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 34 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 34 of 140)
Font size
QR-code for this ebook


APPENDIX A (13)



the various segments of the market and the contenders for them.
Appendix B, following, will deal with the derivation and analysis of
those portions of the market which hold particular current and future
interest for the Bendix Computer Division.



40-927 O - pt. 7-24



5198



APPENDIX B (1)



DERIVATION OF THE MARKET

The preceding appendix, discussing the general aspects of recent
computer market developments, provides a background on which to base
an analysis of that segment of the naarket which is of particular interest
to Bendix and to develop a reasonable forecast of numbers of units that
are likely to be installed over the next five-year period.

The market area of specific interest to Bendix, and hence the
subject of this appendix, is the engineering and scientific field of
application and the size categories within this field of application in
which the G-15D, G-20, and G-25 will have to compete.

1. THE ENGINEERING AND SCIENTIFIC APPLICATION NO
LONGER ACCOUNTS FOR THE LARGEST SHARE OF
COMPUTER TOTAL ANNUAL REVENUE

Industry sources, user interviews, and an individual evaluation of
each machine on the market indicate that there has been a substantial
shift in predominant computer application in recent years. The follow-
ing data present an evaluation of this shift in proportion of total annual
rental value, by application. Exhibit B-I, following this page, illustrates
the history of this shift.



5199



EXHIBIT D-I
BENDIX CORPORATION

PER CENT DISTRIBUTION OF TOTAL

COMPUTER RENTAL VALUE BY APPLICATION

1955-1960




1956



1957 1958

YEAR



5200



APPENDIX B (2)



SS-Jt


36%


40


62


2


2



Application 1957 1960

Engineering and Scientific

Data Processing

Process Control

Total lOOiii 100%



On this basis, the estimated $453 million of total annual rental
value at year-end 1960 is apportioned as follows:

Dollars
Application Per Cent (Millions)

Engineering and Scientific 36% $164.4

Data Processing 62 280. 1

Process Control 2 8. 7

Total 100% $ 453.2

Users, who are becoming more experienced and mindful of the eco-
nomics involved, are requiring dual usage of their equipment by their
technical and data processing staffs. In recognition of this, manufac-
turers are making available machines with capability in both areas.
The former areas of design and sales specialization are now obscured,
and the growth rate of the engineering and scientific computer market,
as such, has slowed accordingly. The following table illustrates the
drop in percentage of new annual rental money going to the engineering
and scientific field during the past five years.

1955 1956 1957 1958 1959 1960

New Money

Going to E&S 61% 37% 44% 38% 38% 23%



5201

APPENDIX B (3)



(1) Large Systems Are Used Predominantly in the Data
Processing Field



Less than 42% of the total annual rental value represented
in large systems is devoted to engineering and scientific appli-
cations. In terms of units. 64% of the available large units,
although having some capability in both applications, are aimed
primarily at the data processing field. Many of the most recent
entries into the market are directed exclusively to data process-
ing work. Extreme prices of these imits cause considerable
concentration of rental value dollars in this size category.



(2) Medium- size Systems Are Data Processing-oriented to
a Greater Degree than Any Other Size Category



Only 27% of the funds directed into annual rental value of
medium- size systems are channeled to engineering and scientific
applications. In terms of units, nearly 80% of all available medium-
size systems are applied primarily in the data processing field.
With the exception of the G-20, the tape-equipped Alwac III-E,
and the Ram.o Wooldridge 400, every other medium-size unit is
applied primarily in the data processing area, while having some
capability in both areas. As in the case of large systems, many
of the newest units out are aimed mainly at the data processing



5202



APPENDIX B (4)

field. The percentage of total annual rental value directed to
engineering and scientific application in the medium- size category
has dropped sharply since 1955.



(3) Snaall Computers Still Are Serving Primarily the Engineer-
ing and Scientific Field, but to a Lesser Extent than Before



About 64% of the total annual rental value accounted for in
small computers is going to the engineering and scientific field
of application. Of the available imit models, 59% are applied
priniarily to engineering and scientific applications, but there
is increasing pressure to provide dual capability in this size
category also. The relatively stable proportions of both total
annual revenue and new money devoted to engineering and
scientific applications in this size category in recent years
indicates a rather rapid m.aturing of this market and that future
growth will likely be based on dual capability, rather than appli-
cation specialization. Here again, manufacturers have recog-
nized this and the most recent entries into this market, espe-
cially the IBM 1620, are designed to sei*ve this expanding dual
usage field. There is a much more extensive present and future
market of small users to be served by dual capability, than there
is to be served by specialized equipment.



5203



APPENDIX B (5)



(4) While Small Computers Continue To Serve the Engineering
and Scientific Area Primarily, It Is Evident That Data
Processing Capability Must Be Provided in All Sizes of
Equipment



As pointed out in the 1957-1958 report, the data processing
market still appears to offer the greatest ultimate growth and
potential. With the trend toward dual usage established by a real
demand in the market, it appears that the opportunity for unique-
ness and specialization in the engineering and scientific area no
longer exists, particularly in the medium-size category where
the G-20 will have to compete.

The following table illustrates the recent history of the
percentage of total annual rental value for each size category
applied to the engineering and scientific field.



Medium



1955
1956
1957
1958
1959
1960



63%


5T^


52'5i.


52


52


40


61


46


43


70


41


41


73


33


45


64


27


41



2. THE ENGINEERING AND SCIENTIFIC COMPUTER MARKET
SHOULD REACH APPROXIMATELY $400 MILLION IN
TERMS OF ANNUAL RENTAL VALUE BY 1965



The amount of money spent for the rental or depreciation of
computers to support the work of engineering and scientific personnel



5204



APPENDIX B (6)

remains the key figure in appraising the probable future of this market.
This dollar-expenditure -per-man method was fundamental to the develop-
ment of our 1957-1958 G-15D analysis.



(1) About 798,000 Engineers and Scientists Will Have Use
for Computers by 1965



The National Science Fovmdation has revised its forecast of
the United States' engineering and scientific population slightly
upward since our 1957-1958 report. This has increased the 1960
estimate from 1. 000. 000 to approximately 1, 030, 000, or an in-
crease of about 3%. The 40% exclusion factor, which was used
in our prior study and which was based on the size of companies
employing engineering and scientific groups, and the size of the
groups themselves, still appears to be valid, ■ Application of
this factor to the revised population forecast indicates that the
users' group will expand from about 618, 000 as of year-end
1960 to approximately 798, 000 by the end of 1965.



(2) It Still Appears That Annual Expenditures for Computers
for Engineering and Scientific Use Will Level off at about
$500 per Man



The actual rate of growth in expenditures per man has
fallen somewhat short of our 1957-1958 projection. The 1958
general business readjustment, manufacturers' inventory



5205

APPENDIX B (7)

adjustments, and the curtailment of new computer commitments
during the change-over from the earlier tube and drum machines
to the new transistorized equipment caused a temporary decelera-
tion of growth between 1957 and 1961. This has resulted in a delay
in achieving the $500-per-man level anticipated in our earlier
analysis.

Based on current industry data, it appears that the market
will resume its expected growth rate in 1961 and the $500-per-man
level will be reached in about 1965. instead of 1962. Exhibit B-II,
following this page, consists of three curves reflecting:



The projected growth in expenditures per engineer
and scientist from our 1958 report.

The actual growth to date and the resulting historical
trend.

Our best estimate of what will probably occur
between now and 1965.



(3) An Estimated 798, OOP Potential Users at $500 per Year
Each Will Provide a $400 Million Market



The combination of 798, 000 eligible users of engineering
and scientific computational aids, and an expenditure level of
$500 per man in 1965, indicates an engineering-scientific market
of approximately $400 million five years hence. Current expendi-
tures are estimated to be $212. 6 million, including approximately



5206



EXHIBIT G-II
BENDIX CORPORATION

GROWTH PATTERNS OF ENGINEERING

AND SCIENTIFIC MARKET BASED ON

HISTORICAL EXPENDITURES



'i t;rtt'>I'b BT:? + i.i



















.sot^'


. - "


'7¥-


1 ^^^
iiiiii"

UTREN


^
















r«!^


>^%


tVo«Vc.


D












J


/


^-




















i


/


/
























1




























1
























































1


























/




























/


























y


f


























/


























/


f- —



























5207



APPENDIX B (8)

$48 million for punched card calculating equipment, which will be
subject to replacement by computers during the next five years.
Exhibit B-III, following this page, shows this projection graphically.
Note that growth slowed significantly during 1958, was relatively
normal in 1959, and slowed again in 1960. The reasons for these
two cutbacks in new commitments were discussed earlier. We
would anticipate that 1961 will be a good year for the industry,
since the second generation computers will be available for
delivery. The market should continue to mature until most of
the potential computer requirements have been filled.

Exhibit B-IV, following Exhibit B-III, consists of a com-
parison between this revised estimate and the forecast contained
in our 1957-1958 report.

(4) A Second Forecast, Based on Federad Research and

Development Expenditures, Indicates a Total Engineering
and Scientific Market of About $380 Million in 1965

In order to cross-check the estimate derived from our
"dollar-per-man" technique, a second, independent forecast was
developed based upon federal R&D expenditures. Historical
annual R&D expenditures and E&S computer expenditures were
reduced to indexes, in an effort to identify any direct correlation



5208



HIBIT R-IU
BEN^.X CORPORATION

HISTORICAL AND PROJECTED GROWTH

OF THE ENGINEERING AND SCIENTIFIC MARKET

1953-1965




100 200 300

ENGINEERING AND SCIENTIFIC EXPENDITURES ($ MILLION)



5209



F.xiiiniT n-iv

BENDIX CORPORATION

ANNUAL EXPENDITURES FOR ENGINEERING AND
SCIENTIFIC COMPUTERS AND
PUNCH CARD CALCULATORS



CUMULATIVE DOLLARS

AT YEAR END

(MILLIONS)

















































(1957-58 ^V^




/■




















E


STIMATE)


\,


























Jr


^»*»*


























/


>*S


























/>-1
























i^^'


























V





























/s^


k*
























i> s




















1957-58 ESTIMATE


/


/






















\




/ >


A
























k.


* ^


X
























^1


^


N


^






















/


^^


NEW ESTIMATE |




















/


/


























^ /
























i


/I


























i


/


























/


/


























/ .


^


























t J




























t-^^


























/]


Y


























/.


f


























//




























*/


























,


f/


























/


f
























J


/


















— •


^


y


/


ACT


JAL

















1958 1959
YEAR



5210

APPENDIX B (9)

that might exist. The rapid growth of the computer field in recent
years and the size and growth rate of the federal R&D budget have
not borne a direct historical relationship. Analysis of the per-
centage relationship between computer expenditures and the budget
does indicate that the former has rapidly risen to a level of about
2. 4% and leveled off.

If we assume that this percentage will rise to 2. 5% and
apply it to an estimated 1965 R&D budget of $15. 3 billion, it
resvilts in a 1965 E&S computer expenditure level of about $380
million. The basis for this forecast is shown as Exhibit B-V,
following this page. The closeness of this forecast of $380
million and the base forecast of $399 million arrived at by the
dollar -per -man method leads us to believe that the attainment
of these levels is well within the realm of reason.



3, IT APPEARS THAT THE ENGINEERING AND SCIENTIFIC
MARKET WILL ABSORB BETWEEN 1800 AND 2300 SMALL
COMPUTERS BY 1965



The market for small computer units for the next five years is
made up of two parts: new expenditures in the engineering and scientific
area for small computers, and replacement units.



5211




5212

APPENDIX B (10)



(1) New Expenditures in the Engineering and Scientific Area
Are Forecast To Reach about $186 Million in the Next
Five Years



By applying a sliding scale percentage of new funds going
into small computers, based on historical performance and
anticipated performance, it can be determined that about $21
million, or about 11%, of these new funds will be channeled into
the small computer area. Based on an approximate annual rental
figure of $18,000 for a small engineering and scientific computer,
it is estimated that between 1, 000 and 1, 500 new units will be
absorbed into this market by 1965.



(2) There Is an Additional Market of about $214 Million ,
Based on the Replacement of Some Existing Punched
Card Equipment by Computers and the Obsolescence
of Existing Computer Installations



In addition to the new engineering and scientific expendi-
tures forecast of $186 million, there is a replacement market,
made up of punched card calculator installations that will be
replaced by computers, and of present computer installations
that will become obsolete for technical or economic reasons,
amoxmting to about $214 million. The combination of these two
markets gives us our 1965 total engineering and scientific market
of about $400 million mentioned previously and portrayed graphi-
cally in Exhibit B-III of this appendix. It is not reasonable to



5213

APPENDIX B (11)

expect that all calculators and piinched card installations will be
replaced by computers within the next five years, and some com-
puters that have been purchased outright will not be subject to the
normal four- to five-year obsolescence factor, but for purposes
of this analysis, it is reasonable to expect that most will. If
about 7% of this $214 million engineering and scientific replace-
ment market is channeled to small computers, as has been
historically the case, and this result is translated into units,
there are 800 additional engineering and scientific units to be
considered as part of this raarket. This then gives us a total
of 1,800 to 2,300 small computer units, new and replacement,
to be absorbed into the engineering and scientific market by 1965.



4. IT IS ESTIMATED THAT THE ENGINEERING AND SCIENTIFIC
MARKET WILL ABSORB BETWEEN 700 AND 800 G-20 TYPE
UNITS DURING THE NEXT FIVE YEARS



This market likewise consists of two segments, new engineering
and scientific expenditures and replacement expenditures.



(1) About $68 Million of the $186 Million Total Engineering
and Scientific Expenditures Is Forecast To Go for
Medium- size Computers



By applying a percentage, based on historical and antici-
pated performance, to the total anticipated annual expenditures
for engineering and scientific computers, we are able to estimate



40-927 O - pt. 7-25



5214



APPENDIX B (12)

that approximately $68 million of this new money will be spent for
medium-size computers - the area in which the G-20 will have to
compete - by 1965. Based on the annual rental price of a repre-
sentative system, it is estimated that this amoxint represents the
addition of about 450 new engineering and scientific units during
the next five years.



(2) There Is an Additional Replacement Market of about
300 Units



As in the case of small computers, not all the medium
units now in use will be replaced, but it is reasonable to assume
that most will be replaced or upgraded into large systems. Based
on historical relationships of the proportion of engineering and
scientific funds allocated to medium-size units, we estimate that
it is reasonable to expect that approxinaately 300 of the engineering
and scientific units now installed will be subject to replacement
during the next five years. This replacement market, combined
with the anticipated new expenditures, indicates a medium- size
engineering and scientific m.arket equivalent to about 750 G-20
type units.



5215



APPENDIX B (13)

5. THE REMAINDER OF THE MARKET INDICATES A TOTAL OF

300 TO 400 LARGE ENGINEERING AND SCIENTIFIC UNITS WILL
BE ABSORBED BY 1965

On the basis of historical and anticipated relationships demonstrated
in the small- and medium- size categories, our analysis indicates that
about $97 million of the $186 million expenditures for engineering and
scientific computers will be concentrated in large scale units. Based
on representative prices of large units, it is estimated that about 250
units will be added to the naarket by 1965. In addition to this, there
should be a replacement market of about 100 engineering and scientific
units, indicating that the market will absorb the equivalent of about 350
large scale engineering and scientific units in the next five years.
Exhibit B-VI, following this page, illustrates the segmentation of the
engineering and scientific market in terms of units and rental value.



This appendix has presented our best estimate, derived by two
independent naethods, of the probable size of the m.arket for engineering
and scientific comiputers between now and 1965. This base estimate has
been further broken down in terms of new (net additions to the population)
and replacement units within the large, medium, and sniall categories.



5216



KXMiiiir ii-vi

BENDIX CORPORATION

ENGINEERING AND SCIENTIFIC COMPUTER

MARKET POTENTIAL

1960-1965



MEDIUM
$437 MILLION




SMALL
$152 MILLION



LARGE
S611 MILLION



TOTAL SALE VALUE OF MARKET $L2 BILLION



5217



APPENDIX B (14)

It should be noted that although these various estimates appear
reasonable today in light of the historical and probable future develop-
ments within this market, they are only estimates and should be sub-
jected to periodic review in order to determine the impact of events
that are not foreseeable at this time.

These estimates provide the basic market facts requested by
Bendix in accordance with our original proposal. Exhibits B-VII, B-VIII,
and B-IX, following this page, present summary tabulations of the basic
statistical data developed during the course of this study and upon which
the market estimates were based.

Appendixes C and D which follow, present comparative technical
evaluations of the Bendix G-15 and G-20 computers. These have been
developed to aid in the determination of the relative competitive position
of current Bendix products.



2 <
§3^



52 »

2g fe



5218



1^



I i



5219



•s a

= 11

g " g

^ a *

1 1^



2 5;=^
m
o

<



r> in


CD


c-


CO


o


o


to


CJ


o


5 c? w

>■

O


^


"




Tf


??


r-


CO


M












S
1


















!• S^


O


o


o


o


o


o


o


O


CO


o


c-


CD


t^


■*


o


r-i



<^-g




€ -Q



- 3 S

3 i s



CO ^



J3 4)

i-S



s s



D. O

^ o
.9 >



5220



EX IT B-IX
BENDIX jRPORATION

PROJECTED PERCENT DISTRIBUTION OF

NEW ENGINEERING AND SCIENTinC FUNDS, BY SIZE

1960-1965




AflCEi AVERAGE t



1961 1962



1964 1965



5221

APPENDIX C

APPENDIX C (1)
COMPARATIVE TECHNICAL EVALUATION
OF THE BENDIX G-15 COMPUTER

As an aid in the evaluation of the probable future market penetration
of the Bendix G-15 computer, a comparative study was made of its features,
performance, and capabilities against those of other units with which it
must compete. This study was done in cooperation with the Bendix Com-
puter Division marketing and engineering departments in order to benefit
from all available data and to minimize possible duplication of effort.

The highlights of this evaluation are presented on the following pages.

1. THE G-15 HAS BEEN MOST SUCCESSFUL IN THE ENGINEERING -
SCIENTIFIC MARKET

Of the three major areas of application for small general purpose
computers - engineering-scientific, data processing and process control -
the G-15 has achieved its most notable success in the technical area.
This has corae about as a result of the closeness of fit between the original
design of the equipment and the needs of this segment of the market. In
addition, the sales efforts of the division have largely been directed to
this type of user.



5222



APPENDIX C (2)



(1) The G-15 Was Developed To Fulfill the Needs of the
Engineering-Scientific User



The G-15 has a number of inherent design features which
make it particularly well suited to the solution of complex engi-
neering and scientific problems. These features were selected
during the design process in order to improve the usefulness of
the machine to the technical user, while keeping the price below
that of equipment designed to be equally adept at scientific or
business data processing. Specifically, these features include:

Highly flexible internal computing facilities.

A wide range of available machine commands.

Augmented storage capacity through inexpensive magnetic
tape units.

Availability of the DA-1 digital differential analyses and
PA-3 graph plotter accessories.

High-speed punched paper tape input and output.

Sixteen words of fast access memory.

Each of these features, along with the interpretative routines and
machine language programs available from Bendix and through the
G-15 users' organization, has contributed to the usefulness and
customer acceptance of the G-15 over the past four years. Approxi-
mately 90% of the more than 300 machines sold or leased to date are
principally used in the solution of problems of a technical nature
that call for the capabilities provided by these features.



5223



APPENDIX C (3)



(2) The Suitability of the G-15 to Business Data Processing
Problems Has Been Improved, but Its Usefulness in this
Area Is Still Limited



In an effort to increase the usefulness of the G-15 for dual
engineering-scientific and business data processing applications,
and thus expand its market, Bendix has added several important
features. These include the CA-1 and CA-2 punched card input-
output accessories and alpha-numeric data-handling capabilities.
While enhancing the acceptability of the system to the user with
limited data processing requirements, the G-15 is still not competi-
tive in those instances where 50% or more of the work load is of a
data processing nature. This limitation stems from the relatively
slow speeds of the card -handling and magnetic tape units and the
disadvantages of binary machine language for simple accounting-
type problems.

(3) The G-15 Has Achieved Little Success in the Process
Control Area

The G-15 is not wholly suited to process control applications
because of its limited number of available input-output connections.
The few sales that have been made in this area have been in support
of tape controlled machine tool installations where the computer is
used as a tape preparation device rather than as an on-line control
unit.



5224



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 34 of 140)