran a program on 21st June 1948? The story begins in the autumn of 1945 and embraces influences from Bletchley Park mathematicians, from John von Neuman’s group at Princeton University, from engineers at the government’s Telecommunications Research Establishment and from Alan Turing’s activities at the National Physical Laboratory. Some of these influences were stronger than others but none had quite as much practical significance as the seminal work of two engineers, F. C. Williams and Tom Kilburn, who devised the first practical RAM—a random-access memory capable of working at electronic speeds. In this chapter we analyse all the influences chronologically, enabling us to show that the Manchester computer’s general architecture was close to that of the machine being designed at Princeton University but that the detailed hardware implementation was unique to Manchester. From this small beginning grew the family of Ferranti Mark I and Mark I* computers.
Simon Lavington

Chapter 15. Appendix B. Mark I and Mark I* Software Details

Abstract
To modern eyes, the programming conventions of the early Ferranti computers seem both primitive and complex, the complexity arising from the fact that coding was based on the 5-bit teleprinter characters. Users were expected to learn the 5-digit binary equivalents by heart, with the added complication that ‘backwards binary’ was employed with the least-significant digit at the left-hand end so as to be compatible with the engineer’s serial waveforms where time flowed from left to right. All this, and much more, is explained in this chapter—to the point where the bold reader can write a small program. The bootstrapping, routine-calling and monitoring processes are also described, along with the scheme for memory management, the integration of the small but fast primary store with the slower but larger magnetic drum secondary store. We detail the differences between the instruction sets for the Mark I and for the Mark I*—the latter being much easier to use.
Simon Lavington

Chapter 16. Appendix C. Mark I and Mark I* Hardware Details

Abstract
At the heart of the Mark I and Mark I* computers lay a memory system based on Williams/Kilburn tubes. These were Cathode Ray Tubes (CRTs) equipped with the means for reading and writing binary information as small areas of electrostatically-charged phosphor. Williams/Kilburn tubes were used for the primary store and all central registers and their characteristics determined the speed (i.e. clock-rate) of the entire machine. We describe how Williams/Kilburn storage worked. Specimen logic circuits based on thermionic tubes are then discussed, along with the magnetic drum backing store, input/output equipment and the computer’s power supplies. In short, we present an illustrated overview of all the hardware that went into the Ferranti Mark I and Mark I* computers.
Simon Lavington

Chapter 17. Appendix D. Naming Names

Abstract
This book has focussed on the computer-manufacturing fortunes of one company. It would not be complete without a mention of all those Ferranti employees who were involved in the design, implementation, use and marketing of the hardware and software of the Mark I/Mark I* family from 1948 to 1958. With the paucity of surviving company records, this is an almost impossible task. We do our best, however, mainly via the memories, anecdotes and photographs of former Ferranti staff. Their collaborative colleagues at Manchester University are also listed, though in the case of the academics there are staff records, theses and published papers that make the task of naming names comparatively easy.
Simon Lavington

Chapter 18. Appendix E. Performance, Cost and Delivery Details of Other Computers

Abstract
Although from 1951 to 1954 Ferranti had, in effect, no competitors in its efforts to sell computers outside the North American continent, from 1955 the market place saw increasing numbers of British and American companies vying for position. In this chapter we list the computing products of relevant competitors, their characteristics and cost, the dates when they first appeared and, for the UK products, the destinations of the first few examples to be delivered. The period of interest, from Ferranti’s view, is the critical window of opportunity from 1955 to 1958—as explained in Chap. 5. Finally, in a separate part of this chapter, we review the performance and dates of the British and American supercomputers that began to appear at the end of the 1950s and the start of the 1960s, spurred on by the high-performance computing demands of the nuclear weapons researchers at places such as Aldermaston and Los Alamos. This section clearly gives added background to Chap. 12.
Simon Lavington

Backmatter

Additional information

Premium Partner

    Image Credits