VP (AS/400 Worldwide System Sales), IBM.
It has been a long journey,
nearly a quarter of a century, for Denise F Buonaiuto to her current job as VP of the
AS/400 Worldwide System Sales division at IBM. A product of the University of Rhode
Island, Buonaiuto joined Big Blue as a marketing trainee in 1974 and has since held
various positions in the organization. On her recent visit to evangelize AS/400 in India,
she managed to spend some time with DATAQUEST to deliberate on the various issues about
the server. Excerpts:
How has IBM positioned the
AS/400 platform in its product space. Is there a clear differentiator with the RS/6000
Broadly, we have segmented our product range into four services. On the top of the ladder
is our large systems division, S/390, catering to large banks and manufacturing
organizations. Again, it is easy and simple to segment the other end of the ladder, the NT
servers, which look after the need of small businesses. Of the other two services, the
AS/400 and RS/6000 do overlap to some extent in the mid-range space. Of course, we
differentiate between these two platforms based on applications. For instance, we don’t do
engineering scientific computing or CAD/CAM-based applications on the AS/400. We leave
that for the RS/6000. Where we have positioned AS/400 very well is the small and medium
businesses, which want to run their business and not their computers. This segmentation is
further facilitated by the second factor. The AS/400 has an integrated architecture and we
have integrated an OS, database, security and communication and then we work with our
solutions providers to provide the applications. So in spite of the overlap, we are very
clear of the market area each platform focuses on.
But, isn’t this space being
invaded by high-end PC-servers?
Yes, it has to some extent. However, the biggest disadvantage with the high-end PC servers
is that they don’t scale too well. Yes, you can definitely do clustering with these
servers, but what our customers are finding is that clusters of PC servers are tougher to
manage. Five or six NT servers together in a cluster, that forms the basis of any
application are not as easy to manage as compared to a single AS/400 server. The AS/400
was designed for nothing else, but doing business. We are finding that customers are
looking for better system management capabilities, easy to use and above all low Total
Cost of Ownership (TCO). And when you put the AS/400 in this model, it is the lowest TCO
system because it requires the least amount of technical expertise to run. TCO would
encompass hardware cost, typically around 20-25% of total cost, and the balance cost is on
the personnel, which is easy to manage. That’s where people are starting to realize the
AS/400 as a benefit.
With limited applications, the
TCO of a AS/400 should be high and not low?
No, actually not. We have tons of applications for the AS/400. But I do agree that people
choose the application before thinking of the technology and this is a gray area for us.
We are making considerable investment in Domino and Java to overcome these limited
problems. Adopting 100% Java will open up a realm of applications for the AS/400, not
necessarily written for the AS/400 but easily used on the AS/400. We can combine Java
applications with this low TCO system to have a differentiated advantage in the market.
Moreover, AS/400 can run both NT and Unix application. We run NT on an Intel-based server,
running under the cover of AS/400. With the help of the integrated PC-server, we can run
NT-based applications and OS/400-based application on the same AS/400. We can share data
back and forth from the NT server to the AS/400 server. One can have 16 NT servers under
the cover of a single AS/400. This is what I imply by low TCO of AS/400. For instance a
single command of `Backup’ will back up my AS/400 and my NT data. Thus the second part of
TCO is of personnel usage is very low for the AS/400 platforms. The AS/400 is the lowest
TCO system available in the market.
Does AS/400 support a
centralized computing type of environment like online, real-time applications, or are
there some limitations?
No limitations at all. In fact, that’s been our strength. Our strength has been to provide
a large system centrally and then have clients anywhere in the world and network them
together. Actually, we are seeing a change in the way people are doing business.
Initially, the spread of mainframes and minis, coupled with high communication cost, saw
the advent of distributed computing. Now we are seeing another swing of the pendulum back
to more centralized computing and consolidation of service. Moreover, I think that TCO is
again driving this consolidation. It’s expensive to manage these little servers in remote
locations. TCO would increase if one considers availability of technical team at such
locations, back up problems and other related problem with remote management. Like many
other platforms, AS/400 offers excellent remote diagnostic capabilities for remote systems
management. It is a plain common sense that it is easier to manage one of anything than
100 of anything. So as people move on with their business, they are also looking at doing
their business in the most cost-effective manner.
Is there any limitation with
respect to ecommerce solutions and AS/400?
Absolutely not. In fact, this is one area where the small and medium businesses have a
level playing field. In the US, Pacific Brokerage, a small telephonic brokerage firm,
moved over to web-based trading within two weeks on the AS/400. This is no mean
achievement as it usually takes around three to six months for an operation of similar
size. The result: they grew their business by 70% within six months such that a bigger
brokerage house looked at their internet-based computing capability as a competitive
advantage and acquired this company. There is no limitation of the AS/400 capability in
Actually, I think that this
platform has many strengths and lends itself to the e-business environment. The AS/400
system is considered as a very secure, reliable and very scalable system. For instance, we
have a notes benchmark, certified by Lotus Corp, that you can put 10,400 Domino users on a
single AS/400 server doing email with sub-second response time. So the traditional
strength of AS/400 of reliability, security, scalability and services are all the same
thing that one needs when you bring your business on the web. Moreover, the moment you put
your business on the web, the business better be up on the web 24 hours, seven days a
Assuming that we already have an
AS/400 installed in our company and plan to deploy an ecommerce application, what would be
the additional hardware requirement?
Nothing. Of course, it would depend on the how you want to implement the ecommerce
applications. May be, you might have to upgrade your OS from version 3.7 to version 4.3.
We continue to improve our e-business and ecommerce capability in each of our subsequent
version. So the more current you are, the better your functionality can be. Then you can
use tools from IBM or other solution providers. For example, the net.commerce initiative
from IBM. Net.commerce is a framework for customers to build capabilities to do business
on the web. However, for document-based capabilities, Domino would be a better option to
How does the AS/400 product
range scale from low-end to high-end?
At the lowest end of the spectrum is the model 150. This system is very small and may
cater to around 10-20 users. But we can grow this system, from 150, all the way up to our
largest 12-way model 650 and this growth is around 330 times in capability terms.
Moreover, from the 600 to 650 models, the growth can be incorporated in the same box.
However from the 150 box one would need to change boxes to scale to the higher 600 series.
But we would look into many issues before advising our customer on the server type to
invest in. We would look at issues like type of data storage, type of transaction per
second/per minute/per hour, datamining, operational transactions and then recommend a
particular type of platform. We would look at the growth of business and if we recommend a
low-end 170, the customer can grow up to 30 times scaling to the high-end 170.
Which are the focus industries
for the AS/400 application platform?
We are focusing on five key industries: distribution, industrial manufacturing, banking
and finance, insurance and telecommunications. These five segments represent nearly 80% of
AS/400 business. Of course it doesn’t imply that we are not into health and other
industries but these five industries are very important in our business plans.
A highlight of the AS/400 system
has been its 64-bit computing capability. Do you see Intel’s 64-bit Merced initiative and
Linux OS combine creating the same problems for the AS/400 as the Wintel combine did in
the PC market?
Well it depends. It remains to be seen what they will actually deliver of the 64. They
would need a 64-bit OS and right now in the Intel space there are no 64-bit OSs. However,
we are already on our fourth-generation 64-bit bit OS processing environment. Again, we
think we have a strategic advantage because the AS/400 was build for 128-bit address. So
architecturally, its not a 64-bit OS but a 128-bit OS, implemented on a 64-bit space. So
we think there will always be a competition but it will take time for the other vendors to
catch up to 64 bit. By this time, we would have outpaced them with our capability of being
able to move to bigger and better computing scales.
With hundreds of people working
on Linux, do you think that it is only a matter of time before a strong competition
emerges in the 64-bit space?
Again it depends. I don’t think Linux runs on AS/400 but runs only on Intel chips. But
here our advantage is based on two things. Firstly, whether the applications that are
already running on current OSs and processors can be ported to Linux and, second, is our
Java initiative. Now if Java delivers on its promise, then it doesn’t matter which OS you
have. It will not matter if your OS is Linux, OS2 or Windows. The whole thing will become
a Java application. So as long as you have Java and as long as Java application can run
anywhere, you have defined openness. The capability for us to grow in future is by having
all these Java-based applications and then the OS is not the issue, but the real issue
would be lower TCO. This is the key strength of the AS/400 and our goal is to continue our
TCO advantage by making it much easier-to-run applications on the AS/400 than any other
With the PC-server market moving
upward with Intel’s Merced and Linux combination and also Unix servers coming down, again
with Linux’s help, do you see AS/400 getting squeezed in between them?
Well, there is always going to be competition and new competition too will emerge now and
then. However, it is our capability to respond to the competition and keep our products
moving in the direction we want. Everybody told us that when Unix came out, the AS/400 was
doomed. Then when NT came out, people told me to look for a new job. That has not
happened, and we are still in business. I am not saying that these are not competitive
threats and I treat each and every one very seriously as a competitive threat to my
Moreover, we still have our
competitive advantage of the 128-bit address capability of AS/400. What this implies is
that if a user has to move to higher operating technologies, he does not have to change a
single line of code. So far, nobody in the business has been able to say this and it
remains to be seen whether Merced will allow this or not. So, if people are thinking of
64bit, we are already ‘old.’ We are already done with 64 bit and are moving to higher
things. So, if your OS does not handle the next generation, somebody will have to do a
whole lot of rewriting in the existing OS. We are pretty confident that we have a platform
that will grow into the future, no matter what kind of technologies comes our way.
With the latest IBM initiatives
on the standardization of Unix, do you foresee the gradual shift of AS/400 and other IBM
products to Unix?
I see IBM commitment to its four server platforms-S/390s, RS/6000s, AS/400s and PCs.
What I do understand about the SCO-IBM Unix initiative is that IBM can provide affordable
Unix in the Intel space. But we are fully committed to the PowerPC architecture at the
high-end RISC 6000 product line and yet we are going to author the alternative of the
Intel-based 64-bit computing at the lower end of the market. This stems from our
understanding that customers want choice to determine what they can do. So we are just
providing alternatives to our customers, but this would not imply that we are not
committed to other platforms like NT or our own OS.
Can you elaborate on the Java
initiatives like the San Francisco initiative and how is it related with the AS/400?
Firstly, what we did with Java, was instead of using the Java virtual machine above the OS
as an application, we embedded in into the OS. So the first initiative is that by creating
a virtual machine as part of the OS we will get better performance in the long run as
customer do multi-credit job and complex Java-based applications. Today, many people are
using client applets and simple Java because the performance at the server level is not
what it needs to be for heavy-duty Java-based transactions. By putting it into the OS, we
think that we will be able to make it a high-performing Java server. Secondly, we are
working with all the Java providers. We are stepping out of the traditional AS/400 space
and working with tool and middleware providers that were not necessarily known to the
AS/400 in the past, but will now have AS/400 capability. We hope this will ensure our
competitive edge in the uncertain future.
and ARUN SHANKAR
in New Delhi.