CEO Roundup: A Closer Look at 2005

As executive leaders forecast the year ahead, they also examine how changes in their respective fields will shape the outlook. Intertwining technologies and economies affect how these CEOs plan to prepare for what most are predicting to be the beginning of the next downcycle.
 
Arthur W. Zafiropoulo
Chairman & CEO, Ultratech Inc.
 
The semiconductor industry is known for its cyclicality, driven by continually evolving consumer demand. Much of this demand has been sparked by technology revolutions, which create changes that, in turn, drive the demand for conversion technology. The first wave was the PC revolution. The second wave — the digital consumer revolution — has been in effect for several years. Driven by global consumer buying habits, this wave shows no signs of cresting anytime soon, although I believe we will never again see the single product or killer application that matches the significant impact of, say, the digital watch, the calculator or the PC.

The current recovery that began in the United States has expanded into Japan and Asia, and Europe will soon follow. In our industry today, it is believed that there will be reduced spending on pure capacity. As we begin to put more capacity into production, it will be absorbed by the digital revolution. Companies with high market share will only grow to the industry average. To remain competitive in this age of digital revolution, equipment manufacturers must provide customers with tools that incorporate essential conversion technologies.

Consequently, companies that are focused on conversion technologies are expected to see growth above the industry average. Advanced packaging (AP) is also another market where conversion technology is taking place (e.g., bump processing is replacing wire bonding, while wafer-level chip-scale packaging [WLCSP] is growing in popularity to help meet the tight form factor requirements for today's digital devices).

Keith Lee
President & CEO, Advantest America Inc.

Legend has it that an old Chinese saying (or curse, depending on your perspective) goes, "May you live in interesting times." If some industry analysts' recent predictions are to be believed, the near future holds some highly interesting developments, to say the least. In November, a Gartner Inc. analyst projected a 20-40% decrease in the number of semiconductor equipment vendors over the next decade because of accelerated consolidation. Concurrently, the Semiconductor Industry Association (SIA) released its annual forecast for 2004-2007, predicting record sales of $214B in 2004, followed by essentially flat sales in 2005. The SIA attributes next year's dampened industry expansion to what it believes will be a less favorable supply-and-demand balance for memory products than has previously been enjoyed.

Without a crystal ball, it's difficult to assess the degree to which such projections are likely to prove true. What we do know is that demand for semiconductor devices used in consumer products continues to grow at a rapid pace, and that the fierce competition within this market drives the need for lower costs, shorter product cycles and lower consumer selling prices. Supplier survival requires not only R&D strength, but also the ability to leverage those resources as effectively as possible.

In the automated test equipment (ATE) sector, this means adopting a platform solution that can be deployed across a wide range of markets and easily reconfigured for the next product cycle. Typically, ATE providers have created new product platforms every three to five years, forcing customers to continually reinvest in new systems, including infrastructure, training and support costs.

Customers can no longer support this inefficient industry structure. They want a test architecture that is extendible for five to 10 years. OPENSTAR is designed to be the industry's first truly open test architecture. With a variety of hardware and software suppliers, including Advantest, already signed on to support it, the architecture is poised to resolve these issues by delivering true multi-vendor interoperability. In turn, it will enable suppliers to focus their R&D efforts not on continually reinventing the wheel, but on honing new solutions and capabilities that will help customers remain competitive while minimizing their test costs.

Richard S. Hill
Chairman & CEO, Novellus Systems Inc.

Looking back at 2004, we've seen the semiconductor industry add significant capacity to meet growing global demands in arenas, such as PCs, communications and consumer electronics, with each segment showing a 40, 20 and 15% growth, respectively. However, a recent flattening of user demand indicates that the semiconductor industry will need to absorb current overcapacity before making further capital investments. While we have already seen a slight softening in revenues during the latter half of 2004, we do think that we'll see some market drivers emerge in the course of 2005 that will translate into demand for new capacity. We also expect the Asian market to continue to drive growth in the electronics and semiconductor industries.

Nick Gutwein
President & CEO, Rohm and Haas Electronic Materials CMP Technologies
Vice President, Rohm and Haas

As we approach the end of 2004, most experts are predicting 25%+ year-on-year growth in the semiconductor industry, making 2004 one of the best years in history. During the fourth quarter of 2004, inventory corrections and perhaps some slowing end-market demand have temporarily decreased semiconductor wafer starts. Recent announcements by the major foundries indicate that sequential growth has been flat to negative since August. Our outlook for 2005 is for a slow first quarter, followed by some momentum building as the industry works its way out of the inventory over-build (two key assumptions here are that global economies will be solid, and that demand for cell phones and PC refresh continues).

As a material/consumable supplier, Rohm and Haas Electronic Materials CMP Technologies feels the impact of slowing wafer starts a few months later than equipment suppliers (the same is true during a recovery). This kind of cycling is not unusual in the electronics space, and our business model reflects this inherent cyclicality.

Jerry Cutini
President & CEO, Aviza Technology

While 2004 saw digital convergence in the global electronic end markets drive demand for faster, smaller and less expensive semiconductors, 2005 will see a pause in growth in order to rebalance the supply/demand dynamics throughout the electronic value chain. This required correction points to a flat year of growth for the IC industry.

Capital equipment makers will feel the effects of this economic correction. But the technical inflection points traversed in the past year to achieve subwavelength IC manufacturing will continue to drive the evolution of Moore's Law. The migration to 300 mm wafer manufacturing and increased adoption of 90 nm design rules will keep the backrooms at IC makers lively as these requirements — along with the need for new dielectric, metal and composite materials — will sustain R&D efforts in fabs across all regions. Development efforts will also continue with emerging technologies such as atomic-layer deposition, which is aimed at the 65 nm node.

As our industry has matured, experienced companies that have weathered previous storms have come to better understand what it takes to manage their business through a period of cyclicality. Remaining aggressive in the marketplace and ensuring that operational and technical initiatives continue to support growth, profitability and shareholder value will remain the focus of industry management teams.

Edward C. Grady
President & CEO, Brooks Automation

When I look at the big picture for the semiconductor industry, I believe that Moore's Law will be valid for many more years, driven by the proliferation of feature-rich consumer products with shorter product life cycles and more silicon-intensive applications. Leading-edge technology is already reaching critical mass at the 90 nm node, with 65 nm fabs not too far behind, and 45 nm to follow. The result is a secular shift requiring semiconductor equipment manufacturers to focus their R&D investment more on process development and core technologies, which in turn requires them to outsource the development and manufacturing of non-core modules and subsystems to their supply chain.

Looking at 2005, in the absence of any true killer app, the industry has a greater dependency on the consumer end market, which in 2004 crossed the 50% threshold to become the largest category for chip demand. This shift makes the industry more closely linked to macroeconomic conditions — such as the price of oil, interest rates and others — that affect consumers' real disposable income. Unit shipments of semiconductor chips are expected to grow between 5 and 10% in the coming year; however, dollar measured growth is expected to be flat because of expected declines in chip average selling prices and falling capacity utilization at fabs above the 150 nm technology node. While the lower utilization would seem to indicate a challenging year for the semiconductor equipment industry, we do expect continued investment at the leading-edge sub-150 nm nodes, and virtually all of this is expected to be at the 300 mm wafer size fabs and equipment.

The results we expect will be a flat to slightly down year in 2005 for semiconductor capital equipment, with pockets of growth driven by new process technology and accelerated right-sourcing to align business models with lower cost. Many companies in our industry are adopting more efficient business models to anticipate and react quicker to market conditions. As a result, it is hoped that some of the extreme peaks and valleys that have characterized our industry will be reduced somewhat going forward, and more of us will remain profitable through all cycles of the industry.

Marshall C. Turner
Chairman & CEO, DuPont Photomasks Inc.

Revenue for the photomask market in 2004 is expected to show double-digit growth over 2003, according to industry analysts. Although revenue estimates for 2005 are under debate, analysts and we agree on one trend: Demand for leading-edge photomasks — represented by 130 nm and below design rules — is expected to grow robustly. Our own order mix showed leading-edge revenue nearly doubling during our most recent fiscal year. That segment of the photomask market is expected to continue to show strong double-digit growth this year and into the foreseeable future.

In 2005, we can also expect the trend toward greater collaboration in the semiconductor industry to accelerate as chip designers, photomask makers and device manufacturers continue to push the limits of optical lithography. As we do so, the financial costs and intellectual challenges will increase and transform photomask supplier-customer relationships into more highly integrated collaborations, enabling us to continue breaking down technical and economic barriers in innovative ways.

History teaches us that there is a basic human tendency (highly developed among political pundits, stock pickers and semiconductor industry analysts) to speak with the greatest certainty about future events that are the least certain. Thus, our industry gatherings tend to include numerous and dramatic pronouncements, especially toward the end of their cocktail hours. But history teaches us something else. Day in and day out, we will do what we've always done in this amazing industry: Deliver photomask and semiconductor technologies that drive world change.

Roger Blethen
CEO, LTX Corp.

Last year at this time, we said that the semiconductor test equipment (STE) industry was poised for recovery. The first-half recovery was strong for STE. Companies in our industry enjoyed strong revenue and earnings growth. The second half of 2004 was another story; chip inventories and capacity grew at an unsustainable rate, leading to a sharp cutoff in capacity expansion and, as a result, STE orders.

In the STE sector specifically, some suppliers will continue to embrace the migration toward scalable, single-platform testers, while others will struggle with this trend, navigating difficult multi-platform product transitions. All in all, test customers are recognizing that the single-platform approach delivers better performance and far more production-floor flexibility. The STE companies can concentrate R&D spending and applications development on an extensible solution flexible enough to address a wide range of test requirements. We believe the growing preference for single-platform systems will lead to significant market share gains in the coming year for vendors embracing that approach.

In terms of the devices to be tested, in 2005, STE companies can expect greater demand for advanced test capabilities that cost-effectively support high-speed interfaces. In addition, RF and DSP will continue to factor prominently into the industry's test needs, as will increased demand for automotive and power applications. The ability to effectively address the test requirements of these device types will prove critical to success, as they will serve as the key drivers when the STE industry returns to growth.

Peter Podesser
CEO, EV Group

This year promises the usual array of challenges, but also significant opportunities in several emerging markets. We believe that new enabling technologies like nanoimprint lithography (NIL), 3-D integration, and chip-to-wafer bonding will make noteworthy progress from R&D to industrial and commercial applications.

Moreover, niche technologies like silicon on insulator (SOI) will continue their march toward volume applications, and along the way they will need 300 mm processing equipment. Thin-wafer processing techniques, originally at home in the compound semiconductor III-V industry, are now also entering the mainstream silicon world with an emphasis on wafer-level packaging.

From our vantage point, there will be strong demand in Japan and Korea for processing equipment with high-tech content, because those countries have laid the groundwork for emerging technologies. China, too, looks poised to place greater emphasis on new fields like BioMEMS, MEMS and nano applications.

We see a stable demand for high-volume production and next-generation technologies in Europe and possibly America. Even so, equipment spending for R&D appears to be slowing in the United States. Whether that reflects a level of market saturation or other factors remains to be seen.

One thing is clear for 2005 and beyond: Collaboration and partnerships will continue between companies with complementing technologies as they work together to provide solutions that go beyond the mere sale of processing equipment.

André-Jacque Auberton-Hervé
President & CEO, Soitec

In 2005, we see our industry more closely linked to the global economy. The predictable cycles of corporate markets are being replaced by the more volatile and fashion-sensitive markets of the mass consumer. This is a positive trend for Soitec, as SOI has entered these markets as a key enabler to device optimization in the upcoming generations of games, phones, cameras and other consumer electronics. To meet the mobility and convergence that consumers demand, these products require higher-performance chips that run on lower power — the primary realms of SOI.

The flash points of energy consumption and sustainable growth are being reached, igniting the quest for advanced, low-energy solutions across the board — from fuel cells for PCs to high-autonomy mobile products. In parallel, we see growth in the RF market, as well as in the convergence of photons and electrons. Advanced materials for optoelectronics and related markets such as lighting will emerge with solutions built on new generations of Smart Cut-enabled engineered substrates. For those looking further down the road, engineered substrates are also the enabling platform for nano and bio-nanotechnologies.

From a technology standpoint, advanced materials have entered the mainstream. Our customers — amongst whom we count virtually every major chipmaker in the world — are fine-tuning their device optimization choices. This may be local or global strained silicon, or some combination thereof, or the integration of strategies such as crystal orientation and non-planar architectures. Whatever the choice, SOI is the platform. Cost is the critical factor, and SOI makes scaling more cost-efficient.

We are in high-growth mode, and will be doubling our capacity in 2005. Those of us in the advanced materials market are looking forward to a dynamic 2005.

Sven Löfquist
CEO, Micronic Laser Systems Inc.

Those of us serving both the semiconductor and display markets see two very different dynamics. The high-end display market is transitioning from a pure PC play to the home consumer market. The result has been major growth in generation 6 and 7 TFT displays, and a market size that is beginning to rival the semiconductor opportunities. We continue to forecast a strong market in 2005 as consumers continue to purchase large displays for their homes.

In the semiconductor market, we see a continuing consolidation of designs and a rationalization of the photomask business, even as IC volumes increase. As designs become more complex, enabling and cost-effective design solutions will be the drivers. This means that all tools and processes will need high levels of manufacturing flow integration to achieve top efficiency and yields.

As the IC continues to penetrate the consumer market, the operative phrase in the next 10-20 years will be product cycle time. The consumer will be constantly looking for new products with an ever-increasing emphasis on services. The hardware will need to be cheap, and will be in a constant state of change. The IC industry will need to focus on low-cost, quick-to-market products. The challenge for our industry is to continue to improve efficiencies in an increasingly complex design and manufacturing environment.

Paul Raymond
Vice President & General Manager, Honeywell Electronic Materials

Recent history reveals that relying solely on delivering a competitively priced product or technology to customers within the global semiconductor industry does not necessarily lead to growth. Honeywell Electronic Materials and other suppliers must have a unique competitive advantage to succeed in a cyclical industry impacted by ongoing cost pressures. Moreover, they must focus investment on R&D, leading to innovation that is market-driven and meets ever-critical customer demands.

However, it doesn't end with materials suppliers. There is increasing emphasis on integration to jointly develop materials between end users, equipment suppliers and materials suppliers. This trend helps the industry face new challenges that include the maturing of the desktop market as well as the flooding of the handheld market, which has yet to provide a winner. The need for standards in the mobile computing marketplace remains a concern that must be resolved.