Volume
9, #7____________________________________________________________
Part 2 - July; 1990
"MGMT MEMO"
was written by Richard Seltzer in Corporate Employee
Communication for the Office of the PresiÂdent. It was
written for Digital’s managers and supervisors to help
them understand and communicate business information to
their employees. You can reach Richard at seltzer@seltzerbooks.com
This is
part two of our coverage of the State of the Company
Meeting, which was held on May 3, 1990 at Merrimack,
N.H. The first part dealt with the first two hours of
the program, which were broadcast live over the Digital
Video Network (DVN) and addressed matters of general
employee concern. This issue summarizes the rest of the
program, which dealt with strategies and markets and the
need of "providing one voice to all audiences."
Client/Server
Computing by Bill Demmer, vice president, Mid-Range
Systems Business
Databases by Hans Gyllstrom,
Manager, Database Systems
Transaction Processing
by Dennis Roberson, group manager, Transaction Processing
Systems Group
Workstations by Don Gaubatz,
group manager, Workstations
ULTRIX by Kurt Friedrich, group
manager, Open Software Group
One Powerful Voice To
Customers by Peter Zotto, manager, Product Communications
A Worldwide Educational
Vision by Pat Cataldo, vice president, Educational
Services
DECWORLD ’90 by Pat
Zilvitis, chairman of DECWORLD ’90
The public
policy of governments is becoming increasingly important
to the success of Digital’s global business.
Public
policy is simply what a people of a country want or do
not want, as expressed through their central government.
Public policies are expressed in laws or regulations, as
well as official policies, settled practices, unwritten
laws, and customs. Basically, public policy is an
expression of self interest on the part of a people, a
country, and its government.
The major
world changes that we read about in the news — the
collapse of the Berlin Wall and the shredding of the
Iron Curtain — signal a new era of more varied and
intense relationships between governments and companies
throughout the world as trade rivalries and alliances
replace national security rivalries and alliances.
The forces
that are shaping and driving governments in Central and
Eastern Europe are affecting almost every government
throughout the world. The major change is the
accelÂerating expectations of citizens. They want an
elected government that is "user friendly", that can
anticipate their needs, provide them with services, and
empower them as indiÂviduals.
Second,
most people around the world — not only those in the
Western democracies — are joining the "consumer
revolution". Citizens of the emerging nations want a
Western, free-market quality of life. There are dramatic
changes in how consumers view their needs and in how
governments view those expectations. In turn, these
developments drive changes in world business patterns
and in the dynamics of the global market. These changes
are magnified and accelerated by the rapidity of
technological change, particularly in the information
technology business.
All of
these changes in dynamics will drive increasing and more
widely varied interactions between governments and
global companies, each of which must act in its own self
interest. How Digital responds to government policies in
the many countries where it does business will be
critical to its success in market areas around the
world.
The rate
of change will continue to be remarkable. One should
remember that both governÂments and companies like
Digital are striving to serve the same people, the
citizen/con- sumer. As a practical matter, consumers and
customers want an information technology company that is
"user friendly", that can anticipate their needs,
provide them with apÂpropriate services, and thereby
provide them with enhanced productivity and empowerment.
Digital’s
new joint venture in Hungary provides a front row seat
at the drama of these major global changes. The
Hungarian venture gives a glimpse of a great variety of
governÂment policies that impact Digital’s business.
A number
of public policies — both of the United States and of
Hungary — were helpful to Digital. Both countries are
seeking to promote the institution of free-market
systems and the privatization of industry. The Hungarian
organizations with which Digital formed its joint
venture had both previously been government-owned
institutions. The public policy of privatization of
enterprise in Hungary drove these two organizations to
seek alliances with companies in the West. Intellectual
property laws and policies in Hungary were also uniquely
suited to help Digital form the venture.
Meanwhile,
the sudden opening of Central Europe spurred the U.S.
and Western European governments to adopt policies
encouraging the alignment of Hungary with the West in
order to neutralize the influence of the Soviet Union
over them. But these policies conflicted with policies
of denial of the availability of strategic technologies
to the Soviet Union. Policies of technology denial
remain. Although controls have been loosened, compliance
with export controls must continue to be an important
responsibility of all Digital emÂployees. On the other
hand, in order to assure that Digital could be as
competitive as other companies in these new markets, the
U.S. endorsed Digital’s request that the export control
regulations be redrawn to eliminate specific
discriminations in the regulations against the VAX-VMS
technologies.
Similarly,
U.S. export rules and regulations were far more strict
than other Western countries in regulating connection
with and service of the installed base in Communist
countries. The large installed base of Digital-type
computers in Hungary - including clones and diverted
equipment - makes Hungary a particularly attractive
market to DigiÂtal. Here again, Digital has been
forthright in asserting its own self-interest in
serÂving the installed base. This interest parallels the
public policies of the United States which seek to
assure the competitiveness of U.S. companies in foreign
markets.
To
adequately assess the public policy and self-interest of
a given country, it is imÂportant to note the variety of
roles that a government may play. A government may be a
regulator of business, a purchaser from a business, a
partner, or a competitor to a busiÂness.
Governments
are, first and foremost, regulators. They tax and
control. However, it must be remembered that these taxes
and regulations not only affect Digital, but affect
DigiÂtal’s customers in that country.
At the
same time, the government establishes and controls and
infrastructure that underÂlies many of the markets that
are important to Digital. With respect to
telecommunicaÂtions, transportation and health care, for
instance, the government is a customer of information
technology companies. The government can be a key
customer, serving as a "lighthouse account". In many
countries, local customers may look at the product
selÂections their governments make and elect to follow
them. Particularly, organizations that are required to
remain consistent with the government’s infrastructure
tend to follow the government’s technological lead. And,
of course, the best argument as to why a given
government should adopt a particular technology is the
success of that technology in the hands of another
government.
Government
invariably, though to a greater or lesser extent, is a
partner of business. This partnership may be based on
aggressive policies and a close alliance between
business and government, such as that seen in Japan.
Other governments take less activist posiÂtions. But
there are several forms of closer partnership between
global companies and individual governments.
Partnership
with a national government to provide a service to the
people of a country can be sensitive and potentially
uncomfortable for the company when the people regard the
success of the project as reflecting on the capability
of the incumbent government. The exercise of political
judgments based on company performance is a vigorous
marketplace test. Digital should not shrink from such
challenges, since it has capabilities to satÂisfy both
governments and citizens with respect to information
technology. The parallel between public policy and
Digital’s self-interest in these situations is identical
— to provide the best for the citizen/consumer.
In a
number of countries, Digital will be required to comply
with public policies dealing with "local content" and
"local participation". These policies are designed to
encourage greater company presence within the boundaries
of the country, and usually center on manufacturing.
Digital works with these public policies in an
enlightened way, recognizÂing that Digital assets for
these purposes are necessarily limited. Digital has
suggested that, where appropriate, manufacturing plants
might become obsolete, whereas engineering or software
facilities would make a more lasting impact on the
infrastructure of the country. As a practical matter,
Digital is in a superb position to help countries build
their infrastructures in more lasting ways than
manufacturing, since in the information technology
industry manufacturing jobs are becoming fewer and
fewer. In bringing our self-interest into parallel with
that of the countries seeking local presence, Digital is
in a position to better serve the citizens of the
country.
While many
public policies compete with Digital’s self-interest,
there are many more areas, particularly in the emerging
countries, where Digital’s contribution can be
proÂfound. Digital values, particularly in countries
that are trying to adapt themselves to a free-market
system, can provide excellent examples of the efficacy
of openness, honesty, fair dealing and respect and
empowerment for the individual. Digital will have a
front row in helping these countries appreciate and
capitalize on the dynamics of the free-marÂket system,
as well as assisting with the structuring of
relationships between government and business in
Western-style democracies.
The
policies and roles of government are many. The
relationships between global busiÂnesses and governments
that will flow from these different policies and roles
will someÂtimes be confusing. It will always be helpful
to recall that governments and companies both seek to
serve the same public — the citizen/consumer.
We can
expect that Digital will have many and more varied
relationships with governments throughout the 1990s and
into the next century. Digital must also be prepared to
deal with the complexities of "super governments", such
as the EEC under which the European countries are
unifying, or regional trade organizations that will
emerge in Central and South America and the Pacific Rim.
The public
sector is a dynamic and growing market, and
understanding of public policies are a key to success
there. Digital must continue to build on its unique
strengths and be prepared to find ways to take advantage
of new opportunities that will open in the global
marketplace. Digital must capitalize on public policies
of governments to help meet the needs of customers.
The
mission of the newly-formed General Systems Business
Group is to maximize Digital’s share of the small and
mid-size business market. For our purposes, ’small and
mid-size businesses’ is defined as companies worldwide
with between 50 and 1,000 employees. Last year these
companies spent $30 billion globally on computer systems
and services.
According
to industry consultants, this is the fastest growing
market in the computer industry because many of these
companies haven’t adequately computerized.
We found
several characteristics that applied to these customers.
The most critical is that applications drive the
computer purchases. Also, since in many cases the money
these customers are spending is directly out of their
own pocket, they are very concerned about
competitively-priced computer systems.
Service
and support are extremely critical. Many of these
companies don’t have the inÂhouse expertise to do the
kind of maintenance or training that they need to
maximize their use of their computers. To do this, they
need quality service, support and staff training.
These
customers primarily buy their technology from what we
call "value-added resellers" or VARs. The VARs have the
expertise in the customer’s industry and particular
application need. Thus the VARs have the customer’s
trust.
To deliver
solutions to small and mid-sized companies, we will be
working primarily with VARs. In order to work with the
VARs, a focused and streamlined sales organization in
each geography will be created. As part of the geography
plan, the sales organization will work with the VARs,
recruiting and managing the relationship. We will
selectively recruit the VARs who will provide the
highest quality complete solution for our customers.
Essentially,
we see Digital’s strengths in three major areas today.
First, we have a broad and mature portfolio of VAX/VMS
applications for small business. We must continue to
work ‘with our VARs to market these solutions
aggressively to our customers.
Our second
clear strength is our service. Worldwide, our Customer
Services, worldwide, are known, and respected, and are a
major asset to our company. This will become one of the
key differentiators for us in this market.
The third
strength is the respect that people have for Digital as
a company. We found that many VARs are enthusiastic
about working with us, and being part of the quality
associated with Digital products and service.
We have
developed a strategy based on our strengths and which
targets areas where we can grow. To win we need to:
o Offer a broad line of
price/competitive products for these customers
o Make available the right set of
applications, and
o Provide comprehensive sales and
services.
Because
applications drive the business, we surveyed independent
software vendors who focus on this market, and learned
that for this market they are making major investments
in three operating systems: MS-DOS,[*]
OS/2* and Unix.*
Of the
three, Unix is the only true multi-user operating
system. It has rapidly become extremely popular in small
business operations. Many of these Unix-based
applications complement our VAX/VMS-based applications.
We decided
to work with a company called Santa Cruz Operations or
SCO, which has a Unix operating system, Unix V3.2. The
driving factor in choosing this system was the over
3,500 small business applications available today on SCO
Unix.
We already
have a number of Complementary Solutions Organizations
(CSOs) committed to providing Digital solutions with
SCO/Unix to expand our efforts in this area. In
addiÂtion, to expand our base or resellers, we are
inviting a large number of VARs to attend DECWORLD in
Boston. Almost all of these VARs are new to Digital. We
are also continuing to recruit VARs through our field
organization.
In the
sales and service area, training is a key component of
our strategy. Since an understanding of SCO Unix and
Intel-based systems is key, we have developed an
extensive on-going training program.
Our first
Intel 386 based multi-user systems running SCO Unix were
recently announced. VARs, customer, analysts, and the
press were enthusiastic about this first step in
expandÂing our product and services focus in small and
mid-size business. There are many more products,
services, and marketing activities well under way.
It is
important to point out that Digital’s progress in
implementing this plan for small and mid-size business
is based on the hard work of many groups working toward
a common strategy. In particular, Jay Atlas and Bill
Armitage from the US Sales organization; Diane Fraiman
from Europe; DickYen, Taiwanengineering/manufacturing;
Henry Ancona, BOIS; John Rose, PCSG; and Don Zereski,
Customer Services, have all been key contributors to
this effort.
Client/server
computing uses a combination of hardware and software,
part of which resides on the client and part on the
server. We normally think of the desktop device as the
client; and, indeed, that is generally the case. But,
with Digital’s style of computing, any computing system
can end up as a client or a server. A server is a system
dedicated to deliver one or more functions needed by a
client. The client requests the work, and the server
does it.
The
Gartner group believes that client/server computing will
be almost universal by the end of the decade. Today the
market is measured in tens of billions of dollars; but
by the mid-1990s, it will be measured in the hundreds of
billions of dollars.
In
timesharing, we developed systems that could provide
various services to users who connected to those
systems. The design often focused on maximizing the
number of users the system could support. In the
client/server model, the balance of computing shifts to
the desktop. Here the focus is on maximizing the
services provided to the client.
Over time,
users began to realize that not all of the functions
they wanted were immediÂately available from their
timesharing system. They wanted access to data lying
elsewhere in a corporate data base or access to some
computing resource that wasn’t directly attachÂed to
their system. Digital, along with the industry, began to
work on that problem.
In the
1980s, as leaders in peer-to-peer networking, Digital
created a large, complex network of systems that would
allow users to gain access to remotely located data or
other computing resources. In many cases, we also
provided network system management capabiliÂties that
were used in such applications as automatic routing of
messages and mail.
Today we
have advanced the state of the art of networking to the
point where our customers have very complicated
multi-vendor network environments. Each department in an
enterprise might have a separate and different set of
computing resources for its own use. In addiÂtion,
individuals in these departments have a variety of
personal computers (PCs) which may be tied together in
various local area networks (LANs). With client/server
computing we want to take all the functionality that
exists anywhere in that multiple vendor complex network
structure and make it available to any user. And we want
to do that in a way that feels no more complicated than
timesharing.
Our goal
in client/server computing is to overcome the
complexities of being able to interact and communicate
with various architectures and various vendors systems.
We want to make it easy for the user to access the
functions that he or she needs, wherever they may reside
on the network. Terminals will still play a major role
in this computing scene.
For
example, say we have five different kinds of desktop
clients and five database sysÂtems, and we would like
each to be able to query any one of those databases. In
that case, it would take 25 different combinations of
hardware and/or software to provide the interoperability
necessary for that query to take place across any
combination of those desktop clients and those database
systems. If you add on top of that the total range of
desktop client architectures that might exist, and the
range of databases and also the range of computing
resources that might exist somewhere in the enterprise
(such as superÂcomputers or special kinds of servers or
backroom office applications), the number of
combinations required to provide interoperability is
very large. As you can see, every enterprise in the
world is going to be different, and we are still far
from being able to provide the level of integration we
need to achieve full enterprise integration.
Digital
deals with that complexity within the context of our
Network Application Support (NAS) activities. Our NAS
program consists of an architecture, a strategy and a
list of services that can be performed to solve each and
every combination that Digital wishes to solve as it
approaches the integration of an enterprise.
Today, we
see three major classes of server. The majority of
servers in use now are PC LAN servers. These are
generally limited to simple file or print services
across a small number of homogeneous architecture
desktops. These have been one of the driving forces
behind client/server computing technology, but I believe
they will not be the dominant factor later in this
decade relative to the overall client/server computing
picture.
Workgroup
servers support
multiple workstations or personal computers with a more
sophisÂticated set of services than PC LANs. These
include database servers, compute servers and other
capabilities.
Another
new and important class we refer to as "integration
servers." In this fast-moving area, we are just
beginning to see some of the server functionality that
we expect will become commonplace later in the decade.
These servers work with other departments and other
server systems to bring the client the capabilities it
needs.
In the PC
LAN and workgroup space, Digital can probably support
more multiple vendor clients than anyone else in the
industry. We can support five major desktop
architectures as clients, ranging from Apple Macintosh*
to MS-DOS and OS/2, as well as VMS and Unix
environments. We support these across a family of VMS
servers that range from our low-end MicroVAX systems to
our high-end products. We also support them with our
RISC/Unix family of systems
We are in
a unique position in terms of the breadth of client we
support in the industry at the PC LAN or workgroup
server level. But our users tell us they need more than
this.
They need
access to bigger databases and larger installations.
Some want access to their company’s supercomputer, or
need to get into the backroom application area and
interoperÂate with an IBM system there.
Digital
has developed all of these capabilities over the last
decade. We have proven strengths in networking,
multi-vendor gateways and other capabilities required to
bring functionality directly to the client from a remote
system in the enterprise. Today’s users are just
beginning to get comfortable with their PC LAN
environment, and are beÂginning to see these additional
possibilities. They are looking to Digital for help.
A recent PC
Week survey showed that the VAX 6000 system was
number one in "ability to support distributed
applications on PCs, ability to function as a server in
a PC network and multi-vendor networking and application
portability."
We are
fast approaching the point where the majority of our VAX
3000 and 6000 systems are being used as servers. In
fact, we’ll soon see some of our VAX 9000 systems being
used as servers.
For
example, consider how our base system platforms are
being used as servers at Chemical Bank, the customer for
which I am the executive partner. MicroVAX systems are
supporting the PCs and teller terminals at each of the
branch offices. Each of these MicroVAXs systems feeds a
VAX 6000 application server that does the branch
consolidation work, and each of these servers can, in
turn, act as a client to several of IBM databases.
If you
look at the Chemical Bank world, the IBM databases,
which were built up over a long period of time, are
broken down by the functions that they support, such as
demand-deposit accounting or loans. If you, as an
individual, go into one of the branches and want
information about each element of your business with
that bank, it is up to Digital’s systems to integrate
the IBM-based data and serve the information back to the
appropriate branch.
In other
words, today we have many instances where our base
system platforms can work as either client or server
depending on the needs of the moment.
You can
expect to see our server development activity going
through a series of functional increases over time.
You’ll see database servers, compute servers,
application-oriented server systems and integration
servers.
The first
task, of course, is to identify any particular area
where it might be useful for Digital to provide server
functionality. We might then announce a product that can
proÂvide those basic services. Then we continue to make
improvements to simplify their use by the client and the
end user.
In
summary, the client/server style of computing has
evolved from the timesharing style at which Digital
excelled. With this sytle, the system is the network,
and Digital’s straÂtegy in this area leads to full
enterprise integration.
* Macintosh is a trademark of Apple
Computer, Inc.
Many of
our large accounts today have "islands of information" —
separate computers systems that have evolved in separate
divisions over the last 10-20 years. It is very
difficult to integrate information across these islands,
with their different hardware, operating systems,
database systems and proprietary applications. In cases
where the database system happens to be the same, the
data is represented differently and there are different
kinds of data.
Three
issues are important to the customer in major
integration projects like this: cost, quality and
timeliness. Since the customer does not have the
technology today to do this integration, we have to use
people power to do it for them. That increases the cost
and also increases the chance of errors, which is a
quality problem. Timeliness is also difficult.
The
solution we offer is a "distributed database system."
That seems to be a contradicÂtion in terms — using a
distributed approach to do integration. But the role of
a disÂtributed database system is to provide a uniform
interface on top of the existing dataÂbases and, hence,
to serve as the integrator of information in the
enterprise.
That
uniform interface is used by applications and users in
accessing data. The interface must be tailorable so it
can be configured for the individual needs of
applications and users.
In
addition, in the 1990s, we expect to see
"individualizable" software on workstations and personal
computers. This interface has to accommodate that trend
as well.
The
distributed database system separates the data from the
users and the applications. This means you could roll
out one of the local databases and roll in a new one
without disturbing the applications and the users.
To
illustrate our distributed database solution, I’ve
chosen a specific project at McDonÂnell Douglas, which
involves three kinds of integration. The first example
involves commercial data, consisting of numbers and
words, for which we use relational technology. The
second step involves integrating documents, image, voice
and audio with data. The technology required is an
extended capability on top of relational which allows us
to deal with multi-media data. The third scenario
includes graphical data, referred to as "obÂjects," and
provides object-oriented capabilities as part of the
overall distributed data base system.
McDonnell
Douglas must report cost-of-data information to the
government about this parÂticular project. But the
project is supported by three different divisions in
three different locations, and each division has its own
information systems that evolved over time. The division
in California has a VAX-based system using a Supra*
database from Cincom. The division in Missouri uses an
IBM system with DB2.* And the division in ArizÂona uses
a hierarchical system, IMS, an older type of database
system.
The
hierarchical system is totally different in orientation
from the other two, which are relational systems. But
the relational systems are very different from one
another, too, because of individual choices made by the
database administrators, who restructured the data
differently. The DB2 database system presents the cost
information in two columns: previous year and this year.
And the California system has it in one column.
Today, a
lot of people are involved off-line, doing
"scratch-work" editing to integrate that data and come
up with the necessary reports. Obviously they need a
system to allow them to enter a single query from a
workstation — "show cost data by task for this proÂject"
-- and get an integrated answer back at that
workstation, without worrying about all the complexities
that underlie those queries.
In the
future, when the query comes in from a workstation, the
distributed database system will break it down into a
form that each of the three local databases will
understand. The data dictionary helps in this task. The
data dictionary contains information on the structure of
the local databases and lets the distributed database
system know how to reformat the request.
In
addition, the distributed database system needs to deal
with local dialects of lanÂguages that are used to talk
to those database systems. When the distributed database
system is successful in sending the queries, the data
comes back in three different forÂmats from the three
different database systems. The distributed database
system restrucÂtures that data so it can be merged and
then combines it. Next, it looks in the data dictionary
to see how the client expects the result to be
presented. Then, it sends the result to the client.
From the
user’s perspective, sitting at a workstation, the screen
shows a menu; and with a click of the mouse, the user
selects "query." A table of projects appears, and the
user selects the right project. Then, the result comes
back on one table on one window.
We are
currently working on projects to extend our ability to
both read and write to the foreign databases with which
we interoperate.
In step
two, we add the complexity of dealing with images,
pictures, voice, audio and text. As part of the major
Department of Defense project known as "CALS," McDonnell
Douglas and other large government contractors are
required to report detailed logistics information. We
are now prototyping for them an Rdb relational database
that is intended to include that logistics database.
Today,
McDonnell Douglas uses two other databases. An Oracle
system on IBM equipment contains part and assembly
information. Since that Oracle database system cannot
store images, the images are stored in a flat file
system, also on IBM equipment, with pointers from the
Oracle database to the flat file system. An engineer
working on this project has to access separate databases
and put the information together if he or she wants to
find out which parts lack maintenance data.
The
solution to this problem once again involves working
with the individual structures of separate databases and
bringing the information back. But in addition, we need
to provide and show the requested drawings. In this
case, the distributed database system is responÂsible
for making sure the image gets delivered from the
server, from the database system to the client. Dealing
with multiple servers of various types and various
hardware platÂforms and multiple clients, we have to be
very flexible in terms of being able to map from
different stored image formats to what the client
expects to receive. Once again the data dictionary is
the key, providing information about the individual
formats.
The user
view should once again be simple. As before, the user
chooses a query, and the results come back in a window —
in this case, maintenance data. Perhaps the last item
has no maintenance data. Then a drawing of the part is
accessed through an icon on this window. With a couple
click of the mouse, the drawing is displayed on the
workstation. Now, we need certain capabilities to
interact with that drawing. Therefore, the menu bar on
top now includes a new choice called "image." This
choice provides the capability to highlight part of the
picture, to annotate it with a text, etc. The
distributed database system is responsible for making
sure the annotation is connected to the pictures.
By the
way, today the vast majority of the data at large
defense contractors like this one is in the form of
paper. CALS requires that they move to electronic
storage, interconÂnected over a computer network. There
is a tremendous opportunity for us with our multi- media
offerings to be the storer of this vast quantity of
information.
The third
scenario involves "object-oriented" data. McDonnell
Douglas, like many other large corporations, today uses
multiple computer-aided design (CAD) applications.
Various departments have different proprietary CAD
applications, many of which have embedded file systems
to store the objects on which the engineer is working.
Say. for
instance, two parts were designed on different CAD
systems, and you need to make modifications on both and
relate them to one another. You could move the one
object from one CAD system to another, but then you have
two copies of the same thing, and if you update in one
place, you need to remember to update in the other place
as well. That becomes a cost, timeliness and quality
issue.
The
solution is to include an "object manager," or an object
database system as part of the distributed database
system. This object manager needs to have a standard
interface and also needs to store information in a
"neutral format" - a format that would be agreeÂable to
the various application vendors for their CAD
applications. Vendors are working on standardizing such
representations. We expect that in the next 12-18 a
standard will be agreed upon that can be used to
implement such an object manager.
As in the
previous example, the distributed database system now
provides the interface between the application and the
object data manager. When it’s time to deliver an object
to the application, using the data dictionary, the
distributed database system maps that object into a
format that the application will understand and can use.
In this
case, the engineer using the system from a workstation
would find a part, click the mouse on the icon that
represents the object, and the part would appear in the
CAD application, available for manipulation and
modification. After modification is compleÂted, the part
design would be updated in one place.
We are working on a number of
development projects to simplify this kind of
integration.
We call
these kinds of solutions the "Information Network." No
longer are database sysÂtems passive "bins" in which you
store bits. Rather, they are active participants in
making the network operate at a higher level, in terms
of understanding the information and being able to
translate the various formats between servers and
servers and between servers and clients.
Our
"Information Network" is inherently distributed. And it
has capabilities that the competition won’t be able to
match for a long time to come.
*Supra is
a trademark of Cincom Corp. DB2 is a trademark of
International Business MaÂchines Corp.
Transaction
processing is a $37 to 40 billion market this year, and
it is growing at about 13-1/2% per year — more than half
again the rate of the industry at large. By 1995, we see
a $70 to 75 billion opportunity in this area. Our
participation in this business arena has seen
significant growth over the past year.
Our
history in transaction processing started in July 1988
when we announced our dedicaÂtion to this market with
our distributed transaction processing architecture, and
related DECtp products. Since then we’ve made a great
deal of progress from both the Product and the Field
standpoints.
From a
hardware standpoint, the addition of the high end VAX
9000 and fault-tolerant VAXft 3000 systems has given us
access to virtually all of the transaction processing
marketÂplace. With the additions to our hierarchy of
storage devices, we now have a full range of disk and
tape products to give us the capacity, performance and
price/performance that we need to satisfy our customer’s
requirements.
We have
also continued to enhance the transaction processing
performance of our VAX 6000 and Micro VAX systems. In
May 1988 the VAX 6200 family offered a maximum of 18.2
transÂaction per second. By December 1989, we had
reached 50 transactions per second, in the same box.
That’s tremendous growth that gives us excellent
penetration in the marketÂplace. We have similar growth
at the low end, where the next MicroVAX system will
offer more than 20 transactions per second. So we’re
moving very rapidly in the performance domain, which has
helped us make radical improvements in price/performance
as well.
Probably
even more significant for us has been the growth in our
ability to support our software and hardware in the
field. Transaction processing is a relatively new arena
for us, and we’ve had to put great investment into this
area to give us the ability to reach
our
customers, help them understand our offerings, help them
design Digital solutions for their problems, and to
provide for their service and support needs. We’ve
invested in education with Digital University Summer
Session last summer, the Digital Institute of Technology
in the fall and Winter School in Europe in February. Our
DECtp University gives an intensive 2-1/2 to 3 year
curriculum to provide solid Transaction Processing
expertise in the Field. We have established a design
consultant activity in engineering, and a network of
expertise centers including the new US Centers in
Chicago, Atlanta, New York and Los Angeles. To
complement our internal EIS efforts we have also
established important Service Alliance relationships
with Anderson Consulting, Computer Science CorpÂoration,
the Arthur D. Little Group, Deloit-Haskins and
Price-Waterhouse, and informal project relations with
many of the major Systems Integration providers
worldwide. These steps have greatly improved our ability
to reach our customers with the skills they need, where
they need them.
At the
same time we’ve added greatly to our software
capabilities, enhancing the breadth, function and
performance of our offerings, improving the design
capability, the ability to handle forms, the interface
to the customer and system management related issues.
We
recently introduced a key software enhancement to VMS
known as "DEC Distributed TransÂaction Manager"
(DECdtm). This VMS service manages the important
"Two-Phase Commit" protocol for our
data-integrity-conscious customers. Prior to having this
service, there were two ways of dealing with the
complicated problem of databases that are physically or
logically distributed and/or databases that use
different methods to manage data. For instance, the
customer might want to update in a single action Rdb,
DBMS and RMS databases at three geographically separated
sites. The first method, the "cross-your-fingers"
approach, was to do three separate transactions in
sequence and hope that nothing happened in between to
disrupt what you wanted to occur as a single closed
transaction. The secÂond, much more commonly used
approach involved a great deal of investment in
application code to prepare for this kind of update.
This code would check after each to make sure it was
done successfully and to prepare for the next piece of
the update. That was the state of the art in this arena.
With our
DECdtm offering (and complementary database and TP
monitor products), the soÂlution has become very simple.
If you want to update multiple databases, you can do so
with a single transaction. The DECdtm software manages
the complexity of the database updates for you while
ensuring consistency. This is a significant enhancement
to our functional capability and an important selling
feature for our customers. Many compliÂcated tasks that
couldn’t be done or required time consuming, complex
solutions are now easily resolved by this capability.
Our DECdtm
offering also enhances our ability to handle
cluster-based transaction proÂcessing with near linear
scaling. This feature coupled with the availability of
the VAX 9000 means we can now handle even very large
centralized computing requirements, in adÂdition to our
heritage of leadership in distributed computing.
We use the
client/server model as a basis for our distributable
transaction processing architecture (DECdta). We can
build a transaction processing system on a single
computer with the DECtp software logically separating
the client and server functionality or on a wide variety
of frontend/backend configurations using the full range
of our VAX product offerings. A person uses the client
software to handle his or her forms management, and menu
selection, while the back end handles the mainline
application, and data management. In this model the
monitor software coordinates (ACMS or DECintact) the
actions that occur between the database and the
workstations or terminals, and provides support under
the application. More generally, it works in conjunction
with the VMS operating system and our DECnet support to
provide the infrastructure for the overall transaction
system.
Using this
architecture, we and our customers can design
applications on a VAXstation system, do prototyping on a
MicroVAX 3000 system, do field test on a VAX 6000 system
(using a MicroVAX 3000 as a front end if that’s
desirable), and then, if it’s a large enough
application, deploy it on a VAX 9000 or 6000 cluster.
With all of this we’d still have plenty of room or grow
into a cluster of VAX 9000 systems or even a distributed
cluster, which could provide capability for over 1000
transactions per second. Using our DECtp architecture,
the application doesn’t change from the point of
inception on the VAXstation system.
From a
performance standpoint, this means that Digital can now
offer systems for even the world’s largest transaction
processing applications.
Finally,
we offer leadership in the area of development
productivity and the integration of the enterprise, or
the ability to integrate a variety of different
applications on a single system; or through our Network
Applications Support (NAS), even on multi-vendor
distributed systems using a variety of Desktop devices.
Customers buy our systems for transaction processing
based on the fact that they can develop applications
more rapidly with us than on competitors’ systems. They
use our systems because of the flexibility we offer not
only in transaction processing, but in the ability to
handle the full range of the customer’s application
needs.
Our
challenge for the coming fiscal year is to continue to
expand and enhance all elements of our transaction
processing program, from our hardware and software
products, to educaÂtion including the important DECworld
and DECville forums, to marketing and applications
procurement, to our Sales and Sales Support activities,
to our Enterprise Integration Services, to our Customer
Service. Together we can continue to rapidly expand our
base of satisfied Transaction Processing customers.
On April
3, we announced the DECstation 5000, Model 200, system.
To understand the impact of this product, we need to
look at the evolution of workstation specifications.
In the
1980s, central processor speed was the workstations’
primary measurement. Vendors quoted MIPS (millions of
instructions per second) as if they were comparing hot
rods. But for the 1990s, workstations must now have more
than just processor MIPS. Vendors must offer
well-balanced systems, with performance in integer as
well as floating point, and graphics performance in 2-D
vectors, 3-D vectors, and 3-D polygons. The balanced
system should also offer main memory expansion without
arbitrary limits.
We have to
redefine the desktop power paradigm so it is meaningful
for users in the 1990s. A system that excels only in
floating-point performance and is deficient in the other
areas is unbalanced. It’s like those early hot rods.
They go fast in a straight line, but otherwise they are
very limited. IBM has announced RISC systems that are
like this — they do one thing fast.
The
DECstation 5000 system is an open workstation for the
1990s. At 24 MIPS, it’s one of the fastest workstations
in the world. We’ve designed it to be balanced, flexible
and expandable. We offer it in four versions: from the
simple low-cost system, up to a spectacular graphics
machine that provides the best desktop, 3-D solids
performance in the industry.
The
DECstation 5000 system provides ten times more 3-D
polygon performance than IBM’s comparable system. It
offers four times more 3-D vector performance than IBM,
20% better integer performance, and four times greater
memory. Compared to the SUN SPARC 1 or SPARC 330, the
DECstation 5000 system is faster in every area. Next to
Hewlett-Packard’s comÂparable model, it is anywhere from
two to eight times faster over a variety of graphics
measurements. The central processor is twice as fast as
Hewlett-Packard’s fastest workÂstation, at a lower
price.
These
numbers come from systems running a real application
through the X-window’s interÂface. We measure the
graphics performance of all workstations through open
industry standard interfaces.
The
DECstation 5000 system is more expandable than any
desktop system from the competiÂtion. This system offers
up to a 120 megabytes of main memory. That’s more memory
than users can get in a desktop system from SUN,
Hewlett-Packard, or IBM. In fact, to get a 120 megabytes
of memory in an IBM RISC system, you’d have to buy a big
desk-side model, the 520, which starts at over $27,000,
twice the price of the DECstation 5000 system.
Users can
scale the graphics performance of the entry-level system
up to the top system simply by changing graphics cards.
The box doesn’t get any bigger. If the applications
demand it,
you can turn the basic CX system into a high-end PXG
turbo model, with color 3-D solids. You simply add an
option card. With this board-level technology, the
upgrade path for customers is simple. They decide on the
combination of graphics and I/O they need, and plug in
the cards. It is "user-upgradeable."
We also
support the new VME bus. Twelve vendors have already
indicated that they will offer VME peripherals for the
DECstation 5000 system. In addition, we introduced the
Turbo Channel, a scaleable, extremely fast bus that we
opened to the industry at no charge. That’s a real
commitment to open computing.
The
DECstation 5000 system offers performance, superior
graphics, expandability, flexiÂbility, storage, and open
buses. But, most important, in this era of "open
computing," it’s a system built on hardware and software
standards. Digital has been driving and supporting open
systems since the first version of Unix was created on a
PDP-7 system, nearly 24 years ago. The DECstation 5000,
Model 200, system incorporates more hardware and
software standards than any other workstation today.
"Open computing" and today’s
workstation market
"Open" is
a relatively recent term in the computer marketplace.
Customers who demand "open systems" are simply
challenging us to deliver value to an expanding set of
standÂards.
Today’s
requirements list for open computing often includes
central processor technology that is available as a
commodity or is licensable from the manufacturer. It
also includes Unix as the operating system. Sometimes
this requirement can be satisfied by a POSIX-com- pliant
proprietary system. The rest of the list includes
standard languages, communicaÂtions, graphics, and
hardware buses.
Keeping in
mind these requirements, let’s take a brief look at the
history of workstaÂtions.
Apollo
started in 1980, with the Motorola 68,000 microprocessor
and Apollo’s proprietary operating and networking
systems. Apollo was alone in the business for almost
three years. They offered an alternative to traditional
timesharing. Later, Apollo designed its own RISC
architecture for their highest performance model, Apollo
was acquired by Hewlett-Packard in April 1989.
Hewlett-Packard
first offered workstations in 1982. They introduced
their RISC technology in 1986, in the form of a
workstation. Hewlett-Packard now offers both
68,000-based workstations and RISC-based workstations.
When Hewlett-Packard bought Apollo, it signalÂled that
they were very serious about workstations.
IBM made
an attempt to enter the traditional workstation market
in 1984 with the IBM PC-RT. This product was simply not
a success as a workstation, but IBM did manage to sell
it as a multi-user UNIX system. Outside the traditional
single-user workstation space, IBM has for many years
delivered workstation-like features by means of graphics
systems connected to mainframes. In February of 1990,
IBM introduced the RS-6000 AIX Power SerÂies. This
family of products includes four workstations and five
servers, and claims to offer leading integer and
floating-point performance.
Silicon
Graphics, Inc. (SGI) started business in 1983. They set
their sights on the three-dimensional graphics market
and, until late in the 1980s, stayed focused on the high
end of the market. SGI has adopted the MIPS
microprocessor and, in 1989, they introduced the
personal IRIS, which is their least expensive product.
They retain a focus on the high end with their recent
announcement of the Power Vision series.
SUN
Microsystems started in 1982. Their first products,
based on the Motorola 68,000, originated as a Stanford
University graduate student project. "SUN" stands for
Stanford University Network. They used Unix from the
start and stayed with the 68,000 family until November
1987, when they added the first SPARC product. They have
now established signiÂficant applications momentum for
their SPARC architecture. After a two-quarter setback at
the end of 1989, SUN is now reporting strong growth and
has returned to profitability.
Digital’s
workstation history starts a few years later than that
of our competitors. Early products were the VAXstation
I, VAXstation II and VAXstation II-GBX systems. In May
of 1987, we delivered the VAXstation 2000, our first
on-the-desk, complete-in-one-box, VAX workstation. This
product, with an entry price of about $5,000, became the
industry’s biggest selling workstation model, a
distinction it yielded only to the VAXstation 3100
system in 1989.
In January
1989, at DECtop University in Littleton, Mass. , we
announced two new VAXstation families and the beginning
of the RISC/ULTRIX DECstation family. The VAXstation
3100 was the best-selling single model workstation in
1989. The VAXstation 3520 and 3540 models brought true
color, 3-D graphics to the product line for the first
time at Digital. The DECstation 3100 system held
absolute leadership for workstation performance and
price performance for over a year. The DECstation 2100
was the second point on Digital’s RISC/ULTRIX
workstation product line, with a little lower
performance and a lot lower price. In November, 1989, we
upgraded the VAXstation 3100 central processor.
The March,
1990 International Data Corporation (IDC) Workstations
Report includes a chart of worldwide workstation
shipments in 1989. SUN is number one with just over 36%
market- share; Digital is number two with 26%; and
Hewlett-Packard, after their acquisition of Apollo, is
number three at just below 25% market share. These top
three vendors total 87% of the marketplace. The report
shows SGI at 2.5% marketshare and IBM 1.4%.
In other
words, before the announcement of the DECstation 5000
system, Digital found itself solidly among the top three
workstation vendors worldwide, even though we started
late and weren’t critically focused on workstations
until very late in the decade.
Meanwhile,
our VAX/VMS workstations are getting better. We have
aggressive plans for their base-platform and graphics
performance. We also have a complete family of RISC/UNIX
workstations that set the standard for balanced usable
performance.
Basically,
with DECstation 5000 system and other products that are
coming, we’re second to no one in innovation, packaging,
advanced design and expandability. These are the most
advanced workstations in the market today.
If you take all of that, and add
Digital’s worldwide support, you have a winning package.
I have two
simple messages: we can’t wait for the things to settle
down to get serious about the Unix industry, and we need
to invest resources today to get ready for the next ten
years.
The Unix
industry, by its very nature, is a merger of everybody’s
business in the same place at the same time. Conflict is
the nature of this business. We have to learn to sell in
this environment of change — it won’t settle down. And
we need to invest in this area now, even though money is
scarce, because over the next ten years our major growth
opportunity will be in this space.
The
industry is undergoing rapid and major change. The
hardware cycles, which used to be three years, are now
down to nine months. The RISC revolution means computing
power or "MIPS" cost very little. Central processors,
peripherals, and systems software - which were Digital’s
strengths for years — are becoming commodities. We are
in a new industry, with tremendous pressure on cost and
margins.
When
General Motors was shipping big Chevrolets and Pontiacs
with wonderful big fins, Volkswagen entered the market,
quickly followed by Japanese car makers. General Motors
did not react in time. They saw they could make the next
quarter’s numbers by shipping their current cars; so
they didn’t make long-term, basic changes. Consequently
instead of riding with that change and taking advantage
of it, they lost position.
Today, in
our industry, customers’ values and perceptions are
changing. "Open" is good. Customers expect price wars
and require application and user portability.
The word
"open" is often used as a synonym for Unix. Buying
decisions are almost always based on applications.
Customers know what work they want done and what
applications can do it. They are looking for hardware
that will run those applications with good perforÂmance.
Entire
market segments are shifting to Unix. There are market
segments that a year ago used 100% proprietary software,
and now have shifted completely to open systems.
We have to
embrace this fundamental change in the industry or have
it force us into change.
The
competition is fierce. Hewlett-Packard’s and SUN’s
strategies are based 100% on Unix and RISC. IBM has
arrived in a big way. And most of the has-been companies
and many of the future start-ups are now all based on
Unix.
There’s
still a lot of confusion about Unix and open systems.
OSF, which we strongly support, is defining "open
systems" as the name of the game, while AT&T
continues to defend the word "Unix." They try to use the
word "Unix" to define open systems rather than an open
process. This is the crux of the current conflict, and
we must use that when selling to customers. We need to
ask customers if they want Unix from AT&T, which is,
in effect, yet another proprietary system, or if they
want to be part of the revolution that Unix started,
which is the open process.
There is a
common ground. Everybody’s Unix today supports both the
Berkeley and the System V fundamentals. The IEEE POSIX
definitions are supported on everybody’s Unix. But,
regrettably, it will probably take two more years before
we move past these current wars over the definition of
"Unix" and get to the more useful parts of the business.
Digital is
a company in change. The corporate strategy has been
based on leadership proprietary systems. That’s still a
huge business, but it’s not growing at the industry
growth rate anymore. We need to grow other parts of the
company.
We cannot
walk away from situations when Unix open systems are
required. The Unix segment of the industry is growing at
26% a year. We must capture our fair share of the Unix
business.
With our
Network Application Support (NAS) strategy, our
strengths in VMS products should help our Unix strategy,
and our Unix capabilities should help our VMS strategy.
Our
biggest problem is that many customers still don’t
believe we’re committed to Unix. Independent Software
Vendors (ISVs) still put us third or fourth on their
list of systems onto which to port their software. That
hurts us. We’ve got to get the volume in order to move
up on their list. Volume is what they care most about.
Our Unix
operating system, ULTRIX, has a poor quality reputation,
partly because of side issues: we lack key software
products to round out the product, and we have a severe
shortage of trained pre- and post-sales software support
people.
To succeed
in this environment, we’re going to focus software
engineering resources on ULTRIX development. We’re going
to improve the quality and build key differentiators.
We’re going to move past the kernels and build the best
system management, CASE, commerÂcial and networking
capabilities.
To get the
applications, we need to gain "mindshare." We all need
to act committed and talk committed to Unix. We have to
offer and attend the best training in the industry. We
must be viewed as serious and very good at this
business.
At the
same time, we have to make the Unix layered software we
build available on everyÂbody’s platform. Customers will
not buy our spread sheets or our word processing if it
only runs on our platforms. That means we have to sell
our software on the competition’s platforms. Sometimes,
this is going to make our business easier; but other
times it will work against us. It’s a double-edged
sword, but we can’t get away from it.
Unix is in
our accounts today. Retreating from the technical world
into the commercial world won’t work.
In a
survey conducted last November and December, IDC polled
129 commercial accounts. They found:
o 84% were using Unix for commercial
applications;
o 85% said Unix is their operating
system of choice to unite decentralized data resourÂces;
and
o 69% said Unix is their operating
system of choice for mission critical applications.
Four years
ago, we never would have believed you would get that
kind of endorsement for Unix in the commercial space.
Unix is
everywhere. Unix is not the way to make next quarter’s
numbers, but it is our growth for the 1990s. We have to
invest time, people and training today to lay the
founÂdation for that growth.
To
increase our Sales "win rate" we must communicate to our
customers by delivering clear, consistent information
and messages through a coordinated set of Education and
Marketing programs. "One Voice" means responding to a
common set of customer requirements with the same
messages — no matter where in the world that customer is
located.
With the
’Digital Has It Now’ campaign, we built market momentum
by differentiating ourÂselves from our competitors and
making our uniqueness an important purchasing criteria.
And we communicated our uniqueness consistently to every
customer, every day.
Our
worldwide communications template revolves around
customers and account plans. We move knowledge and
information from its source to our customers and the
Sales organization. A source can be Engineering, an
Applications group, a third-party software house or a
SerÂvices marketing group. The channel which connects
the sources to our customers must be fast and must
provide timely feedback.
With its
focus on connecting Marketing directly to our accounts,
the new Application Business unit structure will
facilitate implementation of this high-speed channel.
But that’s
not enough. We need a vehicle, an engine to drive
information and transfer knowledge down this important
highway. Digital's Communication and Education Functions
is the engine. The engine is driven by Digital’s
strategies, our investments and the way we communicate
this information in a clear and compelling way. And it’s
the introduction of each product and service which fuels
our engine and delivers on Digital’s strategies.
Senior
managers in Sales, Services, Engineering, Applications
Marketing and CommunicaÂtions, defined several major
marketing campaigns as the way to organize the
activities to communicate and promote our uniqueness and
help the Sales force win increased sales.
On a
foundation of applications solutions and under the
Network Application Support (NAS) umbrella, each
campaign will have an integrated plan for product
introductions, education and training, advertising and
sales promotion and public relations. With the
leadership of Barry Nay in Europe, Bob Schmitt in GIA
and Steve Thomas in the U.S., area and counÂtry-level
plans will be organized for these campaigns.
The campaigns and their corporate
sponsors are:
o Services — Don Zereski and Russ
Gullotti,
o Office — Henry Ancona,
o Production Systems — Bob Glorioso,
o UNIX/RISC - Dorn LaCava,
o Client-Server — Bill Demmer,
o Software Development — Bill
Strecker, and
o Networks — Bill Johnson.
These are
areas of investment where Digital has a unique advantage
or where we need to focus major marketing activities.
Each
corporate sponsor will guide the overall program and
ensure its scope is companyÂwide. For example, Russ and
Don have asked that we have one integrated campaign for
all services so we communicate "One Service" message to
our customers.
For the
account manager to feel the full force of these
marketing campaigns, their account plans should reflect
how they will sell these strategies and products into
their accounts. For example, a manufacturing account
plan might target sales of production systems like MRP
and Order Processing, or target sales for CASE in the
Information Systems Department. The strategies to win in
this manufacturing account will be positioned and
promoted by all communications disciplines — Marketing
Communications, Customer Education, Sales and Support
training, and demonstrations.
Communications
has become as important to Digital as our activities in
Manufacturing and Engineering. It’s not enough for us to
design and build the best products anymore. We must
communicate we have products and services that help our
customers. And we must communiÂcate in a single,
powerful and unified voice.
We believe
these campaigns, pursued aggressively by the entire
company, will bring momenÂtum and excitement to Digital.
And the momentum and excitement that we all will feel
will be reflected by our customers in greater sales.
Educational
Services has been evolving in several directions.
Whereas in the past, we functioned as a discrete
business unit, today we are a key component of the
Enterprise Integration Services (EIS) organization under
Russ Gullotti. This alignment ensures our focus on the
overall success of Digital as a major solutions
integrator. We provide the critical training element
needed to win major projects.
Previously,
our customer training organization was driven purely by
bottom-line profits. Today, as an element of Appications
and Industry Marketing under Peter Smith, we are
implementing an education strategy that will ensure that
customers and employees receive consistent messages from
Digital.
In the
past, we delivered education and training by audience,
for example, to Sales and Sales Support, and relied on
these groups to tell us the courses they needed. Now,
each product, application and service business unit will
assume the responsibility for identiÂfying the training
needs and course content for the education of both
employees and cusÂtomers.
We are
repositioning our educational resources closer to the
business units. For instance, we recently relocated 29
courseware developers to Littleton, Mass., to align them
more closely with the Networks and Telecommunications
organization.
Until
recently, our primary customer focus was on
tuition-based, end-user training; but we also delivered
customer education at no charge in the form of Corporate
Leader Forums, Business Fellowship Programs and
Discovery Seminars. Now, we’re placing much greater
emphasis on all phases of customer education.
Historically,
the company has had many unconnected training
organizations. Over the last six months, we’ve taken
steps to achieve a higher level of collaboration and
overall productivity among these training groups,
through the Digital Training Consortium.
Our goals are:
o to help each business unit to
expand and maintain the technical excellence and
compeÂtence needed to compete in the marketplace;
o to educate our customers so they
understand, appreciate and use our technology and
application solutions;
o to
expand the knowledge base and encourage the personal
development of our employees; and
o to extend our training to over
165,000 customers; and
o to operate as a profitable
business unit.
Our
educational vision is an extension of the company’s
communication model. We will take corporate themes,
messages and introductions and use them to influence our
employees, customers, partners and the academic
community. The result should be increased technical
excellence and increased awareness of Digital and of our
products and solutions. We will implement this approach
through a series of activities and programs, targeted by
audience, to produce a single, integrated, education
plan for the company.
We’re
positioning this extended educational vision for the
company under the title of "Digital University." The
model we are working on consists of three major schools:
one for architectures and technology, one for industries
and applications; and one for skills and management.
Departments will be identified within each school, such
as client/server and networks in the Architecture and
Technology School, and manufacturing and financial
services in the Industry and Applications School. These
departments will develop content
to train
all audiences, such as Sales, Sales Support and
customers. Responsibilities for the ongoing operation of
Digital University will be shared among the University
staff, the business units and the target audiences.
The most
significant example of this new model and partnership
will be at DEC WORLD ’90. There we will show the
integration of Digital’s messages, themes and
introductions and how education and training is used to
bring it all together.
From
UNIX/RISC workstations to fault-tolerant machines and
hundreds of applications, DECWORLD ’90 is a
demonstration of the value Digital brings to a
customer’s business.
We will
give customers the opportunity to learn by
participation. They will choose a discovery center and
flow from seminars to workshops to demonstrations. Our
guests will listen, question, participate and, together
with our sales people, explore solutions to business
needs. We are all going to make discoveries.
The
Manufacturing Discovery Center will feature 35 seminars
and workshops addressing discrete and process
manufacturing, laboratory and engineering applications.
For example, in the workshops and demonstrations on
concurrent engineering and computer-aided design (CAD),
design teams will share data supporting a host of
federal government logistic standards and using DECwrite
software to combine scanned images with CAD graphics. We
want customers to conclude that Digital’s style of
concurrent engineering differentiates us from the
competition and adds value because of improved
development time and higher product quality.
DECBank is
an example of our innovative solution to retail banking
challenges. It is more than just branch automation. It
is a platform that supports the teller, seller and
cusÂtomer service areas. DECBank integrates existing
applications and allows retail banks to do the outbound
marketing they need to do to compete effectively.
A major
focus at DECWORLD ’90 is on information systems. That is
the thread that knits together all of the applications,
resources and services. It’s also part of the evolution
of Digital reaching out to new customers. The
information systems area will feature a command center,
which will manage not only a multi-vendor network, but
also a network of multi-vendor distributed data centers.
As a result of these experiences,
customers will feel the strength of Digital. They’ll know
we understand the multi-vendor environment. They’ll
recognize that distributed systems are the best approach
to their production requirements and that we have
solutions from the desktop to the data-center mainframe.
They’ll see that we’re committed to open systems, and that
the systems integration capabilities to tie it all
together are here today.
[*]MS-DOS is a
trademark of Microsoft Corporation. OS/2 is a
trademark of International Business Machines
Corporation. Unix is a trademark of American Telephone
and Telegraph Company.
seltzer@seltzerbooks.com
privacy
statement