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 77 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 77 of 140)
Font size
QR-code for this ebook


tory pricing.

Interim Relief Measure No. S: IBM shall not make premature announcements
of its computer products.

Comment: Tliese three remedies are noncontroversial, minimal provisions
necessary to preclude IBM from direct, anti-competitive behavior during the
interim period.

//. Remedies Providing Access to IBM's Patents, Know-How, Software, Product
Specifications and Unique Products

The following injunctive provisions would provide rights to license IBM's
patents, software and other know-how, to have disclosure of its product specifica-
tions and to buy its unique products on an OEM basis, all of which rights could,
but would not necessarily, be extended indefinitely as part of the final relief.

Interim Relief Measure No. f; IBM shall not refuse to license its EDP jiatents
and EDP know-how, patentable and unpatentable alike (and, where it has the
right to do so, sublicense its rights under the EDP patents and know-how of
others) to all applicants on a reasonable-royalty, non-discriminatory basis. IBM
shall afford such applicants reasonable assistance in understanding and utihzing
such rights and information through plant visitation opportunities and other
methods, the expense of which may be charged applicants on a reasonable, non-
discriminatory basis, not to include the original cost of producing the know-how.

Comment: This provision would preclude IBM from using its strong patent
and technical position to injure competition.

Interim Relief Measure No. 5: IBM shall not refuse to license, with the right
to sublicense, its software products, including applications programs, to all appli-
cants on a reasonable-royalty, non-discriminatory basis.

Comment: This provision would prevent IBM from using its dominant position
in software to inhibit sales by competitors and to interpose itself between its
competitors and their customers.

Interim Relief Measure No. 6: IBM shall disclose to other suppliers of EDP
products having a legitimate interest in such disclosure, the complete product
specifications of its EDP hardware, software and media products, including
functional, interface, media and performance specifications, and also the ap-
plicable communications procedures, at the time of product announcement, or 24
months before the reasonably anticipated delivery date, whichever is earlier.

Comment: This provision would ameliorate the competitive advantage that
IBM has by virtue of its control of de facto standards.

Interim Relief Measure No. 7: IBM shall not refuse to sell to original equip-
ment manufacturers any EDP product, a substitute for which is not reasonably
available elsewhere, to be incorporated in EDP systems produced by such manu-
facturers, at prices and with delivery commitments which enable them to efifee-
tively compete.



5545

Comment: This provision is designed to eliminate tlie possibility that IBM
could use a monopoly position in a new EDP product to hamper the competitcrs
sales of systems which depend to a material extent on such new product.

///. The Controlled Growth Remedy

The objective of this remedy is to preserve the status quo regarding liiM's
market share pending the effective date of final relief.

Interim Relief Measure No. 8: IBM shall be subject to suitable measures to
limit its growth to that of the market (s), or, preferably, less. Such measures
should be based on data readily available to the Court.

The markets to which this interim relief measure should be applicable would
be selected in light of the District Court's findings (before or after trial) as to
the market definition and the scope of IBM's monopoly power. However, the
markets to which this remedy would apply would not be confined to those found
bv the Court to have been monopolized by IBxM. Rather, the remedy would be
extended to the additional EUP submarkets which would require the protection
afforded bv this remedy, and also to the related markets tor EDP products and
services (such as communications, terminals and professional services) which
are most likelv to be the target of future domination by IBM.

IBM should be required to give priority to its then existing customers so as to
diminish the number of new customers potentially -'locked-in" to IBM equipment
and thus "preserved" for the IBM successor companies.

Comment: As stated in the accompanying Memorandum, the philosophy behind
this remedy is that it is the only sure way of preventing a couiplete domination
of the industry by IBM during the period prior to final relief. The remedy would
be applicable to "the market (s) found to have been monopolized by the District
Court, and also to the related product and service market (s) and sub-market (s)
which would be particularly vulnerable to an IBM takeover if its expansion in
principal market (s) were limited during the interim period.

Exhibit Q.^CDC Memorandum — Rrport to U.S. Department of Justice Re: IBM
Attempt to Monopolize and Monopolisation of the Automatic Data Processing
Industry

In Re : International Business Machines Coep.

I. introduction
A. General

The information compiled herein has been gathered by Control Data Corpora-
tion in response to anticompetitive activities of International Business Machines
Coi-poration and its subsidiaries, including IBM World Trade Corporation and
Service Bureau Corporation, hereinafter collectively referred to as "IBM". IBM,
by far the dominant company in the field of automatic data processing, has l>een
engaged in conscious and determined effort, intensified in recent months, to en-
large its U.S. and world-wide monopoly position at the expense of its much small-
er competitors, and its practices have amounted to clear violations of the U.S. anti-
trust laws. Unless immediate and forceful action is taken to terminate the prac-
tices and to restore legitimate competition, severe injury to participants in the
automatic data processing industry will result.

The material presented herein is the result of informal investigation by one
small competitor and is limited by the resources available to it. Much of the
detail which would be available through discovery in judicial proceedings, par-
ticularly from IBM. is therefore lacking. Also, because of the nature of the prac-
tices and the possible adver.se effect upon sales efforts of Control Data, it has
not been feasible to interrogate customers in most cases, although information
lias been obtained from customer sources wherever possible. Because of the con-
sistency in the great amount of detail that has been gathered and the fact that
on occasions where opportunity to corroborate has existed tlie information re-
ceived has been subsantiated. it is believed that the great bulk of the information
set forth is factually correct, notwithstanding the possibility that in isolated
instances some fact may not have lieen accurately reported, some circumstance
may have been misinterpreted, or some explanation may not have been made
known.

The material set forth in this Introduction is merely for the purpose of facili-
tating an undersranding of the setting in whicli the anticompetitive practices take
Iilace. The descriptions of the market, of the equipment involved and of the trade
abuses themselves have accordingly been condensed and simplified.



5546

B. Description of the Industry

The industry iuvolved is automatic data processing ("ADP"), which utilizes
botli electro-mechanical and electronic equipment for the processing of various
types of information, or data. Historically, the base of the ADP industry was
electro-mechanical ("tabulating"' and "punched card" equipment), which has
been supplemented, in recent years, by electronic digital computers for electronic
data processing ("EDP'). Thus, many modern computing systems have evolved
as a combination of one or more central electronic computers operating with a
variety of electro-mechanical peripheral equipments, including readers and
punches for punched cards and punched paper tape, printers or tabulators of
various speeds, magnetic discs, drums and tapes, optical character readers, etc.
The equipment is used for processing data, such as routine clerical, inventory
and accounting information ; for solving mathematical and scientific problems ;
for solving complex problems in areas such as production and profit optimization ;
for industrial process control ; and for space vehicle launch control. The principal
part of this Report relates to the electronic digital computer systems segment of
the data processing industry, although the technologically-earlier tabulating (or
punched card) part of the industry is a significant part and continues to have
,heavy impact upon the electronic computer industry.

The computer industry has experienced an extremely high rate of growth and
Very rapid technological development throughout its brief lifespan, from the early
lOsb's when large, slow, vacuum tube computers were used, to the present, with
extremely fast, "solid-state", transistorized computers far exceeding in speed and
capacity the earlier models of computers. To illustrate the rapid change in tech-
nology,' the most powerful computer in general commercial use during 1952-55
was the Sperry Rand Univac I Computer, followed shortly by the Unlvac II. Of
over 80 of such computers believed to have been delivered through 1960, when
the last Univac II was delivered, only 34 were still in use at the end of 1064.
By the end of 1964, the most powerful computer in use was the CONTROL DATA




leadership „ _ . ~ -...^a

Jr., IBM's chief executive ofiicer, at a Shareholders' Meeting in April of 19bJ.

The electronic computer industry started during the years of WW II, when
some elementary computers were built for defense applications. It grew slowly
after WW II to the point where by 1952 perhaps one hundred small electronic
computers were installed in the U.S.. a few of which had been installed by IBM.
By the end of 1956. a total of 810 computers had been installed in the U.S., hav-
ing a cumulative value of $83 million. The growth of the electronic computer
industry has been very rapid since that date, to the point where over 25.000
general purpose digital computers were installed in the U.S. at the end of 196o,
plus an estimated additional 10,000 computers outside the U.S. In addition to
these figures, there are several thousand special purpose electronic computers
installed. During 1965 alone, it is estimated that there were over 10,000 new
computers installed, an increase of approximately 30 per cent over the total
number of computers installed as of the end of 1964. Thus, the combined sale and
lease base of the overall industry continues to grow at the rate of approximately
'>0-30 per cent per year, a rate which has been maintained for several years.
Industry sources generally believe that the total market, U.S. and world-wide,
will continue to grow at the same rates in the predictable future, with no plateaus
in sight Because of the large number of leased systems, however, the reported
income (as in the case of IBM) does not immediately show up as such a large
rate of increase.

C. Market Shares

IBM has long been the dominant company in the automatic data processing
indu-try As alleged in the Complaint, Civil Action No. C-T2-3-44. Filed Janu-
ary 21 "l9.52. U7iifc(l F:faies of America, PUiiitiff. r. Infcniotional Busrne.<<s Ma-
chines' Corporation, Defendant, IBM then owned 90 per cent of all tabulating
machines in use in the United States, and was leasing more than 100.000 electri-
cal tabulating machines to more than six thousand lessees at an annual rate of
approximately $100 million. , , , ^. ^- i „

Since 1952, IBM has significantly increased its base of tabulating machines, by
also buildins on and adding an extensive line of more advanced electronic digital
computers. IB:M's annual revenues have meanwhile grown to a 1965 amount of
$3,572,824,719.



5547

IBM, as a matter of strict policy and practice, is higlily secretive, and does
not disclose the total number of its computer installations to the trade press,
industry associates, etc. ; therefore, exact data regarding IBM's current share
of market (based on the equivalent original-sales-value of installed machines,
whether sold-outright or leased) is very difficult to ascertain, except by equipment
cauuirf iiy revealed by known customers of IBM (although most other industry
members periodically publish their census of installations). Total "Industry
Data" as reported by industry analysts, is, therefore probably subject to varia-
tion on the conservative side simply because it is impossible for any outsider to
locate all of the IBM installations. Under these circumstances, it is generally
estimated that IBM's share at the end of 19G5 of the electronic computer
("EDP") segment of the market was at least 70-80 per cent and when the
electro-mechanical tabulating systems part of tlie business is added to this,
IBM's share of the total automatic data processing industry is probably well
above 80 per cent of all ADP installations, both world-wide and in the United
States. IBM reports that it continues to lease 80-90 per cent of its equipment.
IBM has reported an average of 17 per cent increase in revenues per year, com-
pounded annually, since 1952. Its aggregate revenues for 1964 having been
13.239,359,581, IBM reported its consolidated earnings for 1964 at $431,159,766.
IBJI recently preliminarily reported 1965 revenues of $3,572,824,719, up $333 mil-
lion or 10.0% over 1964, with earnings of $476,902,490, up 10.5% over 1964
earnings.

The cumulative sales value of computers installed throughout the world by
the end of 1965 was about $11.2 billion, of which about 75 per cent were in-
stniL'd in the U.S. IBM has reported that, in recent years, the operations of its
U.S. -controlled foreign operating company, IBM World Trade Corporation, have
been increasing at about twice the rate of IBM's U.S. operations, with the ex-
pected result that by the year 1970, IBM's U.S. and foreign operations will be
about equal. Today. IBM's U.S. operations comprise over 2/3 of IBM's total.

As of the end of 1965, no competitor of IBM accounted for as much as 8 per
cent of the total market, the largest being Sperry Rand Corporation, followed
by several smaller competitors, including Control Data, which possessed about
4 per cent of the market. In recent years, the intense competition, particularly
from IBM, has caused most of IBM's competitors to operate at a loss, frequently
quite severe. As a result, a number of competitors, both in the U.S. and abroad,
have been compelled to sell out, merge or otherwise cease competition, e.g..
Underwood, The Bendix Corporation, General Mills. Alwac. Autonetics, General
Precision, Compagnie de Machines Bull of France, Zuse K.G. of Germany,
Olivetti of Italy. A descriptive account of this situation contained in the "Wash-
ington Post" on July 8, 1965, is attached as Appendix 1.

The fact that 70-80 per cent of the total computer systems installed in the
United States and world are IBM systems makes it necessary, in order to attain
a significant position in the market, that other computer manufacturers design
their equipment and the programs which operate them to be compatible with
IBM equipment and programs. Most computer users presently have (or have
(n-dered. or probably will at some future date order) IBM equipment, and these
customers have spent or will spend enormous amounts of money to gather, or-
ganize and exchange the data to be processed by these systems, and to develop
their computer programs to operate their total systems which often comprise
many inter-related computers. They will not consider acquiring a second manu-
facturer's equipment unless and until such data and programs, or substantial
portions thereof, can be used on the other manufacturer's equipment. Further,
they, as well as customers acquiring computer systems for the first time, fre-
quently wish to exchange data or programs with other coiiiputpr useTs, which
users generally have IBM equipment. Such exchange is feasible only if the equip-
ment ("hardware") and programs ("software") to be acquired by the customers
are functionally compatible with that of the other party to the proposed
exchange.

Automatic data processing accounts for about 80 per cent of IBM's total
revenues and 90 per cent of Control Data's revenues. No other company with a
significant position in the industry receives as much as 10 per cent of its reve-
nues from such source, except Sperry Rand, which receives less than 20 per
cent ; General Electric and RCA receive only 2-3 per cent of their income from
computers. Therefore, IBM's anticompetitive practices, while wrongful and
injurious to all its competitors, are particularly injurious to its smaller com-
petitors including Control Data, who rely on the data processing business for



5548

the majority of their income and earnings, and who do not have large sustaining
incomes from such as color TV and steam turbines to carry them through
competitive crises.

The growth of Control Data Corporation (hereinafter referred to as Control
Data) in the industry since Control Data's formation in 1957 was initially the
result of careful concentration in specialized areas, particularly scientific, and
the avoidance of an attempt to enter into general competition with the very
much larger companies in the industry. The technological excellence of Control
Data's equipment, coupled with careful marketing effort, enabled the company
to gain an initial foothold in the industry. The subsequent design, engineering
and manufacture of large-scale electronic digital computers which far exceeded
the capabilities of competitive equipment, along with the acquisition of several
smaller companies in the industry, have aided Control Data in expanding its
total revenues from $19,783,745 in 1961 to $160,473,162 in 1965 (in each case to
June 30, when the company's fiscal year ends). Most recently. Control Data's
growth and earnings both in the United States and abroad have been greatly
inhibited by the anticompetitive practices of IBM as described in this Report.

D. The Extra-Large Computers

The so-called "extra-large" or "giant" computers deserve special mention be-
cause of their importance — their enormous capacity and great speed, their
ability to handle vast quantiti'^s of data and solve tremendously complex prob-
lems, their indispensable role in the nation's space and defense efforts, and the
sheer dollar investment in each such computer or system. These giant computers
are also significant because of their "time-sharing" potential, that is, the capacity
for simultaneous computer use by several users, frequently at locations some
distance removed from the central computer. Because of their importance, IBM
has centered much of its anticompetitive activities about these computers and
has called upon its vast financial and manpower resources in a powerful drive
to capture the extra-large computer business and to eliminate competition with
respect to their manufacture and sale.

Although there is no generally accepted classification of "extra-large" or
"large-scale", computer users and manufacturers agree upon the use of "power
ful" or "power" as an apt description of magnitude, "power" being used to refiect
the combination of machine speed and capacity. The most powerful computer
presently in use is agreed by all to be the CONTROL DATA 6600 (See Appendix
2 for descriptive material), which provides execution speeds of 3,000,000 instruc-
tions per send. The purchase price for a typical 6600 computer is approximately
$5,000,000 to $6,000,000. The first 6600 was delivered in September 1964 and. as
of January 1, 1966, a total of six had been delivered. The CONTROL DATA <i000
Series also includes the smaller 6400 the first of which is expected to he delivered
early in 1966, and the more powerful 6800, which is scheduled for first delivery
in June 1967.

While several other "large-scale" computers have been offered commercially by
manufacturers other than IBM and Control Data, these are far less powerful
than the existing CONTROL DATA 6600, and are generally considered to be in
a smaller class, merely "large-scale" rather than "extra-large". (Late in 1965
Sperry Rand, Burroughs and General Electric announced extra-large computers,
but these have not been a considerable factor as far as competition to date has
been concerned). IBM made a much belated entry in the extra -large computer
market. Although it announced its new 360 Series of computers on April 7. 1964,
(a so-called "third generation"), the then-announced new line lacked an exrra-
large computer. IBM nevertheless attempted to interest customers in such a
computer and began to offer a changing series of computers variously knowii as
360/90, 360/92. 360/95, and more recently 360/85. From time to time, subsequent
to April 7, IBM announced new or varied computers (or modifications of prior
computer:^) in the 360 Series. Many of these are believed to have been nonexist-
ent, as was the Model 92. The Model 92 was described by IB]M as providing execu-
tion speeds of 3 million instructions per second, makins,- it comparable to the
CONTROL DATA 6600. The 360/92 was offered for sale for approximatelv .$6.-
000,000 to •.$9,000,000. No 360/92s have been delivered, and the trade press has re-
cently stated that IBM will no longer have the 360/90 models for sale. Whether
these are being replaced by the Model 360/85 which has been veiled in secrecy
or the Model .360/91 which was formally announced on January IS. 1966, is not
known. The very recent report in tb.e January 21. 1966 WALL STREET
JOURNAL (Appendix 3) completes the full circle of confusion.



5549

E. The Wrongful Practices

The remaiuder of this Report is devoted principally to elaborating upon the
unlawful practices in which IBM has been engaged over the years. IBM has in-
tensified the practices markedly as a result of the impact oi the CONTROL DATA
6600 upon the market. When IBM found itself without a large-scale computer
to offer as a competitive alternative to the 6600, ail of the .stops were pulled and
a frantic, no-hold-barred campaign was ordered at the top level of IBM manage-
ment. IBM accelerated the announcement of the 360 Series. Top management
was reorganized, one key executive being reassigned with instructions to "stop
Control Data".

Before the 360 Series was announced and with the effect of seriously delaying
customer procurement of competitors' equipment, IBM began •"leaking" informa-
tion about its forthcoming "third generation" of computers (which IBM said
would render obsolete existing computers). After announcement of the 360
Series, it became apparent that many computers in the series were non-existeni.
IBM made change after change in models and in specifications, tpioted unrealistic
delivery dates, and deferred delivery dates that had been sclieduled. New com-
puter "models' replaced the recently-announced models, even when the earlier-
announced models had been ordered by customers, and obviously false claims
were made about machine capabilities. How IBM used its "Paper Machine" to
injure competitors and in an attempt to increase its monopoly is described in
the section following this Introduction.

IBM directed extra heavy attention in its Paper Machine campaign to educa-
tional institutions and scientific organizarions, tlie simple explanation for this
being that these are the foremost potential customers for extra-large computers.
In addition to being a prospective customer itself, a university has substantial
and unique strategic, prestigous and actual influence on other potential cus-
tomers, because compatibility among computers assumes considerable import-
ance when the users exchange computer programs, a frequent occurrence among
scientists. Also, scientists, students and trainees at a university acquire a strong
predisposition toward using the same manufacturer's line of computers in their
later employment.

Most important, there are a few key educational institutions, such as Massa-
chusetts Institute of Technology, Stanford University, University of California,
and Carnegie Institute or Technology whose mathematicians and other computer
personnel are particularly skilled and advanced in computer usage. Manufac-
turers are extremely anxious to place equipment at these institutions because
their personnel will then be available to the manufacturer to criticize and debug
the equipment and develop new and imaginative programs or software for operat-



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 77 of 140)