DCI's Publication Date: April 18, 1997

It's a Good Time To Be an AS/400 User

By Thomas Bittman
Research Director, Gartner Group

1997 is a good time to be an AS/400 user. Not because AS/400 is the best platform for all IT issues, but because the midrange market is in utter turmoil, the AS/400 is a relatively safe haven, and many midrange vendors are investing to "win over" AS/400 users. Hewlett-Packard, Digital, Compaq and Microsoft, in particular, consider products for the AS/400 installed base to be a good investment. Fortunately for current AS/400 users, so does IBM, which is willing to invest heavily to keep and grow its AS/400 customers.

The opportunities for change are increasing, as are the potential risks. Is it riskier to migrate to Unix, migrate to NT, modernize your AS/400, or simply wait out the massive technology changes in the marketplace? Unfortunately, all are risky. In general, the best advice is to build a strategic plan that describes your requirements and the potential technology solutions five years in the future, and make investments today that are most likely to give you a return on that investment. Over the next few years, some AS/400 users are going to migrate/modernize to the right technology either too quickly or for the wrong reasons. Others are going to migrate for the right reasons but to the wrong technology. Successful AS/400 enterprises will understand their requirements and the technology evolution taking place in front of them, and will make effective, stepwise, and sometimes bold changes to ride out the vendor battles and avoid becoming a casualty.

Among the many significant strategic changes that AS/400 users must monitor are seven critical paradigm shifts that will impact most AS/400 users.

NT Ascendancy: The most important strategic change over the next five years will be the ascendancy of NT in the AS/400 enterprise. While NT will take time to mature for mission-critical deployments, the momentum generated by Microsoft and most other IT vendors (including IBM) who stand to profit from its growth will ensure that NT will be a major force in the server market over the next few years. How will the evolution of NT change the business models of today's successful midrange vendors--and who will survive? AS/400 users face the diametrically opposed options of ignoring NT, integrating NT into their enterprise, integrating NT inside their existing AS/400, or even migrating to NT. Opportunities to shift workloads such as networking gateways, decision-support and intranet serving will become more and more attractive. At the same time, IBM already is investing in technologies to create equally attractive alternatives on AS/400. Which investment decisions will prove correct?

IBM Business Changes: IBM is not immune to the changes in the marketplace, which already are causing a significant shift in IBM's business model. A few years ago, IBM's core competencies and profit centers revolved around hardware. In the past few years, however, IBM has grown to become the largest services company in the world and has made significant strategic purchases of software companies such as Lotus, Tivoli and Transarc. By 2002, 70 percent of IBM's revenues will be generated from services, software, PCs and PC servers. While IBM's existing installed bases remain strategic to IBM, the technologies they use may not be strategic, and may not all receive strategic technology investments that users expect. IBM's success depends on delaying the impact of NT and Unix on its existing installed bases, while at the same time growing a significant new business model based on platform-neutral software and services--much of which will center around the NT operating system.

Data Warehousing: Combined with widely available general-purpose tools and the ability to access data from anywhere, the decision-support and data warehousing markets are exploding. Over 4,000 terabytes of commercial data are stored on AS/400s worldwide, making the AS/400 market very attractive to tool vendors and to alternative warehousing platform vendors. IBM's response has been to add new features to the database and rapidly increase system scalability (including very-large memory). However, can the AS/400, which was originally designed to be a very effective integrated online transaction system, become a CPU-intensive data warehouse? In the next few years, literally thousands of AS/400 enterprises will be making the decision to use either DB2/400, Oracle, or SQL Server (among others) as their data warehouse RDBMS.

Application Issues: AS/400 has been successful for many reasons, but none is more important than the large number of available packaged industry applications. However, AS/400's application portfolio is becoming less unique, as traditional AS/400 vendors are investing in Unix-based and NT-based solutions. On the positive side, this creates a long-term migration path should that become necessary. However, several vendors have reduced their investment in AS/400, causing problems with RISC migrations and Year 2000 preparation. At the same time, a few new vendors have arrived on the AS/400 scene, such as SAP with its R/3 product. Which AS/400 vendors will survive the next few years? Which will prove to be "safe"?

System/Pricing Evolution: As commodity-priced systems invade the midrange, and as IBM works to resolve its overlapping server product lines, AS/400 will go through massive makeovers through the next five years. Already, just one year after the final RISC models were released, IBM is preparing to completely refresh its product line. Through 2002, AS/400s will become highly based on commodity and shared IBM parts. AS/400s, RS/6000s and PC Servers will likely come off the same manufacturing lines, looking very similar. IBM will attempt to hold onto its successful high-margin business model, built around the Advanced Systems. At the same time, IBM will deliver nearly identical Advanced Servers aimed at new business with a significantly lower price point. High-end scalability will double within one year and triple within two, making AS/400 consolidations possible. What does this high-end focus mean to today's low-end and midrange customers? Should you upgrade now, or wait for the next model refresh? How will IBM discounting practices change over the next few years?

High-Availability: A tremendous technology gap on AS/400 today is its inability to share disk space in a clustered environment--the curse of the single-level storage architecture. While AS/400s are extremely reliable, planned IPLs can take a tremendous amount of time, and unplanned IPLs can be unacceptably long. As AS/400s grow ever larger, how should AS/400 managers deal with the need for increasing levels of availability? Will IBM continue to rely on third-party solutions from Lakeview Technology and Vision Solutions, or will IBM see a need to take a more active role in AS/400 high-availability?

Network-Centricity: The one major countertrend to the growing dominance of Microsoft is the growth of the server- and client-independent paradigm--known as network computing. While in many ways Sun and Javasoft are the visionaries of network computing, IBM is bringing credibility to the concept by leveraging its installed base and providing incentives to vendors. Network computers will become a popular replacement to today's terminals and will also displace underused PCs. These network computers will be capable of terminal emulation immediately, at low-cost. However, the real value comes through new applications written for access by Web browser clients and through applications written in Java. Microsoft's success has created an alliance between many vendors to eliminate Microsoft's lock on the desktop application environment, and the future server application environment. If the Java paradigm becomes real Java desktop applications, Microsoft's hold on the Office market might loosen. If new server applications are written in Java, any existing server will be able to run those applications, including AS/400. Microsoft's business depends on growth (employee retention depends on growing stock options). If Microsoft is unsuccessful in co-opting the Java language (ActiveX, Java extensions), and if 20 percent or more of new desktops become network computers rather than Windows-based PCs, the Microsoft juggernaut may actually slow down. The question now is: Should you buy in to the Microsoft vision of the world, or the network-computing vision of the world, espoused by IBM and Microsoft's many competitors?

While 1997 is a good time to be an AS/400 user, successful AS/400 enterprises will not stand still but will take advantage of some of the many options being made available to them. Unlike many other companies--those that rely on second-tier Unix systems, those that jumped on the NT bandwagon in 1996, those that remain on less strategic midrange systems--AS/400 enterprises will have an easier road into the 21st century and more viable choices open to them. Perhaps that road will lead to tomorrow's AS/400, or perhaps that road will lead to NT. The key is to spot and avoid the dead-ends, and ensure that each new IT investment is strategic.

Thomas Bittman is conference chairman for DCI's AS/400 Environment: Opportunities, Risks and Strategies Conference. Please see our online brochure for program and registration details.