Beginnings of a revolution - The Computing Universe: A Journey through a Revolution - page 10

After many difficulties, they ultimately succeeded in designing and building
the famous UNIVAC (UNIVersal Automatic Computer) machine. With the war
ended, von Neumann returned to Princeton and wasted no time getting funds
to build an EDVAC architecture computer for the Institute for Advanced Study
(IAS). He quickly recruited Goldstine and Arthur Burks from the EDVAC team
and a talented engineer, Julian Bigelow, to help him design the IAS machine
(see Timeline). In 1947, with Goldstine, von Neumann wrote the first textbook
on software engineering called Planning and Coding Problems for an Electronic
Computing Instrument .
While commercial interest in computers was beginning to develop in the
United States, it was actually two teams in the United Kingdom that first dem-
onstrated the viability of the stored-program computer. At Manchester, Freddie
Williams and Tom Kilburn had followed the path outlined by von Neumann
and in June 1948 they had a prototype machine they called Baby (see Timeline
and Fig. 1.14 ). This ran the first stored program on an electronic computer
on 21 June 1948. This success was followed in May 1949 by Wilkes's EDSAC
machine in Cambridge - which was undoubtedly the first stored-program com-
puter with any significant computational power.

Fig. 1.14 Tom Kilburn and Freddie
Williams with the “Baby” computer
in Manchester. The machine had only
seven instructions and had 32 × 32 bits
of main memory implemented using a
cathode ray tube.

Key concepts

Computation can be automated

Layers and abstractions

The stored program principle

Separation of storage and processing

Von Neumann architecture

Cartoon illustrating the requirement for calculating shell trajectories.