United States. Congress. Senate. Committee on the.

The Industrial reorganization act. Hearings, Ninety-third Congress, first session [-Ninety-fourth Congress, first session], on S. 1167 (Volume pt. 7) online

. (page 8 of 140)
Online LibraryUnited States. Congress. Senate. Committee on theThe Industrial reorganization act. Hearings, Ninety-third Congress, first session [-Ninety-fourth Congress, first session], on S. 1167 (Volume pt. 7) → online text (page 8 of 140)
Font size
QR-code for this ebook


it will become available to people as customers and even the transfer
of funds between cities and between banks will be done by computers.

A large amount of educational potentiality exists in the computer
business where computers can be used for providing instant feedback
to the student, as shown in figure 87, and can be used for providing a
much more graphic display, and can go at the pace of the learner
rather than at the pace of the teacher.

I bring back these pictures of a system, figures 88 and 89, to sum-
marize briefly where we are. A computer is a remarkable versatile
machine which can sen-e the interests of mankind in a wide variety
of ways. Figures 90, 91, and 92 tell us that computers are very sophis-
ticated electronic devices, but remember they handle numbers and they
handle numbers only that people have given them ; and they handle
them in a way that people have instructed them to handle them. Com-
puters work extremely rapidly. Very much under the control of the
people who program them and operate them, although operating
systems are tending to take over the routine functions of the operator.
The operating system doesn't take over the responsibility of providing
a particular application to a particular computer. That's still done
bv people.

Essentially, the thing you can say about computers is that they re
used mostly" where large amounts of information must be processed



4880

quickly and when the processing can be described in an exceedingly
simple series of steps which can be very complicated, very long, and
are highly repetitious by virtue of the computer operating on its own
instructions; but they must be all thought through in advance by
people. These are the conditions under which computers are the most
valuable.

If we can have the lights now, I will show you some sample parts
and give you some feeling about the insides of the machines.

Some of the things I'd like to say about the future will come after
I show^ you a few of these pieces now. Take the whole package up.

These are the kinds of things in the early 1960's, late 1950's, there
were about three circuits in that. You had thousands of them, and you
plugged them in big boxes.

That's what made up the contents of those boxes. That's one tran-
sister. You know the size of a pea. It's a good sized piece of hardware.

That little thing inside there is the transistor, and a few years later
we began putting them in little packages like this, and there are maybe
30 to 40 circuits in this package.

This is probably a better example of it with them lined up in a row
like that. Then a few years after that, late 1960's, early 1970's, we
began producing them like this, with maybe 300 of them in there.
( You can just barely see the transistor, little hints of material inside
'.the box, and there may be 300 or 400 in that box.

The prices kept going up for the box but down for the average ele-
ment. This cost $10. This cost $200 when it was first built. They're now
about $25.

This cost $1,000 when it was first built. They are now down to
about $200, but an individual transistor went from $3 to $5 in this
array, doAvn to $1 in this array, or maybe 50 cents ; so the individual
packages cost more but the aggregate costs less per item.

Xow, in the latest technology these are what are called integrated
circuits. Underneath that little plastic chip there are 30, 40, or 50
circuits equivalent to this box. In that little plastic package, in fact,
most of the active part is in the little space of about one-eighth of an
inch square inside that plastic package.

Computers are now beina: built out of tliese things where the wiring
is printed mostly on these boards inside. For example, that is a whole
computer.

In Von Neumann's day, that was unthinkable. The computer was the
size of this whole desk up here. Memories are shrinking. This is an
example of a memory board in today's integrated circuited memory
technology.

Chips on this side contain the individual })its of memory, but there
may be 1,000 of them in there or 4,000. It's beginning to become stand-
ard to have 4,000 bits in that little chip.

Senator Hart. And the cost?

Mr. Parkix. The cost is going down per bit. but it still costs yon
a lot to put this all together.



4881

Here are some cores. Just to show you what they look like. It looks
like dust. Each one of those has a little hole in it.

It's unbelievable what they do. You have to have a magnifying glass
to see it.

Senator Hart. No. I can see it.

]Mr. Parkin. There are some various sizes and some samples. They
put three wires through those.

Senator Hart. The tiniest has a hole in it ?

Mr. Parkix. Yes. And you put three wires in that hole.

jSIr. LoRETTE. This is done by human manufacturers.

Mr. Parkin. We don't have machines to do tliat yet. Mostly people
in Hong Kong, Korea, or such places do this.

Senator Hart. Is any of it done in this country ?

Mr. Parkin. A good deal of it is done here.

These can be inserted in the board with a tool, and the whole board
is soldered by dipping in a bucket of molten solder by a machine.

This can all be assembled by a machine, and it makes the cost of
memory come down very drastically compared to stringing those on
wire.

It's not down yet far enough to compete totally with cores today, but
it will be down cheaper. It keeps getting cheaper all the time.

Let me make a few more remarks about that.

There are two very forceful trends in our industry. In the one case,
hardware is getting smaller and less expensive all the time, and it seems
to be doing so on a steady curve. It just does not stop being less ex-
pensive. If the cost of a certain kind of technology bottoms out, then
a new technology evolves which drives the cost down still further. And
that seems to be a continuing trend to which none of us can see any
stop, any end in sight. In fact, if you extrapolate it, it becomes clear
that at some point, with the cost of the hardware coming down con-
tinuously and the cost of people continuing to rise, which is the other
horn of our dilemma, and it is the software which is done by the people,
the programing, the preparation of data, the handling of instructions
which is causing the cost of computer systems to go up, it is clear
that at some point the computer will be the giveaway item, and the
programing will be what you charge for. In the past, it used to be quite
different. You charged for the hardware and gave away the program-
ing. In my estimation, the continuing expansion of the use of com-
puters is basically economically limited today. That is to say that com-
puters are such a capital-intensive item, from the point of view of the
consumer of the computer; that is, the industry that wants to use a
machine or the Government agency that wants to use a machine.

It is an expensive proposition. The user pays $1 million, or $2 mil-
lion, or $5 million for a big computer, and he pays $1 million, or $5
million, or $20 million, or $50 million for the programing to go around
that machine.

Those are big numbers Avhich must be thought through very care-
fully in making an investment. So in a sense, a computer is not an
ongoing expense like a desk calculator might be, or like a telephone.



4882

A computer is a capital expense at the time you install it for the
job it is going to do, as well as a:i ongoing expense for the operation
of it.

Thus, the spread of computers is limited by its competition for
the demands of other capital ; capital for a rolling mill is in compe-
tition with capital for a computer to control that rolling mill in the
same company that is going to be building it. So the computer be-
comes a major factor in the capital-intensive world, so that its spread
is really basically limited by the availability of capital.

But the spread of its usefulness, the things that computers can do,
are so enormous, the capabilities are so vast that there is no end
in sight to the number of kinds of things that computers will do and
can do for people.

It is purely a question of can we aiford the programing for those
installations, because with the cost of the hardware continuing to go
down, pretty soon it will be so inexpensive that the whole application
of computers will be strictly limited by the availability of people.

Indeed, it really is nearly that way today, but people don't quite
realize it. It is the lack of availability of trained manpower to do
programing and to do applications for computers that is limiting
their spread and their applicability to things of benefit to mankind.

I am prepared to answer as many questions as you would like to
ask. I could talk from 5 minutes to 5 days, or 5 years, on computers.

I have been deeply steeped in the technologj^ of computers for so
long that I'm really prepared to talk about it.

Senator Hart. Well, if I were a computer I would probably have
a lot of questions to ask. I am a little more comfortable after listen-
ing to the explanation and seeing the pictures.

I realize now I should have long ago taken up the invitation, to go
to a place where computers are assembled.

Mr. Parkix. You are more than welcome to come to Control Data
Corp. We would be pleased to have you and tour you through any
facility. It would be interesting to you or any member of your
subcommittee.

Mr. Xash. I have no questions, thank you.

Senator Hart. Mr. Chumbris ?

Mr. Chumbris. I have no questions either, thank you.

I read wnth interest and listened with interest to the charts you have
presented. Fortunately for me, Mr. Jerry Hellerman, Mr. Dave Martin,
and I went to some plants and watched them put some computers
together.

Still, there is much one can learn about computers. We very much
appreciate the testimony you haA'e given us this morning.

Mr. Parkin. Thank you very much.

Senator Hart. Mr. Lerette, did you have anything you wanted to
add ?

Mr. LoRETTE. No ; thank you.

Senator Hart. Well, again, thank you both.

[Tlie following was received for the record.]



4883

MATERIAL RELATING TO THE TEHTIMOXY OF THOMAS R. PARKIN

Exhibit 1. — FigitrcH and churis mcutinncd throughout the testimony of Mr.

Parkin





4884




A COMPUTER IS



A VERY HIGH SPEED CALCULATING MACHINE
WHICH CAN DO jMiTYPES OF MATHEMATICAL
COMPUTATIONS VERY RAPIDLY

THE MOST LITERAL-MINDED DEVICE YOU
CAN IMAGINE



4885



AN ELECTRONIC MACHINE ABLE
TO PERFORM



FIVE FUNCTIONS



SIMILAR TO MAN

• INPUT • CONTR

• STORAGE • OUTPU

• ARITHMETIC



INPUT ARITHMETIC



4886



MAN PERFORMS AN
INPUT FUNGTHW
BY THE SENSES



TASTE-SMELLTOUCH-SEEHEAR






m




4887



CARD CHARACTERS



iiiiiiiii



iiiiiini
III liiiiiii I '

I I

I, ■ I. ■ -I I-

11 i i

I I I I

I; I I I

|> I I I I'

I,' t IM I.'

Il l< I I

III I



I I



i I I



Mil <



nil



nil



4888




4889




4890




4891




^Mp%^




f if^ilo



4892



^



1




4893




DIGITAL COMPUTER INPUTS



• SYMBOLS

• MEASUREMENTS



/"\T'ljr!0



F«S-^®



40-927 O - pt. 7-5



4894



Man performs a

storage function

by memorizing

and recording



Fi8-*'




ELECTRICAL CIRCi:
MAGNETIC DRUMS
MAGNETIC CORES
MAGNETIC TAPES
PAPER TAPE



4895



STORAGE CONTAINS
NUMBERS



THAT REPRESENT

• NUMERICAL DATA

• LETTERS AND SYMBOLS

• INSTRUCTIONS FOR
THE COMPUTER



4896




4897




4898




4899





4900





4901



STORAGE HIERARCHY



MEDIA



M/UN MEMORY
EXTENDED



ACCESS TIME



.0CX)001
.00001



TAPE
CARDS



100.



man and machine

both can pform

an arithmetic m^m




/^^d' ^'




MODERN NUMBER SYSTEMS



DECIMAL


f J i r-M M ? 1 :



1




2

3




4
5




6

7





DECIMAL






1 1




/^//r



4903



DECIMAL NUMBER ARITHMETIC

ADO MULTIPLY


-


1 2 3 4 5 6 7 8 9








1
2
3

4

5

^


12 3 4 5 6 7 8 9




7

9


7


1




1


2


3






1











^ . ;. ,__


/J- ST


i "




4904



ADDING BINARY
NUMBERS



010
+ 011



^<9.^



COMPUTER INSTRUCTIONS
ARE WRITTEN BY THE PROGRAMMER



ARE EXECUTED IN LOGICAL SEQUENCE
INSTRUCT THE COMPUTER AS TO
(1) WHAT OPERATION TO PERFORM
— C2} WHERE THE DATA IS



/r,^.^/



4905




/^,f)-H^



4906



NATURAL LANGUAGE




PROBLEM LANGUAGES



COMPILERS



ASSEMBLERS



MACHINE



^.^.^3




4907



STORED PROGRAM




TO AM 1 1 mil IC UNIT





^^1


^H


^^HIHIioGiMMaH


^H


A COMPUTER PROGRAM M^^
THREE DIFFERENT METHODS J

• MACHINE LANGUAGE

• ASSEMBLER LANGUAGE


\ LANGUAGi^

/ ^ 1
1 MNEMONIC

\ LANGUAGE!


1



4908




4909



M.




^


PROBLEM =


Ik




K


ADD TWO NUMBERS WHICH ARE IN STORAGE


AND STORE THEIR SUM BACK INTO STORAGE


MACHINE ASSEMBLER
LANQUAQE LANGUAGE




010000000000000100000 LOA X


H 2-^'.-'ri fry


an 000000 000000 101 000 aoa y


V


100000000000000101001 STA Z


IL



4-BIT INSTRUCTION WORD



0001



OPERATION

(9 BITS)
PECIFIE8 THE OPERATION
TO BE PERFORMED



/"/^•V^



LOO???«



4910



MACHINE LANGUAGE



010000000



000000000101000



LOAD THE
A REGISTER

OPERATION



WITH THE CONTENTS
OF LOCATION 40



/^,^'SO



ADDRESS



ASSEMBLER LANGUAGE



LDA



OPERATION ADDRESS



4911



COMPILER LANGUAGE



X + Y



ADD X TO Y GIVING Z



pia-sa-




4912




4913




PRO '

GRAMi
A


X


OP
c


ERATI
JYSTEf


NG
i/l






^




N


PRO-
GRAM
< B


PROC
(


5RAM


1"


.^.o^'









COMPUTER MEMORY



OPERATING SYSTEM GOALS

Provide Maximum Job Throughput

Reduce Traditional System Overhead

Provide IVIaximum System
Resources To User

» Provide System Reliability

» Provide For System Maintainability

> Provide System Flexibility



4915



Man and Machine

both have a

Control Function



MnlMnS UONIROL
OFIHE MACHMK



Ft9«^



MAN'S NERVOUS SYSTEM



CARRIES SIGNALS
THAT CONTROL



HIS ACTIVITY



4916



its Internal Operations

with Signals through

Electronic Circuits



Fij-fa*



Man performs



^



^n'J




t<



''K.DO.




F»c^.^-i



4917



A COMPUTER

PERFORMS

OUTPUT BY. . .

PUNCHED CARDS
CHARTS

MAGNETIC TAPE
PRINTED PAGE
•PUNCHED TAPE
■DISPLAY SCREEN
• VOLTAGE FOR CONTROL
OF OTHER ITEMS




4918




4919




4920





COMPUTER OUTPUT

SOMETIMES CONTROLS
OTHER MACHINES



DIRECTLY



4922










«li.7'>.\



4923



/^ry'/H



1







_ IPli \wmm — ^




4924



''^^




^'d-2^ . m



4925




40-927 O - pt. 7-7



4926



if a Computer cannot react like man
and can only perform Five fvnctions
similar to man —



INPUT
STORAGE

CONTROL

ARITHMETIC

OUTPU"^



COMPUTER ADVANTAGES



SPEED



CAPACITY



ACCURACY



RELIABILITY



4927



'tr''^mmi^mmf




A'5- ^^




4928



''^^^^'^^




F'b-^'i




4929




;



fi^'9$




4930




4931





4932



SUMMARY

COMPUTERS ARE

• SOPHISTICATED ELECTRONIC
DEVICES THAT HANDLE



NUMBERS



^'S'io



COMPUTERS WORK



• RAPIDLY AND ACCURATELY



UNDER THE CONTROL OF PEOPLE



WHO PROGRAM AND OPERATE THEM



4033



COMPUTERS ARE USED

.

• WHEN LARGE AMOUNTS OF

INFORMATION MUST BE PROCESSED
QUICKLY

AND



• WHEN THE PROCESSING CAN BE
ACCOMPLISHED AS A SERIES
OF VERY SIMPLE STEPS ris-^a.



Senator Hart. The next testimony will be presented by Mr. James
Peacock and Mr. John P. Breyer of the International Data Corp.

Mr. Peacock is the editor of the EDP Industry Eeport, a subsidiary.

Gentlemen, we welcome you.

We will order the prepared materials you filed with the commitfee
printed in full.

[The material referred to appears as exhibit 1 at the end of Mr.
Breyer's oral testimony.]

Senator Hart. If you have any footnotes you would like to add,
you may do it.

I should explain, lest the interruption jar you, at about 3 p.m.
I will be compelled to take a recess, the reason being another commit-
tee is attempting to conclude the markup and action on a copyright
reform bill.

You may proceed.

STATEMENT OF JOHN P. BREYER, EXECUTIVE VICE PRESIDENT,
INTERNATIONAL DATA CORP., NEWTONVILLE, MASS.; ACCOM-
PANIED BY JAMES PEACOCK, EDITOR, EDP INDUSTRY REPORT:
AND J. THOMAS ERANKLIN, COUNSEL

Mr. Breyer. Good afternoon, gentlemen. I consider it both an honor
and a challenge to be here today as you begin your examination of
the computer industry.

I have with me today Mr. James Peacock and J. Thomas Franklin,
counsel to IDC and a member of the Boston firm of Sweeney &
Franklin.



4934

Under Mr. Franklin's guidance IDC has been active in the antitrust
area as a consultant, witness, and reporter. And IDC's newest service
is the IBM Antitrust Litigation Service, which reports developments
in all the antitrust suits now involving IBM.

Our company, IDC, is 10 years old, half as old as the computer
industry. During the course of the past 10 years we have devoted our
professional lives to providing information about the computer
industry.

This is our business. This is our only business. To accomplish this
task we talk to users, talk to suppliers, listen to scientists and advance
planners.

We look for the relationships between economic conditions, product
offerings, general usage trends, and the increase or the decrease in the
use of the many services and products that are sold to computer users.

We gather this information with mail questionnaires, telephone in-
terviews, and personal interviews. The information thus obtained from
users is compiled and then stored in a computer so that we can analyze
it for our clients.

As a matter of fact, today, 10 years, and many client services later,
the foundation of our work is still the computer census data file, the
original product of our company.

This data file has been updated every 6 months since 1964, and it
contains today a description of somewhere between 60 percent and
80 percent of all tlie computer installations in the United States and
overseas.

One reason I stated such a wide range as to the completeness of our
data files is that no one to my knowledge has an exact count on such
information.

Data available from the industry is incomplete and unreliable.
Annual reports and press releases are published by industry j^artici-
l^ants, but very seldom do they i^'ovide information on specific cus-
tomers, product shipments levels, or the return of off-lease equipment.

For this reason, we Ijelieve the IDC data that is gathered from users
directly is the most accurate and o])jective information available as
to the actual use of comi:»uters.

We admit that our data is not as complete as we would like to see it.

We are, however, continually in the process of increasing the size
of our data base, expanding the scope of our research.

We certainly feel we ha\e our fingers on the industry's pulse. And
by continually gathering information from computer users and analyz-
ing the results we are in a ]iosition to i)rovide objective, quantitative
information and objective, qualitative observations to our customers.

Our ]iroducts and services include the data files I have been de-
scribing, multiclient research studie^: on the computer industry, re-
search services foi- to)i financial iustitutions in the Ignited States, the
{)ublication of a weekly newsi)aiier. Comi)uter\voild. and semimonthly
newsletters such as EDP Industry Ixepoi't.

We are intei-uational in scope and conduct research analysis
throughout Europe and Japan.

Ten years ago, as IDC was starting its data file oiieration, it also
began ])ublication of EDP Industry Re]iort, a newslettei- popularly
known as the "Gray Sheet."

This is IDC's fiagslii]i liubliration. and the best known vehiele for
the dissemination of IDC's user and industrv infoi'uiation.



4935

At this point, I -would like to reintroduce Mr. James Peacock, who
has been responsible for editinjr EDP Industry Report for more than
7 years.

As editor of the Gray Sheet, Jim has been an astute observer of the
development of the computer industry. T believe he is eminently quali-
fied to i)rovide this subcommittee with an overview on the history,
current status, and future trends of the computer industry.

Mr. Peacock. Thank yo>i, John.

Good afternoon.

I am goino- to be speaking from material in the background sub-
mission and from some posters prepared from it which we will be
showing you, I will point out what pages we are on as we get to it.

But, primarily, it is built around that. To start with. I certainly
want to point out that I share Mr. Breyer's sense of responsibility
about this appearance.

You gentlemen are taking a look at several industries m an etfort to
determine their possible need for reorganization.

And I certainly don't envy your task. I hope that my presentation
will be helpful. It is my full-time job to think about one industi-y. a
multifaceted one to be sure, one that actually cuts across all other in-
dustries and virtually all human activity, including the Government,
because its products and services have become essential to the veiy
functioning of our society.

So I am very concerned about the future of the computer industry.
I am concerned, pei-sonally. of course, because it provides me a way to
make a living, but that is not what I am talking about. I am concerned
about the future of this industry because I truly believe it liolds the
keys to progress for the world in which we live. I would not like to see
this progress stifled by forces from within the computer industry;
neither would I like to see man's ability to use computers hindered by
artificial and unnecessary constraints on the continued application of
computers to the challenges of the complex soicety in which we live.
I certainly agree with you. Senator Hart, when in announcing this
week's hearing you said. 'T believe it fair to state that computers
re]5resent the central nervous system of our economy,"

I would merely go one step further. Our computer nervous system
is today truly embryonic. The capability of the computer has barely
begun to be applied.

Technological factors that lie closely ahead will probably bring more
change to the use of computeis and. therefore, to our own lives than
they have in the past.

But I am getting slightly ahead. You've asked me to describe the
computer industry. U.S. companies today, from giant IB]\I. and other
large main-frame manufacturers such as Honeywell, I'nivac, Bur-
roughs, NCR. Control Data, down to the clever manufacturers of spe-
cial terminals that we sometimes call information appliances, the sys-
tems houses that buy up minicomputers and mold them into problem-
solving tools, down to the service bureau which is likely to be found
in cities and towns of almost any size whatsoever, generated some $20
billion in revenues last year. Why? Simply because people today need
computers.

That revenue which flows into U.S. corporate coffers from all over
the world is increasing at about 15 percent annually, doubling every
5 years, doubling and quadrupling us into 1984, at least.



4936

Why? Again, demand; user demand. Let's take a look. I really
thought long and hard, from the day that I discovered the challenge
you had extended to my company and me, about liow to share with you
the understanding I have of the computer industry.

I asked myself —

What is the best way to plump you down in the midst of this very complex
industry, to let you see all the pieces and then help you put it together again, at
least into a recognizable picture, and to do all this in a very few minutes, and
to do it w'th no intentional bias?

It seems to me that the best way to give you this initial understand-
ing is to look at users and how^ users spend their money.

This is something that IDC investigates by field research on a regu-
lar basis. So let's take a look at my first chart that the artist has put
together for us.

That is the chart on page 12. "^ For the moment, don't worry about
the details. Simply look at the four pies we have drawn.

[The chart referred to appears in exhibit 1 at the end of this panel's
oral testimony.]

Mr. Peacock. Notice how these pies, representing U.S. user
spending in billions of dollars, have grown over time. Just 15 years ago
U.S. computer users spent less than $1 billion on computers, includ-
ing the salaries of the people who they paid to run them.

Now% look at the growth in 5-year increments that we have shown,
from less than $1 billion in 1958 to $4 billion just 10 years ago
to about $12 billion, to the $20-billion level last year.

And we at IDC think we know enough about user spending patterns



Online LibraryUnited States. Congress. Senate. Committee on theThe Industrial reorganization act. Hearings, Ninety-third Congress, first session [-Ninety-fourth Congress, first session], on S. 1167 (Volume pt. 7) → online text (page 8 of 140)