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 13 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 13 of 140)
Font size
QR-code for this ebook


Most users must commit themselves for at least five years
to win significant cost advantages on 370s; during 1974
many lessors will stop writing even five-year operating
leases and deal only in full-payout packages.



4979



Greyhound Computer Corporation

DPF Incorporated

Itel Corporation

Boothe Computer Corporation

Randolph Computer (First National
Bank of Boston)

Diebold Computer Leasing

Rockwood Computer Corporation

Reliance Group (Leasco)

Computer Leasing Company (Wyly)

Granite Computer Leasing

National Computer Rental

Dearborn-Storm Corporation (original o\>mer)

Talcott Computer Leasing

Others (including full-payout
leasing by banks)

TOTAL



$ Value of 360 Portf olio~U .S.
($M Original IBM Purchase Price)

$ 295
225



215
205

175

170

155

150

105

100

90

80

60

715



$2,740



+ Lessors are depreciating 370 equipment over an expected
useful life-span of eight years or so instead of the un-
realistic ten to twelve years most lessors counted on
for the 360 generation.

+ Financially sophisticated firms with experience in com-
puter leasing are turning to lease packaging, lease broker-
ing or "agency" leasing — arranging contracts between users
and investors on a fee basis and avoiding direct ownership
of large mainframe portfolios.

Probably the prime example of the new breed of leasing companies
is an old leasing company — Itel Corp. — which was an early high-
flyer in the 360 leasing business. Itel took its losses, revamped
its philosophy and has bounced back to become the largest manager of
leases on 370 equipment — $300 million worth so far, of which about
a third is plug-compatible equipment including memory units from
AMS and disk drives from its former manufacturing subsidiary, ISS
Richard Lussier, president of Itel Data Products, expects these
bookings to double to $600 million by yearend 1975.



4980



In making the jump from 360 to 370 leasing, Itel has transformed
itself from an owner/lessor of computing equipment into a "financial
and business services" organization. The company lines up lessees,
packages leases combining IBM CPUs with Itel peripherals, and finally
finds investors and lenders to finance the deals.

Aside from its overhead, Itel invests no money in these fine-tuned
financial packages. As the lease manager, it collects a percentage
of all lease payments as a fee and takes a share of residual value
of the equipment at the end of its leasing life. The leases also are
vehicles for the sale of the compatible peripherals Itel markets —
but no longer manufactures. Thus Itel has largely removed itself
from the arena of risk, but still makes money from its financial
expertise.

The economic life-span of the 370 family is one of the biggest
questions confronting the leasing industry. Every lessor is prepared
to see IBM introduce a new and possibly very different family of
systems by 1977 — but they differ widely in their views of the impact
FS is likely to make on the market. Itel spokesmen point out that
users are just beginning to use virtual storage and counts the intro-
duction of VS as the start of a new IBM product cycle — betting that
IBM won't take serious steps to obsolete VS equipment much before the
end of the decade. "The U.S. will provide systems growth for the
indefinite future," says Lussier. "The installed base will stay viable
for a long time no matter what FS looks like." The point is important
to lessors not only because they will need to extend or re-lease many
of their current contracts in order to recoup their investments, but
because most have a stake in the residual value of their equipment.
Generally they assume the equipment will really be worth quite a bit
more at the end of their payout periods than the minimal salvage
value they've allowed for auditing purposes.



4981



PLUG-COMPATIBLE MANUFACTURERS



Differences exist between IDC's estimates of the plug-compatible
peripheral market (PCM) and those of IBM's as presented in the written
decision formulated by the U.S. District Court, and repeated here in
total for reference. While IDC's estimates of the entire 360/370
tape drive and disk drive base coincide very closely to IBM's estimates,
the two differ significantly for PCM penetration.

IDC's estimates are based on continual tracking of various prod-
uct lines by individual manufacturers, and on data derived from its
proprietary Computer Installation Data Files which accurately iden-
tify the manufacturer, models, and quantities of plug-compatible
peripheral equipment installed.

While IDC has not yet been able to reconcile why these apparent
differences exist, it nevertheless has a high degree of confidence
that its estimates are accurate, after being carefully verified
against several internal and external sources, including all of
the major PCM manufacturers.

The IBM plug-compatible peripheral market began in 1968 when
Telex tape drives and Memorex disk drives first replaced IBM counter-
parts on 360 computers. Initially created as a means of solely
providing IBM 360 users with a comparable product at a lower cost,
the market is characterized today by plug-compatible products that
also offer performance improvements over IBM counterparts in many
instances, with reliable company support and service behind them.
The market is still in its growth cycle, despite efforts by IBM
(brought out in the Telex case) to counter penetration of its
peripheral base.

As shown in Table 119, by yearend 1972 plug- compatible manu-
facturers (PCMs) had installed 11,050 tape drives and 17,595 disk
drives worth $798 million on IBM 360/370 systems in the U.S.,
giving PCMs a 23.6% and 21.3% share of total tape and disk drives
in use, respectively.



4982



TABLE 118



IBM 360/370 PLUG-COMPATIBLE TAPE & DISK DRIVE MARKETS - U.S.

12/72





Total No.












Drives


No. IBM


No. PCM




Value*




In Use


Drives


Drives


% Total


($M)


Tape Drives


46,795


35,550


11,050


23.6%


$276


Disk Drives


82,550


64,955


17,595


21.3%


$522



TOTAL 129,345 100,505 28,645 22.1% $798

* @ IBM original purchase price.



Growth in various areas within each of these market sectors is
varied due to obsolete equipment that is being replaced by new IBM
peripheral models and/or new PCM models.



Disk Drives

The plug-compatible disk drive market is dominated by three
suppliers who held 74% of the total number of drives installed at
yearend 1972 — Memorex, CalComp (Century Data), and Telex. Memorex
was the pioneer in this market with its 2311 version, and still holds
a 30% share overall; however, weak financial conditions at Memorex
for the past few years have caused the company to lose ground to
both CalComp and Telex — moreso the latter. Table 120 summarizes
overall disk drive shares by manufacturer or supplier

Tape Drives

Storage Technology, by virtue of heavy 3420 shipments last year,
and few replacements of earlier generation STC drives, holds the
number one position in the 360/370 PCM tape drive market with 3100
drives installed.

Telex, with a broad customer mix in all three tape drive product
areas, has experienced significant returns of its earlier generation
drives, which has resulted in a significant loss of market share over
the past two years. (Two years ago it had about 48% of the market.)
However, substantial shipments of 3420s should help maintain market
position.



4983



TABLE 119



PCM MARKET SHARES

IBM 360/370 PLUG-COMPATIBLE DISK DRIVE MARKET - U.S.

12/72



Memorex

CalComp

Telex

Potter

Marshall

Ampex

Others

TOTAL



No. PCM




Drives




In Use


% Share


5,290


30%


4,025


23%


3,730


21%


650


4%


1,550


9%


1,700


7%


1,050


6%



17,595



100%



TABLE 120

PCM MARKET SHARES
IBM 360/370 PLUG- COMPATIBLE TAPE DRIVE MARKET - U.S.
12/72







No. PCM








Drives








In Use


% Share


Storage


Technology


3,100


28%


Telex




2,900


26%


Potter




2,900


26%


Ampex




1,150


10%


Others




1,100


10%



TOTAL



11,050



100%



4984



MINICOMPUTERS



Growth in numbers of minicomputers shipped continued to show an
increase in excess of 50% in 1973, and will maintain levels of 35%
or more through 1976, sustained by volume microcomputer shipments
beginning in 1975. New applications continue to emerge, and exist-
ing users are demanding more and more equipment. The composition
of this market is changing slowly, but DEC remains the leader.



Market Definition

For the sake of review, here are IDC's definitional criteria
for minicomputers, expanded to include microcomputers, which are
simply microprocessor-based minicomputers.

+ The processor is general purpose by design and is sold
by the manufacturer as a minicomputer, not as a terminal,
small business system, or other computer-based product.

+ The current basic price range for a CPU with 4K words
of memory is generally $3,500 to $25,000. Downward
extensions of product lines that fall within this price
class — the PDP-8/A, LSI-12/16, and Naked Mini ~ are
also considered to be part of this market.

+ Word size is typically between 8 bits and 24 bits,
although some 32-bit models exist.

+ Memory is typically expandable from 4K to 32K, but
some models have capabilities outside this range,

+ Software and peripherals available from the same

supplier allow these units to become complete systems.

Table 121 lists all the commercially available minicomputer/micro-
computer models that IDC includes in its market estimates. Hewlett
Packard's Series 3000, for example, is purposely omitted from the
list because its base price is higher than $25,000.



Market

Shipments of minicomputers in 1973 represented a 58% increase
over 1972. Continued rapid growth — compounded at over 30% per
year — will generate for U.S. manufacturers a total market of $2.67
billion in 1978, based on two major factors.



4985



TABLE 121
LIST OF MINICOMPUTER MODELS



CoTOiler SyiC«««. Inc.



Coocrol Dica Corporstloi



DatAcraft

Data Gcnaral CorporaclO'



DlBlae Corporation

Dl|ltal Coaputer ConCrola



lOB/208/808
116/216/8U
Alpha-8/iapha-16
Naked KiDl-8/16/LSI

1700-SC
Elhlt-lOO
Syacaa IT



«-800/820/860



DH-70/W
ECP-18



crodata CorporadOD



HlalCkMqt

Nodular CoapuCar Syacaaa

Hotorola CoMunlcatlona Dtv
Nuclaar Data. Inc.
Omlconp Coaputar Corporatli
Qaniccc (BIT)



74
80
85

7/16. 7/32

HAC-16
MAC-Jr.
SUE-nO

400

800/810/612/820

1600

3200/3230

32/S



ND-812



Digital Equipment Corpon



Digital Sciantlflc Corporac:
Electronic Aaaoclacaa, Inc.
Electronic Procaaaora, Inc.

Ccneral Aucoaaclon, Inc.



CE/PAC

GRI Computar Corpon



PDP-8/8e/8f/81/8&/81/6a/BA

PDP-9L

PDP-ll/05/15/20/ftO/35/'.5

PDP-12

PDP-15



EPl-118/218
Cent auar- 100



CA 18/30
SPC-12
SPC-16
LSI-12/16



II

2100-A

21K-A/B

2U5/A

2n«-A/B/C

2100-S

21Z3-A

112



lipa



Radcor Corporation
Sciantlflc Control Cc
Splraa Syataaa/USH
SYS CoBpucar Corporal

Syatana Englnaarlni 1



RDS-50O
RC-70



960

960-A

980

»eo-A

CP-8A/B/C/D

CoBip-l6/18

520/1
610
620
620/i/£/l



. Technology, In<



Uesttnghouae Ele



4986



+ Microcomputers will allow manufacturers to take advantage
of the extreme demand elasticity in this market.

+ Peripheral equipment will provide the means for increas-
ing dollar shipment volume to both existing minicomputer
users and to other markets currently supplied by mini-
peripheral companies. With far less price erosion, mini-
computer manufacturers will find this market sector quite
attractive.

By 1978, minicomputers will be buried in so many products — both in-
side and outside the traditional computer industry — that the myriad
of future applications would be impossible to enumerate.

Within the computer industry, however, IDC anticipates the
effect of network development provides the greatest source of poten-
tial over the long term. Minicomputers will serve as various types
of components.

+ Large computers will evolve into a series of minicomputer-
like modules.

+ Peripheral equipment and network interfaces will consist
of minicomputer-based controllers.

+ Remote subnetworks will utilize minicomputers as pre-
processors.

And so on. Some will be oriented toward special applications, like
POS systems; others will be general purpose. But the concept will
focus on the use of minicomputers.

Thus, by 1978, worldwide shipments by U.S. manufacturers will
reach about 144,000 units. Growth will still be at the 20% level,
with add-on and OEM peripheral equipment revenues offsetting the
declining processor prices to keep dollar shipments growing at
about the same rate as units.

The international market will continue growing at a faster pace
than the U.S. as indicated by Figure 114. Although this sector of
the computer industry is one of the few that is not dominated over-
seas by U.S. suppliers, these familiar names do hold quite signifi-
cant shares at least of the European market. Most of the business
there is end-user shipments because no single market is large enough
to support the myriad of systems houses that have sprung up in the
United States. Thus, the average system value shipped abroad is
somewhat higher than at home.



4987



Who needs all this processing capacity? The list of applica-
tions continues to grow. About 60% of the minicomputers shipped in
the U.S. (worth almost half the total value) go first to an OEM
manufacturer for inclusion in shared processor (data entry) systems,
intelligent terminals, turnkey systems for small business, and a
growing number of consumer-oriented (autotransaction) products such
as point-of-sale systems, automated bank tellers, credit authoriza-
tion systems, and even automated gas stations.

Minis are used in industrial automation across the board, as
concentrators in teleprocessing networks, for plant security, and in
special applications such as traffic light control. Furthermore,
IBM's entry — the sensor-based System/7 — is broadening the market
rather than increasing competition as it connects more and more
"live" applications to centralized System/370s.

Some dedicated application computers, such as IBM's System/7,
can be relatively expensive; others, like those from Microdata, very
inexpensive. Across-the-board suppliers such as Digital Equipment
have a wide mix, so market share tends to be the same by number of
dollars. The primary reason average system cost continues to remain
upward of $20,000 is because of peripherals, even though a number
of suppliers — Caelus, Centronics, Dataproducts, Diablo, lomec ,
Pertec — are specializing in miniperipherals . Here's a breakdown
of system shipments for the major suppliers of dedicated application
computers in 1973:

TABLE 122



Company


# Units


% Total


$M Value


% Total


Digital Equipment


10,800


31.9


$


242


31.4


IBM


1,500


4,4




90


11.7


Hewlett Packard


2,450


7.2




83


10.8


Honeywell


1,060


3.1




69


9.0


Data General


A, 400


13.0




58


7.5


General Automation


2,550


7.5


$


44


5.7


Varian


1,280


3.8




32


4.2


Interdata


780


2.3




19


2.5


Computer Automation


2,730


8.1




15


1.9


Xerox


140


0.4




14


1.8


Systems Eng. Labs


85


0.3


$


13


1.7


Modular Computer


485


1.4




12


1.6


Microdata


1,550


4.6




9


1.2


Digital Computer Cont


900


2.7




5


0.6


All Other


3,190


9.3




65


8.4



TOTALS



33,900 100.0



$ 770



100.0



4988



Mini manufacturers have been trying to make life easier for
end-users in recent years by adding high-level language compilers,
beefing up operating systems and improving service. But in general,
manufacturers haven't been turning out systems ready to go to work
for the end-user. The user's only choices are to buy a mini and
program it himself, buy his application software from another source,
or find a complete design to fill his particular need. The complete
"turnkey" system is the easiest way out — but since mini makers
lack the expertise and the capital to turn out such specialized
products, they usually aren't available directly from the manufacturer.

This is where the OEM builder will play an increasingly impor-
tant role in the growth of the mini maker. IDC predicts, in fact,
that two- thirds of all new minis will be shipped on an OEM basis
by 1977. By marrying minis to end-user equipment — ranging from
data entry systems to traffic lights — OEMs will bridge the all-
important applications gap between user and machine. In many such
systems, minis will be nothing more than components buried in larger
products.

Mini applications are hard to pin down into categories, but IDC
has come up with one approach based on the visibility of mini-
computers to their users in the applications concerned:

+ General-purpose minicomputers are purchased by end-
users from either a mini manufacturer or systems
house for a customized application. The computer
is highly visible to the user, who usually does
the programming himself. Problem solving/analysis
systems fall into this category.

+ Stand-alone commercial data processing systems also
are highly visible — but come with packaged soft-
ware for general accounting/payroll, etc., within
small companies, or other packages for use in a
specific sector of a large company's operations.

+ Large computer support systems incorporate minis
into some other data processing gear such as key-
to-disk systems, graphic displays and plotters,
or communications controllers.

+ Automation and control systems combine minis with
a non-computer product to improve the efficiency
of the system. Process and quality control are
major applications; turnkey systems for traffic
control and baggage routing are two early examples
of mini-supervised products that are beginning to
move automation out of the factories.



4989



Significantly, the number of minicomputers in use in each of
these application categories goes up as "visibility" decreases —
the "out-in-front" minicomputer is relatively rare compared to
the number buried in OEM products, as the following tables show:



TABLE 123

U.S. USER-PROGRAMMED AND CUSTOM-DESIGNED MINICOMPUTERS INSTALLED

1973 1974 1975 1976 1977

Problem Solving/Analysis 6,000 7,150 8,450 9,885 11,450

Commercial Data Processing 2,600 3,600 4,875 6,535 8,630

Large System Support 7,675 11,900 17,855 25,890 36,250

Automation and Control 9,425 13,200 17,550 22,465 28,070

Mixed 600 900 1,250 1,650 2,050



TOTAL 26,300 36,750 49,975 66,425 86,450

% Total Minis Installed 40% 37% 35% 33% 31%



TABLE 124
U.S. MINICOMPUTER-BASED TURNKEY SYSTEMS INSTALLED

1973 1974 1975 1976 1977

Large System Support 5,025 8,000 12,325 18,550 26,750

Automation and Control 29,675 46,250 68,700 98,625 132,425

Commercial Data Processing 5,000 8,000 13,000 20,000 30,000

Consumer Products _ _ _ AOO 1,375



TOTAL 39,700 62,250 94,025 137,575 190,550

% Total Minis Installed 60% 63% 65% 67% 69%



40-927 O - pi. 7 - U



4990



While these figures reflect changing patterns of minicomputer
use, they do not represent corresponding shifts in revenues for
mini manufacturers. OEM minis are sold at large discounts —
prices are falling rapidly — and tend to be small machines. End-
user systems, on the other hand, are growing more powerful — edging
toward full scale capability in some cases — and the average value
of new end-user systems is declining much more slowly than average
OEM values.

Despite the problems involved in producing turnkey systems for
specialized uses, a number of prospective markets for such systems
are so large and potentially profitable that mini manufacturers
may well be tempted to develop products for them. Some of the
most promising of these sectors are:

+ Printing and publishing, one of the first sectors to
use minis extensively. DEC is producing a complete
typesetting system for publishing houses and diversi-
fying its system for other uses. Some systems in
the works will completely automate newspaper composing
rooms, processing copy all the way from a reporter's
CRT to the printing press.

+ Gas stations and other businesses with dispersed

customer bases. Large mini companies are in a better
position to provide service to such customers than
most OEM houses. Mini-based traffic light systems
have proven very efficient in early trials. Auto-
tellers for branch bank sites may also become popular
mini products.



4991



MAINTENANCE OF COMPUTERS



Customer service has traditionally been a costly, difficult, and
complaint ridden business, and data processing equipment maintenance
is certainly no exception. EDP equipment manufacturers - both large
and small - are confronted with the problems and expense of providing
maintenance services for their customer base, regardless of whether
they are an established company with a nationwide network of customers,
or a new or small company struggling to expand its base.

In addition, other equipment outlets such as third-party leasing
organizations must be able to offer their customers cost-effective
ways in which their equipment can be maintained. And of course, com-
puter users who own their systems require reliable, efficient mainten-
ance services at the lowest possible cost. Few users can provide their
own.

+ For the large supplier, the penalty it must pay for its
widely dispersed customer base is the necessity to pro-
vide adequate maintenance for these users - reasonably
good response and quality of service. For many, it be-
comes economically infeasible to establish a service
office in certain areas where the number and/or size of
customers is limited.

+ For the small or new equipment manufacturer, a reason-
able level of service support is required concurrent
with initial equipment deliveries. The cost justifica-
tion for establishing its own internal service force is
generally totally impractical until a reasonable number
of customers in given areas is built up. The point at
which it becomes feasible varies dramatically, of course,
depending on the sophistication and price of the equip-
ment.

+ For the leasing company - namely, holders of IBM sys-
tems - its marketing edge is lower hardware prices.
Discounted maintenance rates can only serve to enhance
the marketability of its systems.

The overall outlook for third-party maintenance within these
three major industry groups appears bright. Within the past two years,
the participants - as a group - have proved that (1) maintenance by
a third party can be extremely cost-effective, and (2) it oftentimes
can be far better than that provided by the original equipment manu-
facturer. One of the biggest obstacles for any concept in its embry-
onic stage is establishing proper credibility - the independent ser-
vice vendors appear to have overcome this gap.



4992



Users of their services - both the OEMs themselves and the
equipment end users - indicate that they have been pleased with the
quality and responsiveness of service. What remains then is continued
geographic expansion and development of a broad level of expertise to
enable them to be as flexible as possible, adapting themselves to the
needs of the various industry markets. There is no indication at this
time that this direction cannot be successfully followed.



4993



DATA COMMUNICATIONS



Among the more adventurous, forward-looking computer operations,
the advantages of using data communications techniques have gained
increasing visibility for the last five years or so. But recent
events have marked the beginning of an exceptionally high growth
period for this sector.

+ IBM — with three new products, the 3735 printer process-
ing terminal, the 3705 communications controller, and the
370/158 and 168 virtual storage computer systems — has
apparently started to take an active interest in communi-
cations network implementation.

+ MCI's first operating link — from Chicago to St. Louis —
and the threat of imminent competition elsewhere brought
retaliatory announcements from AT&T regarding the avail-
ability of its Data Under Voice (DUV) technique and
Digital Data System (DDS) network, beginning in 1974.

+ Point-of-sale and other autotransaction systems have
begun to achieve significant popularity, suggesting the
potential of special-purpose terminal systems in many



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 13 of 140)