Emerging Keys To The Future
The International SEMATECH Manufacturing Initiative (ISMI) feel this is of vital importance for the industry and has developed the Industry Economic Model (IEM), a sophisticated software tool that can analyse various aspects of semiconductor economics. While the IEM business logic and software are presently proprietary, accessible only by ISMI member companies; its' scenario content is made available to selective collaborative user groups as an industry resource for purposes of building consensus on a set of default metrics and/or demand that best represent the semiconductor industry's strategic paths. In addition, its' principles are being extended to emerging related industries as well, such as nanotechnology and microelectromechanical systems (MEMS), for the purpose of establishing a common methodology for evaluating business strategies.
ISMI brings together global leaders in the form of forums and symposiums to ensure the most relevant information is utilised in the IEM. The recent ISMI's Global Economic Symposium suggests that the recovering chip industry's best long-term prospects appear to lie in emerging industries and applications.
Industry Economics: Reason to Smile Again
After the tech bust of 2000, our industry entered into an earnings slump that was so long and severe that mere survival became the dominant business strategy for many companies. This year has brought a respite that some observers originally thought would be short-lived, but new data suggests the recovery may be more sustained, although with more modest growth.
Dan Tracy, Senior Director of Industry Research and Statistics for SEMI, reported strong bookings and billing trends for semiconductor wafer processing equipment in early 2004. He said the improvement in the capital equipment market is fueled by strong unit demand for semiconductors, and by rising fab capacity utilization levels.
SEMI has tracked equipment data trends for over a decade, and Figure 1 shows the correlation of fab equipment bookings (net) turning into billings in one quarter. Through the first four months of this year, the one quarter bookings-into-billings correlation remains stable, unlike three previous periods in the past where a deviation in this trend coincided with an industry downturn. Stable booking levels are indicative of healthy growth as semiconductor manufacturers invest in capacity expansion that is rational, Tracy noted.
Tracy also said the Symposium should look for industry revenues of $210B in 2005 and $224B the following year -with both totals exceeding the $2004 performance of our last boom year in 2000. In all, the cards appear stacked against another huge downturn anytime soon, though we could well see another hiccup after next year.
…But Don't Get Too Comfortable
In counterpoint, James Feldhan, president of Semico Research Corp., said in a keynote address that a strong first half in 2004 could push the industry past the 27.5% growth rate that Semico had previously predicted for this year. That increase is being powered by strong corporate IT spending, PC upgrades, greater consumer purchases, and restocking of semiconductor inventories, Feldhan said.
But, he warned, past market surges greater than 30% have been followed by significant declines in revenue growth. Macroeconomics suggests a slowdown in 2005, he said, as negative influences - among them greater inflation, higher fuel costs, and diminished PC upgrades - absorb consumer and corporate spending and so take their toll on the chip industry.
Nevertheless, the long-term outlook appears ripe with possibility, with Feldhan projecting as the next growth driver a “portability and connectivity wave” energized by fuel cells, rich media, and multiple wireless devices. Fuel cells for portable electronics will be the next “killer catalyst” for the industry, bringing notebooks with battery life of two or more weeks, and cell phone batteries good for a month. “This,” he noted, “will be a new paradigm and a new price/cost value proposition!”
Nanotechnology: Leadership is Up for Grabs
Early indications are that nanotech will amount to the world's next great industrial revolution, and governments across the earth are competing to capture its center, said F. Mark Modzelewski, former Executive Director of the NanoBusiness Alliance. Academic programs are developing rapidly, he said, with over 50 already in the U.S. and 50 more in other nations. And he noted that venture capital investment in the nascent industry topped $1B in 2001-03 and may reach $700 million in 2004 alone.
“Nanotechology research is happening in over 50 nations, with more focusing on it as a source of economic potential every day,” the speaker pointed out. “Many countries - while suffering economically - are beginning to look at recovery and are turning to emerging technologies such as nanotech to drive future economic development.”
Such rising expectations may not guarantee Americans a leading position in nanotech, which is expected to become a $200B industry by 2006, Modzelewski continued. “Most would concur that the U.S. is leading in nanotechnology,” he said. “However, unlike past waves of innovation, nanotechnology is not dominated by the United States.”
For example, he said spending by European Union governments on nanotech R&D is nearly double that of the U.S., while national initiatives (similar to the U.S. National Nanotechnology Initiative) have been established in Japan, China, Canada, Israel, Switzerland and Australia.
Irrespective of borders, multinationals such as IBM, Hewlett-Packard, ChevronTexaco, Samsung, Mitsubishi, and DaimlerChrysler are conducting internal nanotech R&D efforts of their own. And on the small end of the scale, more than 1500 nanotech startups (most with fewer than 10 employees) have emerged around the world. Modzelewski profiled a few of these companies and products:
· Molecular Imprints, Inc. of Austin, TX pioneered Step and Flash Imprint Lithography, based on the ancient craft of embossing, for semiconductor applications.
· Kovio, Inc. of Sunnyvale, CA developed a radical new semiconductor manufacturing technology to directly print circuit patterns on semiconductors, without the need for etching.
· ZettaCore, Inc. of Denver, CO devised molecular memory chips that use specially designed molecules in which electrons can be added or subtracted to store information.
Although hyperbole and science fiction have clouded the outlook for nanotech, there's little doubt it will someday touch our lives as deeply as electricity and information technology (see Figure 2 for perspective). To hear Modzelweski tell it, “Nanotechnology will affect almost every aspect of our lives - from the medicines we take, to computers we use, to the energy supplies we require, to the food we eat, to the cars we drive, to the buildings we live in, to the clothes we wear.” And he adds that other products, capabilities and markets that we haven't even imagined are yet to emerge.
MEMS: Evolution in the Making
Not quite emulating the flash of nanotech but sometimes burdened with similar hype, microelectromechanical systems (MEMS) are seeping into industrial applications and daily life in ways that will guarantee this new technology an interesting future. That was the message of Ellen M. McDevitt, Managing Director of the MEMS Industry Group, who portrayed MEMS more as a supporting technology than a groundbreaking one.
McDevitt added that the MEMS industry, now reckoned to grow to $9B in revenues by 2007, seeks innovative solutions for existing applications (such as making air bags more reliable) rather than disruptive approaches. “Customers do not want to be disrupted,” she said. “Remember that Thomas Edison made his light bulb look like a gas light.”
In a typical MEMS device, McDevitt said, integrated circuits provide the thinking part of a system while the MEMS component complements this intelligence with perception and control functions. An example is a microsensor, which gathers information (inertial, acoustic, thermal, chemical, etc.) that is processed by an IC component which in turn directs actuators to respond by moving, pumping, filtering, or other action, based on a set of instructions.
But though MEMS integrate well with semiconductors, there are significant differences between the industries. For one, MEMS devices are made in smaller batches than chips, need nonstandard materials, and require more complex process recipes, packaging, and end-stage production.
“MEMS device manufacturers want to use the cost advantages of batch manufacturing methodologies to produce high volumes of smart, affordable, electromechanical devices and systems - the same way IC producers use photolithographic techniques to mass-produce millions of chips,” McDevitt said.
The problem, she noted, is that attractive cost models for transitioning from prototypes to volume manufacturing are difficult, although not impossible, to achieve. It could be that the solutions for growing this industry lie in collaboration, stronger value-chain relationships, and deciding whether potential market size makes a MEMS-based solution worthwhile.
Making Sense of It All with the IEM
As valuable as expert opinions are, corporate leaders still need statistical rigor in making decisions that can make or break a company. For the semiconductor sector and the emerging technologies related to it, the IEM provides a unique tool where great quantities of information can be integrated at an industry level, and manipulated to produce quantitative results that can be analyzed. (Figure 3 shows model inputs and information flow.) Using the IEM, industry representatives can come together as a group to examine the implications of current technology directions. The recent Global Economic Symposium provided an opportunity for several industries to share their perceptions and IEM-based analyses, and ISMI looks forward to enabling similar exchanges in the near future.
ISMI is a global alliance of the world's major semiconductor manufacturers, dedicated to reducing cost per wafer, and ultimately cost per die, through cooperative programs focused on manufacturing effectiveness. The consortium conducts programs in manufacturing infrastructure, methods, standards, and productivity, with the aim of reducing the costs of producing finished wafers and chips and driving solutions to major productivity challenges. ISMI is a wholly owned subsidiary of International SEMATECH of Austin, TX.