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 133 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 133 of 140)
Font size
QR-code for this ebook

for facilitating the resolution of these intei-ferences. Eventually, these late 1955 and early 1956 IBM-SR discussions ex-
panded to include discussion of other outstanding differences between the two
companies, particularly patent licensing and the SR antitrust suit against IB:\r.
In the discussions SR and IBM considered the possibility of a TAB and EDP
patent cross-license and an exchange of know-how.

15.24.3 On January 25, 1956, the United States District Court for the Southern
District of New York approved a consent decree entered in the case of United
States V. IBM. The 1956 consent decree contained numerous remedial provisions
directed at the IBM tab and EDP monopoly. In section IV of the decree, for example. IBM was required to offer
for sale as well as lease all of its tab and EDP machines.


15 24.3.3 In section IX of the decree, IBM was required to afford certain appli-
cants (otlier tlian agents or employees of a manufacturer of tab or EDP ma-
chines) the opportunity to obtain training to repair and maintain IBM tab and
EDP machines and to furnish to such applicants repair and maintenance manuals
and instruction books on all IBM tab and EDP machines. IBM was also required
pursuant to section IX to furnish such repair and maintenance books and
manuals to owners and lessees of IBM machines.

15 24.3.4 Regarding patents, section XI of the consent decree required IBM
to license any applicant under any, some or all IBM's tab and EDP patents and
applications filed prior to 1961. IBM was permitted to charge a reasonable
royalty for most patent licenses, but section XI of the decree also stated that if
the applicant and IBM were unable to agree on what constitutes a reasonable
royalty, either could apply to the Court for such a determination (with the burden
on IBM to prove reasonableness). Section XI of the decree opened the injunction so as to permit IBM
to grant patent licenses under future tab and existing or future EDP patents on
the applicant's granting fair value to IBM, as a reasonable royalty, including
licenses under any, some or all of the applicant's patents.

l.'j.24.3.tj In addition to the information contained in the IBM repair and
maintenance manuals required to be furnished applicants under section IX,
section XIV of the decree required IBM to furnish any applicant for a patent
license with technical information with respect to, and for use in the manu-
facture of, a list of some 9 8IBM "tabulating machines". Basically, this list
(set forth in Appendix A to the decree) of "tabulating machines" included
all of IBM's tab line as of January 1, 1956 of which three were electronic, viz.,
the IBM 604 Electronic Calculating Punch, the 083 Sorter and the 101 Elec-
tronic Statistical Machine. Tlie card readers, card punches and other devices
covered by the decree were ncjt electronic nor used primarily in or with an
electronic data processing system. These were not "EDP systems or machines",
therefore, and the consent decree did not provide that the EDP know-how later
obtained by SR was to be available to anyone. The technical information to be provided by IBM pursuant to
section XIV of the consent decree was to be such as to enable the applicant
satisfactorily to manufacture or assemble the tabulating machinery covered
thereby. IBM was permitted to charge applicants only for the cost of reproducing
the technical information.

l.".24.3.8 Section XIII of the consent decree also prohibited IBM from enter-
ing into any agreement or understanding relating to tab or EDP machines or
systems which, "pro\ades for disclosure to IBM on an exclusive basis of any
invention, formula, process or technical information".

15.24.4 The 1956 consent decree required IBM to grant specifically defined
tabulating systems know-how technology, but not EDP systems know-how tech-
nology, to certain applicants.

15.24.5 Honeywell executives were aware of the IBM consent decree, and
Henry Hanson, Honeywell's EDP patent counsel, prepared a summary of the
decree for distribution to selected Honeywell personnel.

15.24.6 Herbert Bergsou, one-time Assistant Attorney General in charge of the
Antitrust Division, had attempted (for SR) to persuade the Justice Depart-
ment to include a consent decree provision for a public dedication of IBM's
EDP know-how technology in order to correct what he and SR saw as the already
noticeable advantage of IBM in the EDP industry.

15.24.7 SR and Bergson considered it necessary, in order to have viable
competition in the EDP industry, for all competitors to have royalty-free
access to IBM's EDP know-how technology, as well as its TAB know-how

15.24.8 However, SR was unsuccessful in having the EDP know-how require-
ment included in the consent decree for the public : subsequently it began direct
negotiations with IBM to and did obtain the IBM EDP know-how solely for

15.24.9 On March 5, 1956, a conference was held between SR and IBM to
discuss the possibility of settlement of the then pending patent and antitrust
disputes between the two companies.

15.24.10 Because SR believed that the consent decree January 1, 1956 cutoff
date was insufiicient, SR suggested that the cutoff date be extended approxi-
mately a year and that the definition of technical information be broadened to
include EDP, as well as tab.

40-927—75 66


15.24.11 On May 28, 1956, SR and IBM representatives met to discuss the
antitrust problems which might arise out of a then proposed settlement
between IBM and SR which included a complete sharing of know-how tech-
nology in both the tabulating systems and EDP systems area and a settlement
of all Patent Office interferences between the two companies.

15 24 12 At this meeting, IBM submitted a proposal to SR that each company
would grant the other a royalty-free, non-exclusive license under all tab and
EDP patents and patent applications as of July 1, 1956 ; each would release the
other from liability for past infringement; each would furnish the other, at
cost, tab and EDP know-how as of July 1, 1956 ; IBM would pay $2,000,000 in
royalties to SR ; and SR would dismiss its antitrust suit.

15 24 13 As part of the negotiations at that meeting, representatives of IBM
told SR that neither company would get a patent in the "big interference case"
between the IBM "SSBC" computing system and the SR "ENIAO" because of the
fact that the ENIAC machine and any alleged invention therein were in public
use prior to the critical date. , . ^ t. ^r.

15.24.14 It was stated by both SR and IBM that it was m the interest of both
companies to settle their Patent Office interferences.

15 24 15 Shortly after making its May 28, 1956 proposal to SR, IBM counter-
claimed against "SR for infringement of 35 patents, mostly in the tabulating
area, and^ it petitioned the Patent Office for the institution of public use
proceedings against the ENIAC patent application.

15 24 16^ SR management evaluated the IBM May 28, 1956 offer and concluded
that EDP should be left out of the settlement. But it remained interested in tab,
particularly because it suspected that IBM was holding under wraps an IBM
tabulating machine called the World Wide Accounting Machine (for WWAM).

15.24.17 SR officials estimated that patent licenses and technical information
on the WWAM machine would be worth $20,000,000 to SR.

15 24 18 SR sent a counter-offer to IBM on June 27, 1956, proposing a tab patent
cross-license and an exchange of tab technical information as of December 31,
1956. SR also asked for a $20,000,000 payment by IBM. ^

15 2419 SR hoped that the December 31, 1956 cutoff date contained in its
June" 27, 1956 counter-offer to IBM would be late enough to "catch" IBM's
WWAM tabulating machine. , .. ^ x

15 24.20 At meetings subsequent to June 27, 1956, SR reduced its offer to
$10 000 000 and IBM eventually accepted, but one the basis that the agreement
include' licenses under IBM and SH's EDP patents and applications (including
the ENIAC patent application) and that the cut-off date for know-how be
shortened to October 1, 1956. After much discussion, the parties agreed to these
fl. rr & D ff^mG n t s •

15 24 21 Again in the July 1956 negotiations. IBM representatives told SR
representatives that IBM's patent counsel, Sanborn, Brumbaugh & Cooper, New
York City, had forecast that IBM would prevail ultimately on the pubhc use
question and that the ENIAC patent would be invalidated.

15 '>4 '>2 At a meeting with SR on July 16, 1956, IBM representatives took the
firm position that the ENIAC application had to be a part of any settlement

15 24 23 At a July 23, 1956 meeting, SR was successful in obtaining IBM's
agreement to pay an additional royalty if certain conditions were fulfilled.

15.24.24 A July 25, 1956 draft of the agreement contained a provision for
exchange of EDP technical information on machines announced or released
to production as of October 1, 1956.

15 24.25 In July of 1956, IBM's top management indicated a concern that
the largest patent'license deal in history would not be well received by the pubhc.

15 24 26 There were final drafting sessions on August 1 and August 2, 1956.
Bergson testified that he had probably reviewed the final drafts of the agree-
ment and orally advised SR that the agreement was lawful, , . .,

15 24.27 At the August 2, 1956 drafting session, Bergson suggested that the
parties submit a copy of the agreement to the Department of Justice. After the
agreement was signed, Bergson personally delivered a confidential copy of the
agreement to Marcus Hollabaugh, a Justice Department attorney who had been
in charge of the IBM litigation.

15 24 28 In August 1956, a meeting was held between representatives ot bK
and IBM to work out the procedure for settling major interferences and lesser
Patent Office interferences. x i. *. -a

15'>4 29 Interferences were divided into major and lesser to try to avoia
raising antitrust problems over a settlement of major interferences between the


conspirators without presentation of some evidence by IBM; all such show-
ings were essentially a formality.

15.24.30 At this August, 195C meeting, representatives of IBM again repeated
to representatives of SR that the "public use" defense was a good one against
the ENIAC patent.

ir).24.81 On August 15, 1956 the Assistant Attorney General wrote IBM's
counsel and asked for a copy of the agreement. Accordingly, after the agreement
was executed on August 21, 1956, IBM's counsel sent a copy to the Assistant
Attorney General.

15.24.32 Defendants used their EDP patent and application portfolio in influenc-
ing IBM to enter into the 1956 Agreement at a time when the conspirators had
about 95% of the EDP market.

15.24.33 On August 31, 1956, IBM and SR entered into a complete cross-license
under all their EDP and TAB equipment patents and patent applications, a
complete sharing of all EDP and TAB equipment know-how and a complete
settlement of all EDP and TAB Patent Office interferences.

15.24.34 The completeness of the August 1956 technology sharing is evidenced
by the following excerpt from that Agreement :

^''Exchange of Technical Information. Sperry Rand and IBM will each, to
the extent that it has the right to disclose such information, furnish to the
other, as soon as practicable but in no event later than January 1, 1957, all
technical information, both domestic and foreign, in its possession as of
Octol)er 1, 1956, relating to, and for use in the manufacture of tabulating systems
and machines and electronic data processing systems and machines which
shall have been released to production or announced to the public for sale or
lease by it up to and including October 1, 1956.

For the purpose of this Section 5, "technical information" shall mean the
following information :

(1) a complete set of drawings used in the manufacture of all the detail parts,
assemblies, subassemblies, circuits, components, etc., for the product:

(2) a complete set of operation sheets, duplicates of those which the manu-
facturer used in the manufacture of the product ;

(3) a duplicate set of all inspection specifications for parts, assemblies, sub-
assemblies, circuits, components and machines ;

(4) a duplicate set of drawings of the tools, jigs, dies, and fixtures necessary
for the manufacture of the product ;

(5) a bill of all materials necessary for the manufacture of the product and
specifications to cover the purchase of such materials ; and

(6) illustrated seiwice manuals listing all the detail parts of the product and
setting forth mechanical and electrical instructions to enable servicemen to
maintain and service machines.

* * * :^: '!: * *

In the event that either party shall give the other party notice in writing
that the technical information furnished to it is inadequate to enable it satis-
factorily to manufacture the commercial end product involved, the other party will
supply such further written explanation of the information furnished as may be
reasonably necessary for such purpose."

15.24.35 The 1956 IBM-SR Agreement as finally consummated contained, among
other things, the following terms :

(a) an exchange of patent licenses on tab and EDP patents and applications as
of October 1, 1956;

(b) an exchange of technical information on tab and EDP machines announced
to the public or released to production as of October 1, 1956 :

(c) a payment of a fixed annual royalty by IBM of $1,250,000 a year for eight
years ;

(d) a payment by IBM of an additional royalty, if a patent issued on the
ENIAC application prior to January 1, 1956, of 1 per cent of the manufacturing
cost of each IBM EDP machine embodying any invention covered by the claims
of the ENIAC patent manufactured within the United States between October 1,
1956 and October 1, 1964 — after deducting the $1,250,000 payments as a credit ;

(e) dismissal of SR's antitrust suit against IBM and dismissal of IBM's patent
infriiiaement counterclaim against SR.


15.24.36 IBM and SR also on August 21, 1956 entered into a series of pro-
cedural agreements to dispose of the various outstanding patent interferences
between patent and patent applications of the two companies. These agreements
were characterized as "very conventional", and Honeywell (Datamatic) later
joined in several of them.

15.25 I find that the cross-license and exchange of technical information agree-
ment was an unreasonable restraint of trade and was an attempt by IBM and
SR to strengthen or solidify their monopoly in the EDP industry.

15.25.1 Section 1 of the Sherman Act prohibits "Every contract, combina-
tion * * * or conspiracy, in restraint of trade * * *." Nonexclusive patent cross-licensing in itself may be proper. During the negotiation of their 1956 Agreement, both IBM and SR
had expressed the view that they were not particularly interested in exchang-
ing EDP know-how (or technical information) ; each asserted its belief that its
own position in EDP was superior to the other's. Nevertheless, a provision for an
exchange of EDP technical information — on machines announced to the public
or released to production as of October 1, 1956 — became part of the agreement. Shortly before the 1956 SR-IBM Agreement was executed, J. Presper
Eckert , then a UNIVAC executive, expressed opposition to the agreement. Among
other things, Eckert was concerned about the treatment of "some dubious
information known as 'knowhow' (whatever that is)". Eckert also wrote:

'A considerable emphasis has been placed on the value of something known as
'know-how' (presumably this means manufacturing drawings, etc.) to us. I can-
not find anyone in Engineering who actually knows how we can use this so-called
valuable know-how. The only work I know of at IBM which would be of great
interest to us would be that of high speed circuity and the study of new electronic
storage techniques. Since, however, none of this is at the present time being used
or offered for sale we would not get anything on the things that they are doing
which would be of real value to us * * *." Notwithstanding Eckei't's demurral, the 1956 Agreement was con-
cluded and IBM and SR went about working out details for the exchange of
technical information. IBM and SR representatives met again several times in
the fall of 19.56 for this purpose. Since neither IBM nor SR really knew what it wanted or might expect
in the other's technical information, it was ultimately agreed to have a broad
exchange. However, a large number of obsolete machines were removed from the
lists of machines on which information was to be exchanged. Pursuant to the August 21, 1956 Agreement IBM and SR did, in late
1956, exchange thousands of documents containing proprietary, technological and
production know-how of the kinds described in the Agreement with respect to
the EDP and TAB systems and machines listed in the Exhibits A to the Agree-
ment ; at least as late as 1962, SR requested and received technical information
from IBM pursuant to the 1956 Agreement. In the fall of 1956, SR received IBM manuals called for by the agree-
ment and also received the manufacturing information on the IBM 604 Electronic
Calculating Punch (a tab machine listed on appendix A to the IBM consent
decree). The remainder of the manufacturing information which IBM supplies
was received by SR in late December of 1956. After this information was received,
it was placed in the custody of Robert Kalb at a UNIVAC facility in St. Paul. The technical information SR obtained from IBM did not disclose
design alternatives that had been considered by IBM and rejected, cost data, price
data, software, third-party proprietary information or information on IBM's
field experience. In October of 1959, McNamara (General Counsel of SR's Remington
Rand division) approached IBM's Birkenstock, stating that perhaps the time had
como to discuss extension of the 1956 cross-license vrith an exchange of know-how
Birkenstock replied that he did not feel that an exchange of know-how was practi-
cal .and that :

"We were able to include a know-how exchange in our present cross-license
agreement because we had already been required by our Consent Judgment to
make January 1, 1956 level know-how available. It was relatively simple for us
to merely update it to a level consistent with the date of our cross-license agree-
ment. It would, however, be a tremendously expensive and time-consuming job
to update it again to a current level. Furthermore, If we were to make current
know-how available to RemRand, we would feel obligated to make it available to
others who request it as well. The cost of maintaining the records would be too

5873 No furtlier exchange of unpatented know-how (or technical informa-
tion) between IBM and SR was entered into after 1956 except under the patent
exchange of 1965.

15.25.2 None of the other EDP competitors ever received either SR's or IBM's
EDP know-how.

15.25.8 After receiving the technological and production know-how, from each
other, both SR and IBM studied and used the information ; SR refused to sell or
allow use of any of its equipment made in consequence of the technological merger
by any EDi* industry member at any price; IBM sold or leased such items at
retail prices ; by reviewing and studying the technical information shared under
the 1956 Agreement, SR and IBM were better able to know and to evaluate the
options available to each of them and to decide what equipment to build and how ;
having such opportunities involves time and cost saving shortcuts in the evalua-
tion and selection of alternative routes, One instance where manufacturing information on an IBM EDP
machine was examined by SR was on the IBM RAMAC device. The drawings on the RAMAC were incomplete and, in any case, SR
had prior to 1957 already developed considerable competence in random access
storage, both disc and drimi. Indeed, SR's Appleton patent, filed on February 2,
1955, and licensed to IBM but not to others, discloses an invention bearing a
striking similarity to the IBM RAMAC. SR made no specific use of the IBM RAMAC drawings as it could
have ; instead it continued to follow the drum approach to random access, which
it had used prior to 1957.

15.1:5.3.4 The SR technical information was shipped in late 1956 to IBM's
plants at Endicott and Poughkeepsie, New York. The SR EDP material went to
Poughkeepsie. An assignment was made in February of 1957 to H. A. Miissell of
IBM to review the SR drawings that had been sent to Poughkeepsie. After months
of prodding by Mussell's superior, Phelps, a one-page report was ultimately
issued on July 8, 1957. The report stated :

"Detailed studies of the subject information were not made since it was felt,
in all cases, that the information to be gained would not greatly benefit IBM.
"In all cases, it was felt that the material might have proven more valuable, if :

(1) Information available was better organized.

(2) A key to readily interpreting the Sperry Rand circuit coding were available.

(3) Missing prints and information were available." In September of 1957, Phelps was asked by H. T. Marcy, the head of
engineering at IBM's Poughkeepsie facility : "Are we using any of the data we
obtained from Sperry Rand as a result of our settlement with them"/" Phelps
responded :

■'For a few weeks * * * there was reasonable activity in that various individ-
uals came to the library and studied parts of the data. For some time there has
been almost no activity with these files at Poughkeepsie." In the spring of 1957, IBM was in a "bind" on its high speed printers —
they were wire printers giving IBM "fits from a mechanical performance point
of view". IBM examined the 1956 SR technical information on the UNIVAC high
speed printer, but concluded that IBM should solve its problem by going ahead
with its own development, the chain printer. IBM thus eventually elected to
introduce its very successful chain printer, which was "as different as day and
night, from a functional and mechanical point of view" from the UNIVAC printer.
No other mainframe manufacturer had this choice. IBM also examined 1956 SR technical information on magnetic drum
plating and magnetic drum bearings. However, IBM found no information that
would advance IBM's state of the art and so put the drawings "back in the file
and forgot it". No other mainframe manufacturer had this opportunity. IBM put the 1956 SR information into dead storage in 1958. In 1960 J. Svigals, an IBM employee, wrote to Birkenstock :
"Several years ago, I believe the year was 1956, IBM received detailed descrip-
tive material from Sperry Rand. The material described all of their announced
products at this time.

"There has been little or no use of this material. As part of this department's
File Retention Program, it is desirable that this material be discarded. May I
please have your advice on the disposition of this material. Attached is a list de-
scribing the contents of these files."

The EDP information which IBM had received from SR was subsequently de-
stroyed and the tab information remained in dead storage at Endicott.

5874 IBM could have but never built a copy of any of the SR machines
on which it had received EDP technical information and it could have but never
used or incorporated into any of its equipment, designs, details, features or proc-
esses specifically derived from the SR technical information. No other mainframe
manufacturer had such opportunity.

15.25.4 There was a competitive advantage to be gained by both IBM and SR
by receiving the respective packages of EDP know-how.

15.25.5 That same competitive advantage was also possible for any other EDP
competitor had it received that EDP know-how.

15.25.6 SR and IBM attempted to, and did, gain a competitive advantage by the
receipt and subsequent ability to consider and decide upon utilization of that
EDP know-how.

15.25.7 Then other EDP competitors, by reason of their non-access to the EDP
information, did not gain or have a competitive advantage.

15.25.8 Since SR and IBM gained a competitive advantage and since the other
EDP competitors did not, the others were at a competitive disadvantage com-
pared to either SR or IBM or both.

15.25.9 Thus, the effect of the technological sharing, embodied in the 1956

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 133 of 140)