It is a cloudless, balmy day in Florida. The hotel room's balcony overlooks a
lake that resounds with the calls of egrets and ducks. Before I can take in the
view, I hear a noise behind me and turn to greet Bob Diesùthe AS/400 Division's
new general managerùas he enters the room.
Dies is a man who is hard to ignore, and not merely because he runs a division
whose revenuesùif it were to become a separate companyùwould rival those of
Hewlett Packard or Digital Equipment Corporation. There is a presence about
this 26-year IBM veteran that commands attention and respect. He carries his
six and a half foot frame with a calm equanimity that shows he is comfortable
with himself, yet he also emanates a steely determination to make things
happenùand a confidence that he will.
We shake hands and exchange some opening pleasantries. Dies looks out at the
lake as he takes off his suit coat and folds it over a nearby chair. We sit at
a table in the center of the room, and I turn on my tape recorder.
How does it feel to be a part of the AS/400 Division after almost five years of
It's good to be back, though in some ways I never really left. I spent most of
those five years marketing IBM's entire product line in Europe. Since that
region is the AS/400's largest market, I talked a lot with the AS/400 division.
Still, it's fascinating to be back. Some things are the same as they were when
I leftùthe compatibility between AS/400 systems, the ease of use, and the whole
system design philosophy. But all the technology under the covers has changed
and a lot of exciting things are happening. That's why I'm thrilled to be
working where I am.
When you met with IBM Chairman Louis Gerstner to discuss your new
responsibilities, did he set any goals for you or the division?
He didn't set any specific goals, but Lou's been very vocal about the success
of the AS/400. He feels IBM should be less product-oriented and more solutions-
oriented, and the AS/400 has that characteristic. So it's been a favorite topic
of his at our senior management meetings.
In what way do you mean a favorite topic?
He feels that the AS/400 Division is good at focusing on total solutions, so
that makes it a likely example when we're talking about changes that could
benefit IBM as a whole. We've talked about how the division packages hardware,
software, and services so that people can more easily use them. Lou has also
said that the AS/400 Division has an awfully strong product and that IBM needs
to support and accelerate it through the end of the decade. From a general
management standpoint, he's saying all the things I would like to hear as the
incoming general manager.
What are your major objectives for the AS/400 and the AS/400 Division over the
next several years?
I have two objectives. First, I want to make sure we keep doing the one thing
that has made us so successfulùlistening to our Business Partners and
customers, then turning what we hear into products. My first priority is to
My second priority is to accelerate the AS/400's move to open client/server
computing, but in a way that avoids all the complexity and the higher staffing
requirements of client/server. As the AS/400 transitions to RISC processors and
new software releases, it must remain a world-class system that is easy to
install, use, and support. As a result, everybody will enjoy the added
performance and function with none of the headaches that you get when other
vendors change the fundamentals of their systems.
In other words, you want to keep the system simple.
That's right. But at the same time, we'll bring in a lot of the UNIX
capabilities, such as the Single UNIX Specification [formerly called SPEC 1170]
and POSIX, to open up the AS/400 to a much wider application portfolio. We'll
also ship more application engines like the File Server I/O Processor (FSIOP)
that let customers put servers under the covers of the AS/400. By extending our
architecture in this way, we could take a huge bite out of the complexity of
client/server computing and give customers the performance, security, and data
integrity of the AS/400.
So "keep the system simple" does not mean "hold your ground." It means "Charge
full speed ahead, but keep the troops organized so you don't cause headaches
for your customers as you go." We don't intend to just hold our ground. We are
gaining market share, and we will continue to do that.
Is that a third objectiveùgaining market share?
Absolutely, and we intend to do it at both geographical and functional levels.
Geographically speaking, there's a large part of the world where we are not
active yet. If you think about the emerging marketsùwhere there is often a
scarcity of trained critical resourcesùand then consider the attributes of the
AS/400, there is an obvious fit. I see lots of opportunities in China, Central
Europe, and Latin America.
From a functional standpoint, there are many PC server sites that could become
our customers. As these servers get larger and start handling network and file
serving, many computing tasks that were easy to do turn out to not be as much
fun. These sites are an excellent market opportunity that we are already
focusing on with the Advanced Server models. Later, when we introduce more
application engines, it will become incredibly easy for these sites to install
an AS/400, drop some of their PC servers, and save themselves a few headaches.
So far, we've talked about three objectivesùkeep listening to the AS/400
community, move to open client/server computing while keeping the system easy
to use, and expand the AS/400's market share. As you strive to attain these
goals, what are the major challenges and obstacles you will face?
On the first objective, I don't think there are any major obstacles at the
AS/400 Division. Our whole culture is based on working closely with partners
and customers, and we just have to remember to keep doing that every day.
As for the second objectiveùtaking major steps into new technologies and
opportunitiesùit's not a trivial matter to incorporate emerging technologies
such as UNIX and application servers into the AS/400 without dragging the
complexity in with them. Taking the giant steps that will get us into broader
market segments is not a problem. Doing that in a way which preserves the
simplicity and the outstanding function of the AS/400 is world-class computer
science, and I think Rochester is up to it. It will be a challenge, as nobody
else in recent history has totally changed the underpinnings of a system
without affecting the application code.
When it comes to expanding the AS/400's market share, my big problem right now
is supply constraints. People have ordered more new AS/400s, upgrades, and
Advanced 36s than we anticipated, and we are sold out. I would like to be able
to respond to an order in two or three weeks, but it is taking more than a
month and in my business that is not a good idea. I intend to solve this
problem by the end of the second quarter.
Over the long term, what challenges will you encounter as you work to enter the
emerging markets you talked about earlier?
In emerging markets, the major challenge is to make sure we have Business
Partners and application providers present in those countries with solutions.
It's easy to take an application for hospitals to Russia, but then you have to
create a Cyrillic version of it. That's a tremendous opportunity for the
application providers and the other people who work with us, but we need
trained people in those countries to work with customers and prospects and to
install the systems.
And what challenges will you face in marketing the AS/400 to sites that are
outgrowing their PC servers?
The major challenge is to develop some momentum behind OS/400 V3R1, because it
provides a lot of the client/server function. We will be rolling out more
function as the year goes on, and I am optimistic that we will do a good job of
absorbing multiple servers under the covers of the AS/400. I am also hopeful
that the added function will generate sustainable volume growth for the AS/400.
I must admit that I'm not as hopeful as you are. Those companies that run their
businesses on PC and UNIX-based servers have made heavy investments that they
want to preserve in their databases and application development tools. And
while the AS/400 still leads these systems in areas such as systems management
and ease of use, PC- and UNIX-based servers are definitely catching up. If
that's the case, how can you win converts?
Here's the proposition we're making to the market. Let's say you just want to
do file serving, or network serving, or queries, or transaction processing. If
you want to do just one of these four things, the AS/400 isn't necessarily the
cheapest choice from a hardware and software standpoint, though it may have the
lowest total cost. However, as you start to do more than one of these tasks at
a time, you find yourself adding things to your system that [IBM has] already
built into [the AS/400's] architecture. In other words, you start getting what
you and I would call a midrange system.
I would suggest that if you start with a midrange system and optimize that
system to handle multiple server tasks from the start, you'll probably have
better function and performance than [other vendors who] patch on those
capabilities later. Over time, you will see more demonstrations like the
client/server benchmark, where we showed that the price/performance of our
Advanced Server exceeded that of a UNIX and a PC server when all three systems
ran multiple tasks. That's how we can effectively position the AS/400 as the
server you should get if you run more than just one type of task.
I would like to talk about some other challenges the AS/400 Division faces. As
you know, IBM's efforts to downsize its workforce have involved major
reorganizations in the sales and support groups that work directly with AS/400
customers. Our subscribers are telling us that these changes have made it much
harder for them to buy IBM products and get service and support for them.
Frankly, many of our subscribers are angry. What do you plan to do to improve
sales and support for the AS 400 community over the next year?
We're doing three things. We are dedicating our AS/400 specialists around the
world to the system. In some situations, we have had AS/400 specialists who
also had training in two or three other areas, and they spent time working on
all these areas as the needs of the branch office or the customer dictated.
Now, we are segmenting these people so that they will be dedicated to the
AS/400. That means that they will understand the product better, they will be
better trained, they will be better educated, and they will know who to call.
We have also decided that we must do a better job of communicating with our
Business Partners, because many of our customers' questions go to them and they
must come to us. We are currently setting up channel organizations in each
geographical area to make sure that we are dramatically improving those
Finally, we have installed an Internet server. Anybody with a question or a
problem can put it on the Internet and we'll get back to them. Over time, we
will also consolidate all of our phone numbers in the United States into a
single 800 numberùthat's already the case in every other country. In the
meantime, the Internet will be a great communications vehicle. [See "AS/400
Internet Web Site," Significa, MC, May 1995.]
How long will it take for you to implement these improvements to the sales and
We have already set up the dedicated sales force. We have lists of the AS/400
specialists we are working with in other countries to improve their training
and education. We'll make sure they get regular communications from us that
tell them what's hot, what's coming, what they need, and who they should call
with their questions. That program is rolling out as we speak. It won't make
everybody immediately smart, but it will certainly improve the education level
of our specialists in the field. You should see an improvement in the
effectiveness of our specialists before the middle of this year.
Back in January, we announced and implemented the channel organization strategy
to better support our Business Partners. My guess is that it will take 9-12
months to get it sufficiently working. It is something easy to organize and
hard to execute.
The AS/400 community is beginning to realize they will not be able to run the
RISC versions of OS/400 on non-RISC AS/400s. Customers and Business Partners
will have to live with two different code bases, and many of them are concerned
about this. What level of compatibility between those two code bases will you
maintain over the next several years? How do you plan to bridge any
incompatibilities that do exist?
The functionality we will be delivering in the new [RISC and non-RISC] OS/400
releases will tend to be on the existing Advanced Series as well as on the
RISC-based series. As we add more functions, you will see us effectively adding
to both code bases for the next 12 or 18 months. Over time, I will have to
decide whether I want to continue to have backward compatibility on new
When you make that decision, what criteria will you use to help you decide?
I haven't crossed that bridge yet, so I don't know. I will deal with the issue
as I begin to make suggestions and changes to the 1996 and 1997 release
schedules. But I can assure the Business Partners and customers that the
applications written on our first systems will still run. We haven't gone to
this kind of trouble just to change our strategy and leave everybody at the
last port and not have them on the boat with us as we move forward.
What are the most profound changes that will take place in the AS/400 Division
over the next few years?
That's an interesting question....When I was here at the initial delivery of
the AS/400, most of the software and hardware was developed in Rochester. In
return, we got tremendous integration capability. We now believe we can get
that capability and not have to develop every component in Rochester. For
example, RISC processor performance is now high enough that we can convert from
Rochester-based processors to PowerPC processors. As a result, we can move
resources that would have been working on processors to other things that our
customers are asking us to do. That's a long way of saying that the AS/400
Division will have a greater ability to adopt things that are functionally
unique from other parts of IBM or from the industry.
Such as SPEC 1170, the POSIX components, the PowerPC architecture, and the open
application serversùwhether they run on Intel or PowerPC chips. We will more
quickly adopt new technologies from the industry and make them available on the
AS/400 without building them ourselves. But we want to adopt them in such a way
that we don't create added complexity.
As you manage the AS/400 Division, what will be your least enjoyable and most
My least enjoyable task will probably be making sure that we view our markets
on a worldwide basis. Obviously, we already do that, or Europe wouldn't be our
largest market. But it takes continual conscious effort for a United States
division to think beyond this country. It's not fun to keep asking when they
will have the Chinese or Cyrillic languages available, but that's the kind of
opportunity we have and it's one of our growth markets. While it might be
mundane and boring to keep the worldwide view in front of our development team,
we have to keep doing that.
My most enjoyable task will be renewing old acquaintances. I have already had a
chance to meet a tremendous number of old friends at headquarters and the
laboratory as well as at conferences, and I personally enjoy that. It's a
delight to be part of a division that is doing well.
I want to personally thank Bob Dies and the many other IBM representatives who
made this interview possible.
Lee Kroon is the industry analyst for Midrange Computing.