A message about the medium

Following the late 1990s global information technology boom, spending in the IT sector slowed dramatically. As a result of this slowing growth, and of other factors as well, the IT industry is restructuring. Some once-dominant players, such as Sun Microsystems, are now but a shadow of their former selves. Others, such as Microsoft Corp., are still large, but growth has slowed.

Nevertheless, both Sun and Microsoft remain within the ranks of the largest global IT players.

This begs the question, what is a global IT player?

For the readers of MSI—interested in the use of information technology to improve productivity in manufacturing and supply chain—a global IT player is one that generates its revenue selling IT products or services to business enterprises, in one or more of several market segments.

To see a chart listing of the Global IT players, and further information on how the list was derived, turn to page 20.

The primary IT market segments, vis-à-vis business enterprises, include the following:

• Consulting & maintenance services
• Computers
• Storage
• Software
• Networking

For each of these market segments, four topics—which also can be seen as the parameters impacting the industry's restructuring—can be addressed, as follows:

• Market segment scope
• Key segment drivers
• Recent competitive developments
• Issues to watch

Consulting and maintenance services include IT-related business consulting, IT consulting, outsourcing, and IT equipment maintenance.

Key segment drivers:

1. New accounting, financial reporting, warranty, and environmental regulations are forcing standardization of many business processes and systems.
2. In response to improved global physical logistics over the last 25 years, and improved global digital communications over the last 10 years, companies are disaggregating. Large, self-sufficient, centrally located, company-owned facilities are increasingly a handicap. Instead, multicompany collaboration and demand-driven digital supply networks have the competitive advantage.
3. In this context, internally focused IT architectures and applications are a negative. The advantage rather goes to architectures and applications that facilitate multicompany collaboration.
4. New IT technologies—such as on-demand, grid, and utility computing—better support right-sizing IT infrastructure, optimized IT facilities, and staff outsourcing, and are broadening and transforming the IT services market segment.

Recent competitive developments: Changes in the consulting and services business can be seen in the breakup of the large audit/IT consulting firms, excepting Deloitte. Accenture and BearingPoint now are independent businesses. Ernst & Young Consulting is part of Capgemini. And PriceWaterhouseCoopers Consulting is part of IBM. There's also been major growth in the services business of leading computer hardware suppliers, including Dell, Cisco Systems, Hewlett-Packard, IBM, and Unisys.

Another factor impacting growth in services—post-Y2K—has been a somewhat nuanced shift from the purchase of commercial, off-the-shelf business applications to targeting specific opportunities or requirements by means of business process automation, bringing with it the return of custom application development.

Issues to watch: Going forward, for IT consulting services to continue to grow, we'll need answers to the following:

• Will growing security concerns accelerate or retard company disaggregation and outsourcing?
• Can utility computing—i.e., purchasing computing capacity and application functionality on an as-needed, externally supplied basis—continue to grow, or will technical, regulatory, and economic challenges kill it?
• If utility computing does grow, will viable shared multicompany or shared industry-specific solutions emerge? When external forces—such as Wal-Mart or regulatory concerns—hold sway, shared consulting support and utility computing could have political and economic advantages.

Computer infrastructure includes all general-purpose computers—servers, desktops, notebooks, and other portables. It does not include appliances like cell phones, industrial controllers, or external storage systems.

Key segment drivers:

1. Semiconductor technology advances—for microprocessors and memory chips in particular—contribute to a permanent technology revolution. The good news is that semiconductor cost/performance improvement will continue for at least another 20 years. The bad news, for vendors at least, is that these same advances transformed the computer industry from a very high gross-margin business to an essentially commodity business.
2. Data-center cost reduction is a second major driver. Through the 1980s and 1990s, many data centers expanded horizontally—adding ever-more small and medium servers. That brought additional costs through the sheer number of servers to be housed and managed. As processor size and speed increases, server consolidation is a major cost-saving opportunity.
3. For desktops and portable devices, balanced size, cost, performance, and power consumption require processors with smaller physical size, fewer support chips, and better power management.

Recent competitive developments:

1. A major shakeout is under way in non-x86 processors. The new IBM 64-bit mainframe processor technology likely will be around for a long time, but most mainframe 32-bit clones will disappear. The IBM POWER processor family—used in iSeries, pSeries, and Apple Macintosh servers—is a leading candidate to gain market share. The Intel/HP Itanium family, which is essentially the next-generation HP PA-RISC with x86 compatibility added, is the major challenger. HP's future as a server vendor may hang on the adoption of Itanium as PA-RISC is phased out. Sun and its SPARC architecture are having a hard time. The long-term future success of SPARC remains questionable.
2. The x86 architecture is going 64-bit. AMD first introduced 64-bit extensions in its Opteron processors. Intel is introducing similar extensions for its x86 servers, desktops, and portables. This creates a short-term positioning problem for Intel versus the 64-bit Itanium. Longer-term, given the x86 architecture's complexity due to its 30-year history of backward-compatible redesigns, ever-faster x86 chips may be evermore difficult in the future. Intel predicts that multicore Itanium chips will outperform x86 chips at the same price point by 2007. They imply Itanium will move ahead on cost/performance beyond 2007.
3. There is a growing diversity of processors for portable and embedded applications. Windows CE runs on more than 10 of them. Likewise, various versions of embedded Linux support most embedded processor families. Two of the most interesting are the Intel XScale family and the Transmeta x86 compatibles. XScale is being designed into a wide range of telephones, PDAs, and embedded devices. If rumors of a portable version of Microsoft Longhorn are true, there might even be an XScale version of Longhorn—possibly the first full Windows breakaway from x86. Meanwhile, the innovative Transmeta processors provide x86 compatibility, extremely low power dissipation, and Transmeta upgradeable code-morphing software, or CMS (essentially on-chip microcode). Via CMS upgrade, existing Transmeta processor-based products can be upgraded in the field to add new Transmeta-supported x86 processor features.

Issues to watch: As fiercely competitive computer hardware battles continue, the outcome remains uncertain. Specifically:

• Can Intel/HP make Itanium succeed as a PA-RISC or x86 replacement?
• Can the new Sun/Fujitsu development partnership save SPARC? With Sun introducing x86-based servers, and Fujitsu supporting mainframe and x86—as well as SPARC server—neither appears irreversibly committed to SPARC.
• Will Longhorn run on Intel XScale?
• Will small, specialty processor innovators, such as Transmeta, survive?

Storage infrastructure includes all freestanding hardware storage systems and connectivity networks, such as fiber channel, storage area network (SAN), and network-attached storage, or NAS. It includes software when embedded.

Storage systems are dedicated server appliances with a dedicated operating system (OS), back-up media, disk drives, storage system connectivity to multiple servers, and management software. Complementary products include disk and back-up drives, back-up software, and virtualization software. Storage virtualization software allows multiple brands of storage resources to be configured, backed up, and managed as a unified whole.

Key segment drivers: It's all about growing storage demand and, at the same time, the need for more physically compact, cost-effective solutions. And, like processors, storage components—especially disk drives and tape drives—are becoming commodities, which is driving industry consolidation. Vendors are scrambling to meet growing demand while seeking improving margins through, among other things, research & development (R&D) around advanced virtualization and management software.

Recent competitive developments: Companies have been exiting the basic disk-drive business. Quantum stayed with tapes and storage systems, but sold its drives business to Maxtor. IBM is selling its drives business to Hitachi. Seagate has continuing financial difficulties. IBM and others are developing next-generation, nonmechanical storage devices.

At the storage systems level, while EMC and IBM reported good storage systems sales for the quarter ending June 30, HP sustained a significant dip in sales. In response to margin challenges, vendors are supplying storage system software with or without their own hardware. Microsoft has introduced a server appliance version of Microsoft Windows server. Several vendors are using this Microsoft server appliance base to introduce low-end storage systems. Others are offering low-end storage systems based on embedded Linux.

Issues to watch: Answers to the following questions will tell the tale:

• As server vendors consolidate their server and storage systems businesses, they should achieve economies of scale on components, manufacturing, OS, and design. How will the freestanding hardware vendors, such as EMC, do in this environment?
• Will tape, optical, and disks continue to be the primary storage media, or will new nonmechanical media finally be commercialized?
• To achieve market expansion, what companies—new or old—will go after the small- and medium-enterprise market?

Software infrastructure products include system software (OSs, databases, and middleware); development environments; and systems management applications. Software infrastructure does not include desktop and enterprise applications.

Key segment drivers:

1. With the emphasis on business process automation to achieve demand-driven supply networks, most enterprises require a mix of legacy, custom, and commercial off-the-shelf applications. Accordingly, middleware is the current focus of industry innovation and competition.
2. OSs and databases are becoming commodities. As such, return on continued R&D investments may decline. The only exceptions are Microsoft Windows and Apple OS X.x , which bundle traditional OS, graphical user interface, utilities, and soon the database into one integrated product.
3. Model-based development environments and automating IT operations are areas of innovation and competition.

Recent competitive developments: Historically, computer hardware vendors developed and maintained proprietary OSs. With the availability of licensable UNIX and MS-DOS OSs in the early 1980s, computer makers gained a make-or-buy choice for OSs. Maturing of the open-source movement has made the Linux OS an additional viable option. Other open-source options such as Apache, Tomcat, and Mono Web server technologies; MySQL, Computer Associates' Ingres, and IBM's Cloudscape relational databases; and Eclipse as an integrated development environment, are additional alternatives to vendor software development.

IBM, with the acquisition of Rational Software; and Microsoft, with the next version of Visual Studio .NET, target next-generation development environments.

System management tools from BMC Software, Computer Associates, IBM, and others monitor and manage enterprise applications, and eventually will do the same for actual business-process execution.

Issues to watch: As IT spending recovers, software infrastructure may become an increasingly volatile and competitive area.

• The outcome of the SCO-centered legal battles will help determine the legal viability of open source. If the open-source approach prevails, more revenue may move from up-front license costs to consulting and support. If open source is legally crippled or killed, the license-based software model may last a while longer.
• There are too many middleware alternatives, each requiring application and system software support. A middleware shakeout appears to be developing. IBM is gaining market share. BEA's recent senior executive departures and an earnings warning have raised concerns.
• Market acceptance and adoption rate of model-based and declarative application development may greatly affect business process automation and application integration trends the next five to 10 years. Watch for big IBM and Microsoft pushes in this area.

Enterprise networking infrastructure includes wired and wireless equipment such as network interface cards, routers, switches, and VoIP (voice-over-Internet protocol) hardware.

Key segment drivers:

1. Supply chain complexity and enterprise disaggregation are driving continued growth in wide-area and corporate-bandwidth demand. The largest single impediment to even more projects is the lack of low-cost, last-mile broadband connectivity for homes and small to midsize businesses. Some smaller communities still lack affordable T1 service and most are without DSL and cable options. Solving this problem is a prerequisite for broad-based deployment of broadband-based consumer marketing, demand-driven supply networks, and VoIP applications.
2. New applications, including improved network security and higher-resolution video, will consume additional network bandwidth and require greater processing power in servers, routers, and other network equipment.

Recent competitive developments: While Cisco has a clear lead in most enterprise network equipment markets, competition is growing. China-based Huawei Technologies is an example of the emerging Asian competition. A Cisco suit against Huawei—for allegedly cloning Cisco products—was recently settled.

Wireless standards are improving and broadening. New wireless standards—IEEE 802.11n, 802.15 UWB, 802.16, and 802.20—collectively promise more than 10 megabits per second for wireless networking links, with ranges from 30-meter local connections to wide-area network connections.

The combination of satellite, terrestrial wireless, common carrier fiber, and telco- and cable-wired last-mile connectivity may yet achieve ubiquitous broadband connectivity in most developed nations within this decade. If so, expect further disruptive changes as, for example, previously tangible products—like publications, CDs, DVDs, and software—all move to intangible delivery only.

Issues to watch: Deployment of new networking technologies and growth of new applications depends on having broadband, which is constrained by political, tax, and legal issues.

• Will governments err on the side of protecting incumbent common carriers or let the necessary—albeit disruptive—restructuring of telcos occur?
• A good benchmark is VoIP. Independent of who the provider is, VoIP promises integration of data (Internet); voice; and video/television over a converged IP infrastructure. Unconstrained, this could quickly kill conventional telephone service. In the U.S., the states derive considerable tax revenue from local telephone taxes. They currently are prohibited from taxing Internet services. How this contradiction is resolved will influence the rate at which ubiquitous broadband comes about. How quickly will these and similar issues be solved?
• Which telcos, if any, will successfully convert to converged digital services? Best case, the companies will evolve into broadband end-to-end providers of converged, IP-based networks for data, video, and voice. AT&T, MCI, and Sprint are making significant moves in this direction. Failure may leave a confused and unstructured common carrier market. How will required global connectivity be preserved?

Conventional wisdom has been that IT is maturing and the greatest innovations are behind us. It's more likely that continued semiconductor innovation, and the consequent continued growth of computer technologies will support innovation in all walks of life, business enterprises, and manufacturing and supply chain. We ain't seen nuthin' yet!