Behind the Giant Brains (Part 2)
February 1957 Radio & Television News Article
Television News ran a two-part article on the state of the art of computers in the late 1950s (this is part 2).
It had only been since ENIAC's (Electronic Numerical Integrator And Computer) debut in 1946 at Massachusetts Institute
of Technology (MIT) that the public (or science community for that matter) was getting used to regularly hearing
about computers in the news. By 1957 there were many companies popping up with electronic computer offerings. Originally
the exclusive purview of university research labs and defense installations, the size and cost of computers was
moving into the realm of affordability by corporations that used them for accounting and bookkeeping, and in some
cases even rented idle time to outside users. Desktop PCs and notebook computers were still the realm of crazy dreamers.
of Contents]These articles are scanned and OCRed from old editions of the Radio & Television News magazine.
Here is a list of the Radio & Television News articles
I have already posted. As time permits, I will be glad to scan articles for you. All copyrights (if any) are hereby
See all available vintage Radio News articles.
Behind the Giant Brains
Part 2. Advantages and characteristics of electronic computers, along with a general survey of the field and
some predictions of the future effects on us all.
Last month we discussed the development of the computer and its basic functions. Now we proceed with the advantages
and a survey of the field.
Typical of the 700 series of scientific and commercial computing systems now being produced by International
Business Machines is the IBM 701, which was also the first of the series. Primarily a scientific computer (the
commercial members of the family are the 702 and 705), the 701 is extremely fast in computing speed, but limited
in flexibility. since it depends primarily on punched cards for input.
The reason computers are valuable, and increasingly more valuable, in our time,
has nothing to do with their innate ability, which is extremely limited. It has to do with their speed, which is
fantastic, and their relative reliability.
A man figuring out his own pay, for example, might spend five
minutes or more multiplying rates by times, computing and subtracting deductions, and finally arriving at his salary
figure. A desk calculator permits him to reduce this to about two minutes. A punched-card calculator could make
the same computations in a few seconds. An electronic computer like "Univac" could do the same work in less than
a tenth of a second.
The man may tire of doing this work in an hour; certainly by the end of a day most
men would be bored beyond distraction. Industry has discovered that most people cannot perform such repetitive work
for more than three or four hours at a stretch without a sharp rise in the incidence of error. Machines, whether
mechanical or electronic, never tire, never need to break for coffee or lunch, and never get bored; but after a
few thousand operations, mechanical parts begin to wear. Electronic tubes can be slammed from cut-off to saturation
millions of times a second (as they are in many electronic computers), and still operate for months without fatigue;
new developments in electronics indicate even higher orders of efficiency. Solid-state circuitry, such as the transistor
and Sperry Rand's still-newer "Ferractor," seem capable of almost limitless operation without fatigue. So, just
as a mechanical device is better than a human being for repetitive tasks, an electronic device is usually better
than a mechanical one; it lasts relatively longer at a high operating efficiency.
The first requirement for any job that is to be done by a computer is that it be capable of precise description.
Or, as Dr. John Mauchly, co-inventor of ENIAC and "Univac," once remarked, "any activity that can be precisely described
can be done by a machine. You're already well on your way to designing the machine in formulating the description."
If anyone could analyze and define the complex operation we know as thinking, for example, the engineers could develop
a machine to do it for us. Until that time, the name "thinking machine," or the concept of machines that think,
merely wastes time.
The Burrouqhs-built UDEC (Unitized Digital Electronic Computer), shown in the Wayne University's Computation
Laboratory in Detroit. This computer, also basically a scientific computer, is much used by the automotive industry
for engineering problems.
Putting a job on a computer first requires a complete analysis of the job, and an exact
definition of its scope. Then every step that a human being performs in doing the job, every decision he makes,
every value he weighs, must be translated into terms that the computer can recognize, steps it can perform, or values
it can weigh. A complete listing of the step-by-step instructions, recognizable to the computer's blind and dumb
hardware, must be drawn up. Then the computer can do the job.
This operation, consisting of the analysis
of the problem and the synthesis of the instruction routine, is known as programming. As is perfectly clear, the
computer's decisions are really the programmer's decisions; its criteria for evaluation are given to it by the programmer.
Drawing up such a program can be a costly and time-consuming job. That is why advanced programmers concentrate
some of their efforts on a technique called automatic coding, which gives the computer a library of simple, frequently
used, chunks of programs, and makes it collaborate in the formulation of its own program of instructions.
It is also the reason why "repeatability" is one of the major criteria for determining which jobs will be done
by a computer. The cost of making the program can be amortized if the job is to be repeated over and over again.
A company payroll, for example, which must be computed every week or two, is a far more likely candidate for mechanized
or electronic treatment than is the design of the earth satellite; but many of the myriad computations incident
to the satellite design are being done by computers simply because of time a computer can save.
The Field Today
There are still some companies - but their numbers are decreasing - who are reluctant
to submit their paperwork problems to electronic treatment, some because they do not trust the machines, others
because they are not convinced of the economies of electronics. That computers are economical when the job is big
enough is now an established fact. Evidence of the economies of electronic data-processing has been available ever
since the first "Univac" was bought by the Census Bureau more than five years ago. The evidence is growing daily.
And the acceptability of electronic record-keeping has even been tested - and accepted - in the courts.
The manufacture of electronic computers and computing systems has become a major industry, and a hotly competitive
one, too. Led by Remington Rand's "Univac," which, started in 1949, was the first production-designed business computer
on the market; and by International Business Machines, which has concentrated millions of dollars in the design
and production of its 700 series of computing systems, the industry has grown in very few years to become a giant.
Some fifty companies are now manufacturing complete electronic computer systems or major systems components such
as high-speed magnetic-tape units, magnetic-drum storage systems, instrumentation and data-presentation systems,
and so forth. Business machine manufacturers, such as Underwood, Burrouohs, National Cash Register, and Royal-McBee,
and electronics manufacturers, such as RCA, Philco, Raytheon, and General Electric, have all contributed to the
progress. And no one can overlook the contributions made by the Bell Telephone Laboratories in basic research and
logical design of information-handling systems.
Business and industry are gradually accepting the machines.
Not considering the countless analogue computers in use all over the world, and in Army, Navy, and Air Force fire-control
and missiles-control equipment, the big digital computers alone - million-dollar systems all-form an impressive
roster: Remington Rand's fifty-odd "Univacs"; International Business Machines Corporation's seventy-odd 701's, 702's,
704's, and 705's; Burroughs Corporation's two UDEC's and a scattered shot of university, research center, and one-time
industrial designs. And within not too long we can expect to see RCA's "Bizmac": the "Datamatic 1000," being built
by the joint efforts of Minneapolis-Honeywell and Raytheon; and Remington Rand's much-heralded LARC, which was "commissioned"
by the Livermore (Calif.) Atomic Research facility.
This besides the increasing flow of small systems, such
as Burrouqhs' E101, National Cash's CRC series, Underwood's "Elecom" 50, Remington Rand's "Univac" 60 and 120, IBM's
CPC, 607, and 604; and the medium-sized systems, such as the "Univac" File-Computer, the"Elecom"125, IBM's 650,
Burrouqhs' "Datatron," and many more. These compact and efficient machines are bringing the advantages of electronic
data-processing to the small and medium-sized business. The market is ripe for the computers, and more companies
enter the field daily.
Communications is also becoming an increasingly important consideration, and Western Union and AT&T, on the
one hand, and the computer makers on the other have cooperated on a number of plans to facilitate the transmission
of data from place to place. These plans range from the conversion of data to telecommunications code and regular
transmission over ordinary telegraph wire, to the direct transmission of the very-high-speed computer codes over
special coaxial lines.
New markets for computing systems are being tapped by the medium-sized general-purpose computer, typified by
Underwood's Elecom File Processor. The Elecom system reduces the contents of 1600 conventional file drawers
to less than three cubic feet of space; savings in space such as this, added to time saving, plus the different
kinds and configurations of management information which computers provide, have given impetus to the furor
of interest in electronic computing.
Naturally, all the activity has strained the creative facilities of the small nucleus
of scientists and engineers who first launched the computer business such a short time ago. Every computer research
center in the country is straining at the seams, and every engineering staff is heavily over-burdened. Computer
engineers and experienced computer programmers have, within the last two years, discovered that they can practically
write their own tickets.
Computer research has borrowed heavily from every known science and technology,
and has managed to solve many of the most pressing problems. Frequently, however, each solved problem turns, hydra-like,
into a hundred more questions. The name mushrooming technology is apt: computer research has frequently tried to
grow in all directions simultaneously.
Industry and business now have heavy investments in the development
of newer and more capable electronic tools, and our economy is gradually accustoming itself to depending on them
more and more. Properly used, they can make life simpler and easier for all of us - and more rewarding, too, as
the time-consuming and deadly dull repetitive tasks which are part of so much of our commercial and industrial effort
are given over to the machines. Their growth was inevitable, because there were too many jobs to be done, and too
few manhours to do them in; without manpower, we must inevitably fall back on the machine. And whether for better
or for worse, automatic controls and electronic computers are with us to stay.
the Giant Brains Part 1 appeared last month)