Information Technology Reference
In-Depth Information
silicon valley
In the
s Bell Labs engineer William Shockley returned to his
hometown of Palo Alto to set up a company to exploit the invention
of transistors. Invented soon after the War, the transistor did the
electronic switching work of the valve far more efficiently, as well as
being far smaller and consuming far less electricity. It was almost
immediately adopted by producers of consumer electronics, such as
televisions and radios, and by the late
1950
s was starting to be used
in computers. Shockley sited his firm on the West Coast not just out
of sentiment. By the
1950
s, partly because of the presence of Stanford
University, the area just south of San Francisco was already starting
to become a centre for the microelectronics industry. Unfortunately
for Shockley his management skills were unequal to his engineering
brilliance, and within three years eight of his most important
engineers, including his most gifted employee, Robert Noyce, had
left to start their own company, Fairchild Semiconductors. While
they concentrated on producing transistors to make money,
Fairchild also put a lot of research into trying to solve the problems
inherent in the technology. Though an advance over valves, transis-
tors still presented many problems, especially in relation to military
uses. Some of the systems used for guidance in military planes had
up to
1950
transistors, and, with the kinds of usage associated
with such systems, the connections between transistors often came
loose. Furthermore, the more complex machines became and the
more wiring they involved the slower they became. The solution
to these problems became a priority of military computing research.
Simultaneously with another company, Texas Instruments, they
hit on the means to solve these problems. Both the engineers at
Fairchild and Jack Kilby at Texas Instruments realized that the
solution was to make the entire circuit out of one block of a semi-
conducting material, such as silicon or germanium, preferably all
at once. This idea, coupled with the notion of making the circuits
20
,
000
