Information Technology Reference
In-Depth Information
Neumann, the peripatetic physicist, mathematician, and economist who,
during the Second World War and its immediate aftermath, was inti-
mately involved with the development of both the electronic computer
and the hydrogen bomb. In the course of his work at Los Alamos on the
modeling of thermonuclear reactions, von Neumann became aware of
the ENIAC project at the University of Pennsylvania. There he began
working with the ENIAC designers on a successor machine called the
EDVAC, which was in concept the fi rst modern, stored-program elec-
tronic computer. In 1945-1946, von Neumann circulated an informal
“First Draft of a Report on the EDVAC,” which described the EDVAC
in terms of its logical structure, using notation borrowed from neuro-
physiology. Ignoring most of the physical details of the EDVAC design,
such as its vacuum tube circuitry, von Neumann focused instead on the
main functional units of the computer: its arithmetic unit, memory, and
input and output. The “von Neumann architecture,” as it came to be
known, served as the logical basis for almost all computers designed in
subsequent decades.
By abstracting the logical design of the digital computer from any
particular physical implementation, von Neumann took a crucial fi rst
step in the development of a modern theory of computation.
55
His was
not the only contribution; in 1937, for example, Turing had described,
for the purposes of demonstrating the
limits
of computation, what would
become known as the Universal Turing Machine. Eventually, the
Universal Turing Machine would become an even more fundamental
construct of modern computer science. According to the Church-Turing
thesis, fi rst articulated in 1943 by the mathematician Stephen Kleene,
any function that can be physically computed can be computed by a
Universal Turing Machine.
The abstraction of the technology of computing in the theoretical
construct of the Turing Machine mirrored the shift toward software that
was occurring in the larger commercial computing industry. Independent
of the work of theoretical computer scientists, working programmers—
and their corporate employers—were discovering to their chagrin that
computer software was even more complicated and expensive to develop
than computer hardware. It was the growing number of data processing
departments and commercial programming houses that provided the
majority of employment opportunities for the graduates of fl edgling
programs in computer science. Establishing computer science as a disci-
pline substantially different from computer engineering had been rela-
tively easy given the growing (and visible) distinction between software