New Book Confronts Issues Challenging Semiconductor Industry

A wide range of issues important to the U.S. semiconductor industry—and a strategy to correct what some experts see as a potentially dangerous trend of declining long-term basic research and development—are discussed in a volume of scientific and engineering papers being published today.

The book will be made available during the American Vacuum Society's 45th International Symposium, Nov. 2-6, in Baltimore. The volume, totaling more than 1,000 pages, is entitled Characterization and Metrology for ULSI Technology. It was published by the American Institute of Physics (AIP #449) and edited by six scientists led by Dave Seiler, a physicist at the Commerce Department's National Institute of Standards and Technology.

"What's at stake is American jobs and the preeminence of the U.S. semiconductor and electronics industry, which is, in a broad sense, the single largest employer in the nation," said Seiler, chief of the Semiconductor Electronics Division of the NIST Electronics and Electrical Engineering Laboratory. "People are not doing the basic research that will be needed 15 to 20 years from now. Who is going to pick up that research? Where will the breakthroughs come from?"

The papers specifically address issues concerning how to measure and understand the materials and technologies involving "ultra large scale integration," or the production of larger and larger circuits that contain more and more densely packed electronic components. Among the issues discussed in the papers—many of which are highly technical—is the concept of meeting future research-and-development needs by forming partnerships among industry, universities and government agencies. The strategy of forming consortia is referred to as a "cross-functional approach" to the problem.

But who might fund, organize and support the consortia? Universities are finding it more difficult to maintain costly semiconductor research facilities. At the same time, government and industry funding for long-range semiconductor research has declined in recent years.

In a sense, today's chip makers are living off of the ideas of past years, when large American companies could afford to sink large percentages of their annual revenues into research considered high risk. But in today's climate of increasing international competition, economic problems in Asia and smaller overall profits, those same U.S. companies are laying off or retiring many of their research scientists.

"They are doing it for survival," said Thomas Shaffner, a physicist who spent 21 years in the semiconductor industry before joining NIST in September. "It's a cost- containment phenomenon."

Meanwhile, the technology for semiconductor electronics is getting ever more complex, requiring an increase—not a decrease—in research and development. New materials are emerging, and chips with smaller components and denser circuitry are being designed in efforts to stretch the physical and economic limits of current technology. If those limits are reached before new ideas can be developed and exploited, the semiconductor industry will be unable to maintain its fast growth, a pace that has seen the computing power of chips doubling every 18 months.

"In my opinion, the industry is moving away from basic research just at the time that it's needed the most," said Shaffner, one of the book's editors and leader of the Materials Technology Group in the NIST Semiconductor Electronics Division. "Everyone has the same problem, even overseas, to some extent," Shaffner said. "So, the overall effect is that the progress of the whole industry could slow down."

The current price for a state-of-the-art semiconductor fabrication facility is about $2 billion. As the demands grow for smaller and smaller devices, so, too, will the cost of fabrication.

As a non-regulatory agency of the U.S. Department of Commerce's Technology Administration, NIST promotes economic growth by working with industry to develop and apply technology, measurements and standards through four partnerships: the Measurement and Standards Laboratories, the Advanced Technology Program, the Manufacturing Extension Partnership and the Baldrige National Quality Program.