The Advanced Materials and Processing Program and the Restructuring of Materials Science and Technology in the United States: From Research to Manufacturing (1995)

Some 20 years ago the United States was the leader in the automotive, consumer electronic, machine tool, steel, and semiconductor industries. Subsequently leadership was in large part lost to overseas competitors. Dr. George described in his talk what the semiconductor industry is doing to reassert itself in this fiercely competitive arena.

Electronics has become the largest U.S. industry. Semiconductor technology is the driving force for the information age. The U.S. semiconductor industry must maintain its leadership if this country's information-based industries are to remain competitive in the global marketplace. Of the top 10 semiconductor manufacturers in the world, only three are still U.S.-based. Additional Pacific Rim countries are showing increasing strength. Japan, Korea, and Taiwan haven't had the trouble we do differentiating between politics and economics. These countries represent our greatest competition but, at the same time, our greatest opportunity for growth. If the United States is to remain competitive, serious technological and manufacturing issues have to be confronted. Additional progress toward these goals can best be achieved through improved teamwork among industry, academia, and government on precompetitive, nonduplicative technology issues. Today's semiconductor manufacturing technology, as an example, is almost too complex for a single company to have all of the capabilities to compete on a global basis. A major trend, over the last few years, has been the cooperative linkages, alliances, and partnerships of corporations from the United States, Europe, and the Pacific Rim to develop and market specific semiconductor products.

In 1987, the U.S. semiconductor industry and government embarked on an experiment designed to reassert the semiconductor industry 's leadership and our manufacturing capability. The Semiconductor Industry Association (SIA) formed a separate organization called Sematech (for Semiconductor Manufacturing Technology), which also included the government through ARPA and the academic community. It was designed to enable industry members to share precompetitive technology. Sematech has become one of the benchmark examples of what collaboration between government, industry, and academia can accomplish if a single mission is agreed upon. Sematech's mission is focused on bringing fundamental change to the semiconductor-manufacturing equipment industry. Estimates by the SIA indicate that the subsequent turnaround in U.S. semiconductor market share worldwide can be equated to about 17,000 jobs. Specific successes include development of a single set of qualifications for semiconductor manufacturing equipment, the establishment of software standards, the improvement of existing equipment, and the development of new diffusion, deposition, and etching equipment, as well as standardizing equipment interfaces.

Listed below are some of the major principles underlying the formation and operation of Sematech. First, it has joint industry-government funding. The investment and return on combined industry-government R&D is substantially greater than what is realized by isolated companies or even groups of companies within one industry. Secondly, Sematech represents over 80% of the semiconductor construction industry, which gives credibility to programs and provides access to the nation' s best technology and best people. Third, the consortium uses member

company personnel, which facilitates technology transfer. About two-thirds of Sematech's personnel are assignees from the member companies. Fourth, Sematech acts as a champion for the industry through participation, direction, and support of its executives. Fifth, Sematech has a clear mission (to create fundamental change in semiconductor manufacturing). Sixth, a long-term commitment is required for a collaborative effort to have a benefit. Seventh, it must be possible to measure success in a quantifiable manner. In Sematech's case success is gauged by improved market share for the U.S. semiconductor and semiconductor-manufacturing equipment industries. These criteria also imply the need for a clear timetable for the development of techniques and technologies, in order to take advantage of windows of opportunity in the market. Lastly, it is necessary to develop a total quality process. The criteria developed for the Malcolm Baldrige Award are used as the foundation of Sematech's quality programs.

Skeptics and critics state that, because global companies form global alliances with global competitors, there is no need for national support for this partnership between government and industry. This view ignores several realities. First, globalization of the market and the intensity of international competition have forced companies to compete and to cooperate around the world. If it is to prosper, any major semiconductor manufacturer must have multiple avenues for product-specific technology development and application. But there is a fundamental difference between product-specific alliances and precompetitive R&D consortia that's being overlooked. Alliances between member companies and foreign corporations are endeavors formed to design or produce a certain product and create market access, which would otherwise be impossible. Such joint ventures allow companies to share risks and rewards with international partners.

One of the peculiarities in modern politics is that many politicians spend most of their time on issues that have little or no bearing on the future prosperity of our nation. Diplomacy continues to outrank economics and finance, which, in turn, outrank science and technology, and yet we are living in the age of the microchip. Since the Depression, technology change is reckoned to have accounted for about two-thirds of the rise in our global standard of living. Many of our economic rivals, notably Japan, recognized that fact long ago and did something about it, and yet we continue to take U.S. industrial dominance for granted and pay no attention to economic competitiveness.

Our collective attempt to reorient U.S. federal science and technology policy is still in its embryonic stage. What is now happening in the United States has the potential to become a model for a new breed of intelligent industrial policy—if, that is, more people wake up to the importance of science.

Dr. McGroddy described the changes being made recently at IBM research. He noted that these changes are driven by forces that affect all of us, and that more changes are coming. For example, the number of major independent fabricators of semiconductor chips will decline significantly over the next 10 years. The survivors will be large global companies.

The research organization at IBM has thought of itself as having two goals: to do world-class science, and to become “entangled” in the affairs of the company by forming tight links with the product development organizations. The research organization acts to accelerate the pace of technology in the company, especially where major changes are taking place. The research organization also thinks of itself as being responsible for looking at what IBM isn't doing, figuring out what it should be doing that it isn't, and driving the company into new business enterprises. This latter mission is essential, because looking 5 to 10 years in the future, half of the growth will come from things being done now but the other

half will be associated with currently nonexistent activities.

IBM is in three different kinds of businesses. The first is an integrated solutions business, that is, working with a company to help it use information technology to compete more effectively. Although this business area is small (about 10% of IBM's worldwide revenues), it has the fastest growth and is the area on which IBM's customers place the highest value. Second, there are the more traditional businesses that IBM is popularly identified with, software and systems products of various sorts. And third is a set of technology businesses. IBM is one of the largest semiconductor and packaging producers in the United States, but traditionally these products have been aimed at internal customers.

Customers of the information industry place the greatest value on the use of information technology to effectively improve their businesses. The value customers place on products is lower. The technology in the products is unimportant to a customer in itself. In the past at IBM, technical resources have been largely directed to the lower portions of the value chain. A year or so ago, only some 5% of the technical resources went into this highly valued sector of services, applications, and solutions, whereas about 40% went into the products and software businesses. Semiconductor technologies and storage received most of the balance despite the fact that there are fewer streams of technology to develop and more collaboration in developing them. About 3% went to basic science.

Technical resources are now being redirected to the area perceived as having the greatest value to the customer, namely, to services, applications, and solutions. There is a rapid increase in the amount of work being done that is aimed at the kind of result that can differentiate one customer's business from another's. Resources for basic science, for work that does not have any immediate connection with any of IBM's products, technologies, or services, will remain at about 3% of the total technical resource.

In the area of technology, Dr. McGroddy discussed the example of IBM's new laptop computer. Rather than try to build up from nothing a liquid crystal display manufacturing capability, IBM has formed an alliance with Toshiba to co-develop the technology and build a joint venture manufacturing company. Both IBM and Toshiba get their displays from that factory but compete in the marketplace for the laptops. IBM has shifted its investment in terms of technology to areas where there appears to be far more opportunity for progress and for competitive advantage. Dr. McGroddy noted in passing that it is rare to find anyone in U.S. universities who is working on problems connected with liquid crystal displays. There has been an enormous focus on III-V compounds, where the downstream opportunity is clearly not very large at this time.

Several examples were given of research innovations that deliver high value to customers, such as the use of three-dimensional car design data to make high-definition movies of what the car will look like and thus eliminate the use of clay models (a technique now finding other applications, as in the movie industry), and the use of imaging technology to do presurgical planning for surgery and development of a set of robots for use in the operating room to implement what the surgeon has decided upon.

The mission of research at IBM has changed over the years. Many years ago a primary value was placed on publishing papers and receiving invitations to give talks. Later it was seen that it is necessary to become connected to the business side of IBM and build joint programs to connect research to other parts of the businesses. Now there is a move to take results straight to the customer. In some cases, that has required entities to be set up directly from IBM research in the marketplace without much involvement of an IBM intermediary.

Dr. McGroddy was asked what changes need to be made in the education of scientists to enable them to play this new role. He responded that it is necessary to set expectations that are broader than they are at present. Universities should focus on an educational experience that prepares students for life in an

incredibly dynamic world, a world in which people are going to continue to have to learn new things. If, for example, you're training physicists, you shouldn't simply set the expectation that the students will do only physics. Rather, the focus should be on what you can do if you're trained as a physicist. Dr. McGroddy also stressed the value of broad curiosity that leads people to learn things they don 't need to know. Progress takes place when people are willing to try something new and have learned enough about it that they're not afraid to take that step across the line into something totally new in character.

In his talk, Dr. Brinkman described some of the overriding technology changes that are transforming the nature of research institutions.

The first of these are the large integrated systems that are dominating the computer and communications industry. These include integrated circuits, flat panel displays, and magnetic and magneto-optic storage. All follow similar Moore-type plots. Photonic transmission, though somewhat different in character, also fits a Moore plot if the parameter is bit rate through a fiber. Photonic transmission is becoming ubiquitous. No slowdown is envisioned in the development of these systems until after the year 2000. The issue is not what the devices are going to be, but rather how to make them. Continually developing the necessary processes will drive materials science to a large extent for the next 10 years. The cost of fabrication of integrated circuits, panel displays, optical fiber, and other systems is escalating. These enormous fabrication costs are driving changes in research behavior. Research cannot be done independent of manufacturing and development.

A revolution is taking place because of what silicon can do. Today a single semiconductor chip will hold a book. Tomorrow it's going to hold an encyclopedia. The number of integrated circuits and discrete devices in home computers, modems, and cellular phones has plummeted. Speech processing used to be thought of as hard. Today it's done with a microprocessor. According to the business plan, when the new fiber-optic TAT-12 is laid across the Atlantic in 1996, the cost for one voice channel for a year will be $400. Voice communication is becoming a commodity.

A second technology change that is affecting the nature of research institutions is the advent of high-leverage, high-technology products. The semiconductor transmission laser for high-speed transmission is an example. Sales of 10,000 units a year constitute big business in this area. Other examples are the erbium amplifier and high-speed electronics. Since the volume of sales of such items is low, the research laboratory effectively becomes the factory.

The ubiquitous high quality of electronic products is a third technology change that is transforming research institutions. There is very little differentiation based on hardware. Most electronics are commodities. Corporations are finding out that captive suppliers don't work, and so the case seldom arises now that a captive supplier is supplying a company with something that no one else has. Research now has to work with an organization that is very much stressed by competition.

A final transforming factor is multimedia frenzy. The enormous bandwidth of fiber opens up the possibility of transmitting a vast amount of data. The future of the telecommunications and computing business is a combined information infrastructure enterprise: multimedia network computing, electronic markets, multimedia information services, and multimedia home services. The problem is that today no one really knows how to do all this from a networking point of view. There is an enormous amount of research that needs to be done, with more emphasis on systems, software and algorithms, and switching technology.

From a technology point of view, these are a few of the things that are driving research organizations today. These dynamic technology changes are really changing our industry.

Mr. Lovell traced the history of Boeing's commercial airplanes and then described planes now being developed, especially the 777, and plans for future aircraft.

The 777 is a huge plane, with a horizontal tail section wider than the wing span of a 737. It will involve about half the total assets of the company in development costs before the first plane is delivered. The design of the plane has been totally paperless. It was entirely computer designed, including preassembly and tooling. The 777 contains a large number of materials never used in a commercial plane before, but Boeing has significantly reduced its in-house testing of materials by providing specifications and test methods for materials and then (after a qualification process) accepting the suppliers' test results. When a new plane is being developed the suppliers are called in and given the requirements, and they then do all the materials development. Boeing does not compete with the materials suppliers and thus has complete access to all their confidential information. While the results of expensive in-house fatigue tests, for example, are kept proprietary, the materials specifications and standards are in the public domain. The more other people want to purchase the materials that Boeing is using, the cheaper they will be.

Various parts of the 777 are designed and built in a number of countries around the world. The design technology, for example, was transferred to Japan. Japan is building the body section of the 777. Such a global operation is essential in order to be able to sell planes abroad. Ninety-five per cent of the planes of Japan's airlines are Boeing, thanks to this sort of cooperation. Assembly of the planes has remained in Seattle, and thus a large number of high-wage jobs are kept in the United States.

Total quality management is essential when dealing with the various teams in a big project like the 777. Boeing used to think its core competencies involved objects such as wings. Now it realizes that the core competencies involve processes for how things are done.

Dr. Williams' talk concerned materials for jet engine turbines. In this context, “lightweight” means a density that is low compared to that of nickel-base alloys. Such materials must have high stiffness to permit the use of close tolerances between the rotating and stationary parts of an engine, the capability to withstand high temperatures to improve efficiency, environmental stability in highly oxidizing conditions, and ability to withstand impact forces. Several promising advanced materials were discussed: polymer matrix-graphite reinforced composites, metal matrix composites, and ceramic matrix composites. In all cases, there is a lack of an industrial base for making these materials. Until the cost comes down they will not be used, but price reduction will occur only when there is a widespread market for them. This is an area where some help from the government is needed, along with cooperation among the members of the aircraft engine industry to concentrate on developing only an agreed upon, select few of the most promising new materials. As the industry moves to the new engineered materials, new capitalization is going to be required. Without an adequate market, the market pull will not induce the private sector to invest enough capital to produce a significant supplier base. A strong vendor base is critical to retention of the

competitiveness of the U.S. aircraft industry. New materials are key to better products, but these new materials are going to be fewer and further between because of the realities of cost and implementation.

A few years ago, nine U.S. companies with interest in optoelectronics founded the Optoelectronics Industry Development Association (OIDA). This step was a response to an action taken a number of years earlier in Japan. In 1980, with substantial funding from MITI, the Japanese industry formed OITDA. OITDA focused industry investments on those opportunities with the greatest long-term potential. Japan's share of the optoelectronic component market went from 30% in 1980 to 70% in 1990 while the U.S. share shrank from 50% to 15%. OIDA quickly realized that it must imitate OITDA and try to become the focal point for the vision of where optoelectronics is going over the next 10 to 20 years. Preliminary data are presented here from Phase I of the technology road map being put together by OIDA. It deals with markets.

In the huge computer market, the portable computer is enabled by an optoelectronic component that dominates its cost, the flat panel display. In the area of peripherals, the optoelectronic content is very heavy in optical storage as well as in laser printers and displays. The optoelectronic equipment market also has been studied for the telecommunications market and for the industrial, medical, energy, transportation, military/aerospace, and consumer electronics markets. Today the computer and consumer electronics markets dominate but over the next 20 years all but the military market will grow to a level of respectability while the military will disappear into the background noise. Display-based equipment represents approximately two-thirds of the equipment market enabled by key optoelectronic components.

The market value of the optoelectronic component in a laser printer, a compact disk player, or an erasable disk represents only a few per cent of the total cost. However, in the case of display-based applications (computer monitor, TV, portable computer), the optoelectronic component makes up something like a third of the cost. Ninety per cent of the content of optoelectronic components is in displays, and this situation is not likely to change in the next 20 years.

Future challenges and areas of growth include faster and cheaper laser printing in color, color copying with lasers, high-density optical storage, multimode-based fiber optic communications, and, most important, cheaper and improved flat panel displays.