Adapting to the New World Order
Date: Monday , November 17, 2008
In the current downturn in the technology sector, it has become fashionable to doubt the abilities of the Intels and AMDs of the world to survive new winds of change in the industry, including the proliferation of mobile computing, the saturation of key markets, or ever declining margins in PC sales. While these questions are pressing for the future of an industry that hit more than $30 billion in sales last year, it is useful to pause before examining them, and revisit the factors that drove the industry to its present form. The historical backdrop provides a useful context for examining the future of PC semiconductors in the coming years.
The remarkable growth in semiconductors over the last few decades is the result of two principal factors: the increasing share of electronics in national output (GDP), and the increasing penetration of integrated circuits (ICs) in electronics products. Nowhere are these dynamics more dominant, and perhaps more illustrative, than in the rise of the PC industry during this period.
The ascendance of the personal computer in the ‘80s and ‘90s drove rapid growth in the microprocessor industry. This created strong, sustained demand for PCs through operating system upgrades (successive generations of the Windows OS), and through the proliferation and adoption of software applications. The microprocessor market has grown rapidly over the past 20 years, achieving sustained growth of over 30 percent a year for the decade of the 1990s. This astonishing pace was almost double the growth rate in any other IC category during this period (figure 1).
During this period Microsoft Windows established itself as the dominant platform for personal and business computing. The resulting user “lock-in” to the Windows architecture, coupled with natural economies of scale in engineering and manufacturing chips, created a virtual monopoly in PC microprocessors, securing Intel as the dominant player. Recent Trends
In recent years, however, growth in PC chips has begun a gradual decline, due to a number of key factors (figure 2). First, near saturation in the developed market (50+ percent home, 70+ percent business penetration in the US [IDC], for example) has reduced growth in key areas such as Western Europe and the United States. Secondly, growth in these markets has been further softened by the shift in performance bottlenecks from processors to network latencies and bandwidth constraints, resulting in softer demand for PC upgrades.
A third cause of decline has been the lack of new “killer applications” to impel growth. The rise of the Windows operating system, office productivity applications, and the Internet are examples of killer applications that drove strong growth cycles in PCs over the past 20 years. With Internet-related PC spending now partially having penetrated into key developed markets, PC vendors are left looking toward a horizon of more incremental (or less revolutionary) technologies such as Windows XP or videoconferencing to stimulate slower growth in sales.
Finally, the increasing variety of information appliances, such as gaming devices, handheld computers, smart phones and Internet terminals, may begin to erode PC sales as these devices usurp PC functionality. Although it is not clear what the rate of substitution is like for other mobile information appliances, it is estimated that in the US shipments of information appliances are projected to exceed PC shipments this year (figure 3).
Interestingly, while the super growth of the PC industry created a huge winner (Intel) and a secondary winner (AMD), the legacy of the PC IC industry is a battlefield littered with once famous companies (3Dfx, Cirrus Logic, S3, etc.) that have had to refocus on other markets in order to survive. In fact, one could count on one hand the number of profitable IC manufacturers that have survived the PC wars (Nvidia, ATI, etc.). This situation is unlikely to improve as the PC market saturates.
Momentum and PC Sales
Against the backdrop of these recent events it is easy to see why there has been so much discussion lately on the future of semiconductor companies in the PC segment. Could it be possible that under threats of substitution, eroding margins, and saturating markets, the PC industry is spiraling into decline? What options are there for the processor manufacturers to navigate paths to growth under new industry dynamics?
We begin by pointing out that growth rates of PCs versus other computing devices often hide the sheer momentum of sales in the PC industry. Although sales of information appliances are growing rapidly, PC sales still outweigh sales of other appliances by a factor of more than 15 to one in dollar terms (figure 3). This dominance is projected to continue well into the next few years. The PC remains the workhorse of corporate computing in developed nations, and continues to achieve reasonably strong sales growth in emerging markets such as Asia and Latin America, where desktops render the best price-performance value. All of this points to the fact that the market for PCs, while slowing, is not going to disappear any time soon.
Options for Strategic Growth
Despite the continued sales momentum of PCs, microprocessor vendors are under continual pressure from the market to sustain strong growth. As a result, a number of manufacturers have recently announced new strategies for pursuing growth in the coming years. These recent initiatives follow three primary lines.
First, some vendors have attempted to broaden product offerings and build processors into newer consumer devices such as PDAs, gaming consoles and digital cameras. At the beginning of this year, for example, Intel and ADI announced a jointly-developed DSP chip intended for deployment in mobile telephones, digital cameras and handheld devices. Intel has also worked with Microsoft to incorporate its Pentium processors into the Microsoft X-Box gaming console. Other companies active in this area include MIPS (gaming consoles, PDAs); ARM Research (digital television, home networking, PDAs); and Texas Instruments (cell phones). While Intel and potentially AMD are likely to become bigger players in these markets, key challenges to overcome for these initiatives include the lack of architectural lock-in, and the lack of vertical markets expertise.
While Intel and AMD may be able to leverage their scale, manufacturing know-how and process technologies for high volume applications, the x86 architecture is disadvantaged vs. ARM, MIPS and other RISC implementations that were customized with these new applications in mind.
A second growth strategy for the x86 processor players is to continue their assault on the high-end computing markets. The increasing popularity of Windows NT/2000 and Linux in enterprise server environments has created opportunities for Intel and AMD to develop faster processors for the high-margin server segment, effectively enfranchising NT to compete against proprietary UNIX based solutions (figure 5). However, growth in this segment is unlikely to overcome the declining unit growth rates and declining ASPs in the desktop arena.
On the third strategic frontier, Intel specifically has begun making a major foray into the communications arena, purchasing Level One, DSP Communications, VxTel, NETBOOST, Softcom among others. The market for communications chips has grown rapidly in the past few years and, despite recent debacles in the communications infrastructure market, is projected to remain the fastest-growing segment for IC microcomponents in the future (figure 6). The communications segment is presently larger than the embedded area and thus is perhaps a better growth candidate for PC chip manufacturers. However, processor companies will face stiff competition from a range of entrenched incumbents in this arena, including AMCC, Broadcom, Globespan, PMC Sierra, and Vitesse Semiconductor. These communications players have established customer relationships, tremendous systems level knowledge, strong technical competencies, and may possess sufficient resources to vigorously resist new entrants. However, even extraordinary growth in the communications markets for the Intels and AMDs of the world may not be able to offset the slow growth of PC related sales, and this may have a profound impact on valuations.
Navigating the Future Despite all of the caution, we continue to believe that changes in technology are inexorable and inevitable. Currently, there is no new “killer app” in the industry, but there will be one at some point in the near future. And that could dramatically alter the competitive landscape. The PC IC companies as well as their communications counterparts must adapt nimbly to take advantage of these strategic inflection points when they occur.
Dipanjan Deb and Keith Toh are respectively founding partner, and associate at Francisco Partners. Francisco Partners was founded by David M. Stanton, Sanford R. Robertson, Benjamin H. Ball, Dipanjan Deb and Neil M. Garfinkel in 1999 to pursue structured investments in technology companies.