"A Walk Through 'Visible Storage'", section 2 of 6,
by LEN SHUSTEK
From "CORE 2.3", a publication of
The Computer History Museum.
Forward,
Backword,
Table of Contents,
Docent Page
ELECTRONIC BEGINNINGS
It wasn't until the 1940s that electronic
devices we recognize as being similar to
modern computers began to appear. Here is a
small part of one of the first, the
ENIAC,
Photo by Michael Dubinsky
("Electronic Numerical Integrator and
Calculator"), designed during WWII at the University of Pennsylvania
to compute ballistics tables for the Army.
Unfortunately ENIAC, a room-sized
monstrosity with 18,000 vacuum tubes, was
finished too late to help with the war effort.
And, it wasn't really a computer in the modern
sense, because it didn't have a program
stored in memory that could be easily
changed.
The "stored program" breakthrough occurred
June 21, 1948 at the University of
Manchester on a test computer called "The
Baby" that at the time wasn't considered
important enough to preserve so it no longer
exists. But starting in 1949 and based on that
idea, true computers as we know them today
began to appear. The
Johnniac,
Photo by Michael Dubinsky
was one of
the first generation of computers in that modern design, and the
only one ever named for John Von Neumann,
the brilliant Hungarianborn mathematician who
played an important role in the invention of
the modern "stored program" computer. The
Johnniac was built by the Rand Corporation of
Santa Monica, California, and was an
approximate copy of the machine built under
Von Neumann's supervision at the Institute for
Advanced Studies in Princeton, New Jersey, It
wasn't exactly the same, but that was ok
because in those days the notion of
computers sharing program "software" (a term
not yet invented) was not an issue-if you had
a computer, you wrote programs specifically
for it and no other machine used them.
These new contraptions were clearly going to
be useful for many different things. But in the
early 1950s if you wanted a computer for,
say, calculating some physics equations for
your PhD dissertation, you had a problem.
Computers had been invented, but you
couldn't buy one. If you were determined
enough, like Gene Amdahl at the University of
Wisconsin, you simply built one for yourself.
This is his
WISC
Photo by Jessica Huynh
from 1952, the "Wisconsin
Integrally Synchronized Computer." In
the process Amdahl decided that building the
computer was more fun than doing the
physics, and he went on to design many
important computers that were manufactured
first by IBM and later by his own eponymous
company. But this early handcrafted WISC,
like many of the objects in our collection, is a
one-of-a-kind item. If you look closely you can
also see that it is the only
object in our collection that is perforated with
bullet holes, a punishment many of us have
wished but not dared to inflict on our own
computers. For the real story on the bullet
holes, visit the Museum and ask a docent.
MEMORY MAKES IT WORK
The biggest impediment to building computers
in the early 1950s was the lack of a good way
to store data-which was now both numbers
and programs. Early machines experimented
with a wide variety of bizarre schemes, from
vacuum tubes that conducted a current or
not, to CRT screens with spots of light and
dark, to this strange-looking
delay line from
the UNIVAC I,
Photo by Jessica Huynh
that
stored information as sound waves traveling
through metal tubes filled with liquid mercury.
The biggest impediment to
building computers in the early
1950s was the lack of a good
way to store data which was
now both numbers and
programs.
The 1953 breakthrough that caused
computers to flourish was the
magnetic core:
a small ferrite doughnut that could
be magnetized either clockwise ("zero") or
counter-clockwise ("one"). An Wang at
Harvard pioneered the use of core, and
Jay Forrester
at MIT made it
practical by inventing a matrix scheme using
two wires at right angles to read and write
individual cores without having a separate
wire for each one.
Magnetic core became the dominant computer
memory for 25 years until semiconductor
memories were invented. Forrester, who was
inducted as a Museum Fellow in 1995, decided
shortly after his invention that all the really
interesting problems in computer hardware
had been solved, and he moved on to other
fields where he made equally brilliant and
seminal contributions.
One of the first large computers that core
memory made possible was a huge system for
the military with the combat-speak name of
"Semi-Automatic Ground Environment" or
SAGE.
Photo by Michael Dubinsky
This photo
shows only a few of SAGE's 51,000 vacuum
tubes, every one of which had to be working
simultaneously in order for the computer to
work.
There were 46 SAGE
Photo by Michael Dubinsky
Photo by Michael Dubinsky
computers built. One
machine was installed in each of
23
underground bunkers located in the U.S.
and Canada, and each location had a hot-
standby backup waiting to take over. Their
purpose was to process radar data and detect
Russian piloted bombers coming over the
north pole toward the U.S. Despite all the
tubes, these machines were incredibly reliable
and were operated until the early 1980s. The
fact that by then Russia had long since
developed Intercontinental Ballistic Missiles
(ICBMs) and SAGE was not fast enough to
track them usefully didn't put them out of
business. Perhaps the Russians didn't know
SAGE's limitations.
go to top
Forward,
Backword,
Table of Contents,
Docent Page
|