Below are the first 10 and last 10 pages of uncorrected machine-read text (when available) of this chapter, followed by the top 30 algorithmically extracted key phrases from the chapter as a whole.
Intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text on the opening pages of each chapter. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.
Do not use for reproduction, copying, pasting, or reading; exclusively for search engines.
OCR for page 76
R&D in the Framework of the New Transatlantic Agenda Kenneth Flamm, Moderator Brookings Institution In setting the stage for the session Dr. Flamm said that the semiconductor industry's conversion from the 200-mm wafer to 300 mm raised not only interest- ing technical issues but also important organizational and cultural issues for the industry. Some might argue that the I300I initiative represents the first international effort to develop cooperatively a new equipment set for semiconductor produc- tion. Dr. Flamm said that usually a single company in the industry "bites the bullet," invests in new equipment, and debugs it at its own facility. The very large scale integration (VLSI) project in Japan in the 1970s was, however, an early example of cooperation in equipment development and deployment. The VLSI project did not deal with the entire suite of semiconductor production equipment, as I300I seeks to, but did address a large portion of it. It would be hard to dispute the notion, said Dr. Flamm, that the VLSI project contributed greatly to Japan's market dominance in semiconductors in the 1980s. One issue to consider in think- ing about I300I is how unique cooperation in semiconductor equipment develop- ment really is. Nonetheless, Dr. Flamm said that the I300I effort is the first truly interna- tional and formally organized enterprise on joint standards and tool development. The international element makes I300I distinctive and, Dr. Flamm hoped, a sub- ject worthy of vigorous discussion. Among the issues that I300I raises are: . Economic fundamentals. As Gordon Moore pointed out in a prior session, it cost $200 million to $300 million in the 1980s to build a fabrication facility. Currently that figure is in the $1.5 billion neighborhood, making 76
OCR for page 77
R&D IN THE FRAMEWORK OF THE NEW TRANSATLANTIC AGENDA 77 it "extraordinarily expensive to do anything in the industry." Rising costs and risks make alliances and cooperation attractive options and looking to international partners becomes a strategy to consider seriously. · Cooperation versus competition. This issue is paramount to the I300I project. On the one hand, the semiconductor industry has always been very competitive, and this has driven innovation. On the other hand, coop- eration looks increasingly attractive as fabrication costs continue to rise. Dr. Flamm noted an interesting pattern with respect to cooperation. Coop- eration in the semiconductor industry was pioneered by the Japanese with the VLSI project in the 1970s. The ensuing Japanese success resulted in a U.S. cooperative response in the 1980s, namely SEMATECH. SEMATECH was largely a process of experimenting with the right orga- nizational approach to make cooperation work in a U.S. setting. Just as SEMATECH was winding down as a government-industry collaboration in the 1990s, Japan, inspired by its apparent success, embarked on a new round of cooperative programs, such as Semiconductor Leading Edge Technology (SELETE), to develop tools to build 300-mm wafers, and the Association of Super-Advanced Electronics Technologies (ASET). Cooperation and competition across international boundaries. Dr. Flamm recalled negotiations between the United States and Japan on the Semi- conductor Trade Agreement in the early 1990s during which Japanese negotiators argued that national boundaries were no longer relevant. The concept of borders was an "atavistic anachronism that we [the United States] had to free ourselves from." The irony today is that Japan's new cooperative efforts, such as SELETE and ASET, are basically exclusive programs for Japanese firms. In contrast, with the I300I enterprise, SEMATECH has pioneered the idea of "international technology policy cooperation," and SEMATECH has even considered allowing foreign firms into the consortium. The crux of the national policy problem, Dr. Flamm said, is a tradeoff be- tween technology diffusion and national advantage. Cooperation in certain areas can push the technology frontier forward at a faster pace for everyone. If that is done, however, the notion of national advantage in the control of a technology is set aside. This tradeoff could become more evident in I300I, and it has become an . issue in the United States in the controversy over the extreme ultraviolet lithogra- phy consortium, an effort by several U.S. companies to use technology generated by U.S. national labs in conjunction with Japanese companies to advance lithog- raphy technology. In concluding, Dr. Flamm recalled from his experience at the U.S. Department of Defense (DOD) the attitude toward foreign suppliers. He said that DOD operated under three unwritten rules in considering work with foreign suppliers: · Monopoly. DOD did not want a key component monopolized by a firm
OCR for page 78
78 NEW VISTAS IN T^NSAT~IC SCIENCE AND TECHNOLOGY COOPERATION . from one country. If there was a diverse source of suppliers around the globe, DOD saw little problem with using foreign suppliers. · Plausible threat of supply cutoff. If a choice was made to use a single foreign supplier or a set of foreign suppliers, is there a plausible threat of being deprived of a key component? If such a threat existed, DOD would prefer not to use foreign suppliers. Technical advantage. In some cases, DOD wants to be better than anyone else in a certain technical area but realizes that it cannot be the best in all areas. When the DOD wants strategic superiority in a technical area, it might prefer a U.S. supplier. If that is not the case, it might welcome cooperation with international partners. In turning the program over to William Spencer of SEMATECH, Dr. Flamm observed that business may have a similar set of "unwritten rules" when consid- ering whether to engage in international cooperation. THE 300-MM INTERNATIONAL INIATIVE William Spencer SEMA TECH Dr. Spencer said that his remarks would focus mainly on the process of bring- ing the I300I program together, as opposed to the technical details of the conver- sion from a 200-mm wafer standard to a 300-mm standard. He agreed with Dr. Flamm that the I300I initiative represents a unique form of international collabo- ration, and Dr. Spencer emphasized that the collaboration was on technology, not basic science. I300I is a bottom-up initiative of private companies and receives no funding from any government. Dr. Spencer also noted that I300I focuses on precompetitive technology for member firms, although what is precompetitive for Gordon Moore at Intel is very competitive for a semiconductor manufacturing equipment supplier such as John Shamaly at the Silicon Valley Group. Lessons from SEMATECH In developing I300I, which is a separate subsidiary of SEMATECH, man- agement decided to stay away from intellectual property issues. This was a lesson from SEMATECH, which has filed relatively few patents in its 10 years of exist- ence and has no trade secrets. The movement of people has been an important contributor to SEMATECH's success; SEMATECH has over 200 member-com- pany employees, known as assignees, in its work force. For I300I about 50 em- ployees from member firms are assigned to the project. Using people as the main tool for technology transfer will be as prominent in I300I as it has been in SEMATECH.
OCR for page 79
R&D IN THE FRAMEWORK OF THE NEW TRANSATLANTIC AGENDA Importance of the Semiconductor Industry 79 Dr. Spencer then turned to the topic of semiconductors in everyday life and observed that he would cover some of the same ground that Gordon Moore had in a prior session. The typical automobile today contains more computing power than the lunar module that landed on the moon nearly 30 years ago. Early com- munications satellites, a technology Dr. Spencer worked on, had approximately 2,000 transistors and orbited about 100 miles above earth. Today, communica- tions satellites orbit at 200 times the distance and contain roughly 100 times the number of integrated circuits as the first satellites had transistors. The technology that enables these advances is the semiconductor. The semiconductor industry supports a worldwide electronics industry whose sales are approximately $1 tril- lion, with 15 to 20 percent of that being semiconductor sales. Over time, sales in the semiconductor industry have grown rapidly, as the performance of devices has improved. In addition to the growth of semiconductor industry sales, the equipment and materials firms that support the semiconductor industry have grown quickly. Dr. Spencer stressed two points about the future growth of the industry. First, if the semiconductor industry is to adhere to the pattern of making things "faster, smaller, and cheaper," technology in the semiconductor equipment sector must advance. About 75 percent of future improvement in device functionality would come from better manufacturing equipment, with the remaining improvement coming from changes in the structure of the transistor and other design advances. Second, future growth would depend on transition to a 300-mm wafer, but the slump in industry sales beginning in 1996 and continuing through 1998 has af- fected the I300I initiative. There are fewer resources available for wafer conver- sion. Even with the current slump, long-term growth in the industry is expected to be 15 percent, approximately the historic growth rate. Conversion to 300-mm Wafers In discussing the change from 200- to 300-mm wafers, Dr. Spencer stressed the improvements in functionality of the semiconductor. In terms of cost per func- tion, the semiconductor industry has been advancing at a rate of 25 to 30 percent per year for the past 30 years. That is, the combination of more circuits per chip and less expensive chips has yielded productivity improvements approaching 30 percent per year for 30 years. Important drivers to this productivity have been larger wafers and better interconnection technology between circuits. When the industry retooled for 200-mm wafers, moving from 150 mm, one U.S. firm purchased the full suite of tools for 200-mm fabrication and paid for much of the tool development. When it was suggested to that company that it do the same thing for conversion to 300 mm, the response was emphatically negative because of the huge cost of doing so. In fact, continued Dr. Spencer, it has been
OCR for page 80
80 NEW VISTAS IN T^NSAT~IC SCIENCE AND TECHNOLOGY COOPERATION estimated that the industry-wide cost of converting to 300-mm wafers is between $10 billion and $20 billion. This does not include the $2 billion cost for building a single fabrication facility. Dr. Spencer also remarked that the United States is currently dependent on either European or Japanese suppliers for the silicon wa- fers from which semiconductor devices are fabricated. National Technology Roadmap for Semiconductors An important part of the I300I cooperative process is the industry's main planning tool the National Technology Roadmap for Semiconductors (NTRS). The NTRS was first published in 1994 and has been updated every two years since. Dr. Spencer observed two recent important developments with respect to the roadmap: (1) it will become international in 1999 as engineers from all over the world will be asked to participate and (2) it will be updated on an ongoing basis; as technology changes, the roadmap will be revised, with revisions avail- able on-line (already, the NTRS is available on the World Wide Web at http:// www.sematech.org). The roadmap does not propose technical solutions but speci- fies capabilities that the industry must meet and the time by which they must be met, for historical productivity rates to continue. Issues in the Conversion to I300I Conditions in the world economy and the semiconductor industry have caused a delay in attaining 300-mm goals. The initial goal, which was widely publicized, called for a 300-mm pilot line to be completed by 1998, using 180-nm line widths on chips. However, the semiconductor industry has had a year of unexpectedly poor sales, largely because of the Asian economic crisis. Combined with the overcapacity among chip makers, there has been a glut of dynamic ran- dom access memories (DRAMs) and a drop in prices that has squeezed profits. Dr. Spencer noted that as of early June 1998 the price on the spot market for 16- megabyte DRAMS was $1, whereas the price for 64-megabyte DRAMS was $6; both levels are below production costs. Such tough economic conditions in the industry have caused a delay in investments to convert to 300 mm. The problems have created repercussions for semiconductor equipment manufacturers. The equipment makers were trying to meet the 1998 deadline and indeed were anxious to do so to have their equipment adopted widely in fabrica- tion facilities. Now, instead of a 1998 deadline for 300-mm wafers using 180-nm line widths, equipment manufacturers are being told that the deadline is 2001, with 150-nm technology instead. This makes investment planning very difficult for equipment makers, and further delays could negatively impact the finances of semiconductor equipment firms.
OCR for page 81
R&D IN THE FRAMEWORK OF THE NEW TRANSATLANTIC AGENDA The I300I Project 81 Members of I300I are Hyundai, Philips, STMicroelectronics, Siemens AG, and the Taiwan Semiconductor Manufacturing Corporation. Dr. Spencer said that the Asian financial crisis has prompted two Korean firms not to participate in I300I. The reason the U.S. semiconductor industry has sought international coop- eration, Dr. Spencer continued, is because the business is global. Any semicon- ductor company must be able to build a factory anywhere in the world and draw on equipment vendors worldwide. In fact, much of the equipment for 300-mm production will come from European and Japanese firms. If an equipment firm involved in I300I were to be prohibited from selling globally, it could not be a true leader in the industry, and it would not have the resources to invest in re- search and development. When I300I was being formed, every semiconductor company in the world was invited to participate, from Europe, to Japan, to Korea, Singapore, and Tai- wan. Japanese firms declined to participate, so there are two parallel I300I efforts in the world. The Japanese effort, called SELETE, includes 10 of Japan's largest semiconductor firms. Keys to Success Dr. Spencer listed a number of organizational and technical factors that will be crucial to I300I's success. Assignees Member companies send assignees either to I300I's headquarters in Austin, Texas, or to facilities of equipment manufacturers for tool development. Some of the best engineering talent resides at equipment manufacturers, and it makes sense to develop and debug equipment with the engineers of equipment manufacturers. This is distinct from the Japanese approach to 300-mm conversion, which has a more centralized structure. Standards At one time it looked as if there would be two sets of standards for 300-mm development. It was a major accomplishment of the I300I project, working with the Semiconductor Equipment Manufacturers Institute, to reach agreement with Japan on a single set of standards. Among the many standards to be settled are flatness, alignment, and impurities.
OCR for page 82
82 NEW VISTAS IN T^NSAT~IC SCIENCE AND TECHNOLOGY COOPERATION Automated Materials Handling Unprocessed wafers are expensive, about $1,000 to $1,200 each, and with integrated circuits on them they are even more valuable. Twenty-five 300-mm wafers with Intel processors etched on them are worth $1 million. It is therefore important to automate the handling of wafers to prevent costly breakage. Cost The I300I project must develop a 300-mm process technology that is initially no more than 20 percent more expensive than 200-mm technology. Otherwise, the conversion to 300 mm is not worth it for the industry. This cost target must be met across more than 200 pieces of equipment. Return on Investment Participating companies must receive a return on investment of about two to three times what they invest in I300I. Meeting that goal will require getting the best people from member companies as assignees. Dr. Spencer recalled that SEMATECH had problems in its early days in recruiting people from member companies to serve as assignees. That situation is now reversed, as SEMATECH usually has many more applicants than slots for assignees and, moreover, re- serves the right to send substandard assignees back to the member companies. Such practice will be carried forward in the I300I project. In conclusion, Dr. Spencer said that SEMATECH and I300I have shown that fierce competitors can come together to cooperate on precompetitive technolo- gies. It requires a commitment of resources, people, and time, but such coopera- tion also promises large payoffs. DISCUSSANTS John Shamaly Silicon Valley Group, Inc. Mr. Shamaly began his comments by observing that the I300I project is about productivity for the semiconductor industry. A simple calculation shows that, with the same throughput of wafers, switching from 200- to 300-mm wafers would increases productivity by 2.25 times. In conjunction with shrinking line widths, in accordance with Moore's law, productivity would grow even more. Finally, the Silicon Valley Group (SVG) sees improvements in step-and-scan technology as fueling additional productivity advances. Together, Mr. Shamaly said that these technological changes could increase industry productivity by 10 times in the next five to seven years. He observed that the demand for chips would have to
OCR for page 83
R&D IN THE FRAMEWORK OF THE NEW TRANSATLANTIC AGENDA 83 keep pace with these changes to support the sales of semiconductor manufactur- ing equipment, which in turn fund tool development. With respect to cooperation, Mr. Shamaly said that, as a manufacturer of automated wafer processing equipment, SVG serves a worldwide market and thus favors global cooperation when appropriate. There are guidelines that SVG follows in considering cooperation with other firms: · adherence to U.S. government regulations with respect to cooperation with foreign firms; · keeping its obligation to stockholders and employees, namely, to earn enough profits to maintain or increase employment at SVG, while stock- holders receive a good return on their investment; and maintaining competitiveness for SVG and the entire equipment industry. Increased cooperation, along with productivity increases in the semicon- ductor equipment sector, may limit the size of the world market. Much depends on the ability of worldwide demand for semiconductors to keep pace with productivity improvements. . In summary, Mr. Shamaly said that SVG favors cooperation in principle and recognizes its potential benefits. However, SVG is concerned about overcapacity in the industry and the negative economic consequences it has in the form of low chip prices. As with other productivity-enhancing efforts in the industry, Mr. Shamaly concluded, international cooperation must take place in the context of growing chip demand. Robert Hance Motorola Mr. Hance said that his remarks would focus on another collaborative effort in the industry, specifically one that is designed to address materials issues in semiconductor fabrication. Before discussing this, Mr. Hance noted that Motorola is a member of I300I, continues to participate in SEMATECH, and has benefited from its participation in both consortia. As the semiconductor industry switches to 300-mm wafers, Mr. Hance said, there is also the challenge of improving the materials that go into semiconductor fabrication. To address this challenge, the European Commuission has founded a consortium called the Hector 300-mm Project, which stands for High Epsilon Materials Cluster Tool for Optimized Rapid Deposition of stacked capacitors on 300-mm wafers. The Hector 300-mm Project is managed by Directorate General III (DG III) and related to the European Union's information technology research and development program, ESPRIT. In effect, this effort adds value to the 300- mm conversion project by making sure that appropriate materials are available by the time 300-mm manufacturing processes are on-line. Any time a change in
OCR for page 84
84 NEW VISTAS IN T^NSAT~IC SCIENCE AND TECHNOLOGY COOPERATION wafer size is made in the industry, Mr. Hance said, new materials must be devel- oped to mirror new manufacturing processes. Motorola is able to participate in this European consortium because of its manufacturing plants in Europe. The consortium began operation in March 1998, is expected to last 28 months, and will receive 10 million ecu, or $11 million (U.S.), from DO III. The members of the consortium are Aixtron, ASM Lithogra- phy, Steag-AST, Siemens, Motorola, Lucent, Austria Mikro Systeme, the Fraunhofer Institutes, and RC Juelich. In concluding, Mr. Hance stated that the European effort was in tackling 300-mm issues at the "module" level rather than the global level at which I300I operates. Both efforts are necessary for the industry to take full advantage of the switch to semiconductor fabrication on 300-mm wafers. Michael Borrus University of California at Berkeley Because the prior discussants, in addition to William Spencer, put so much content on the table in terms of international collaboration, Mr. Borrus said that he would concentrate on the context of international collaboration. Cross-border collaboration is not only useful for promoting innovation but increasingly essen- tial as the economy becomes more global. International collaboration is part of a nascent effort to globalize the innovation process. Mr. Borrus said that he would discuss three reasons for the growing globalization of the innovation process. The Growing Cost and Risk of Innovation Growing costs and risks were mentioned by Dr. Flamm and Dr. Spencer and are part of what Mr. Borrus termed the Business Week image of innovation. That is, the need for quick time to market within firms, narrow market windows, and costly research and development (R&D) creates incentives to look worldwide for partners to meet these challenges. In this view we have a borderless world in which multinational firms work together to push the frontiers of innovation. Increasing Specialization of Technical Skills A consequence of the globalization of innovation has been that regions main- tain and strengthen their hold on unique technical skills. Mr. Borrus recalled an anecdote regarding the sale by the American conglomerate ITT of its German subsidiary Standard Electric Lorenz (SEL) to the French company Alcatel. Mr. Borrus asked a German friend if there was any worry over the sale; his friend responded that "SEL was German when it was American and will be German when it is French." The story underscores how a company or a region's innova
OCR for page 85
R&D IN THE FRAMEWORK OF THE NEW TRANSATLANTIC AGENDA 85 live character is "sticky"; it is embedded in the people, the network of suppliers, and local institutions (such as universities) in a particular region. Mr. Borrus continued by identifying areas in which such "stickiness" is ap- parent. Precision engineering skills reside largely in Europe and Singapore; cer- tain materials technology thrives in Japan and Europe; software development and chip design are very strong in the United States. Such global exploitation of know- how results in certain places being the home for certain types of know-how. For this perspective on innovation to take hold, cross-border collaboration is essen- tial. . Mr. Borrus emphasized that in this image of innovation international col- laboration does not result in greater diffusion of technology but greater special- ization of skills in specific regions. This suggests that further globalization will increase specialization and product differentiation, not result in more homogeni- zation. If this is true, the tradeoff between global benefits of collaboration versus national economic advantage, raised earlier by Dr. Flamm, is a false one. We can, argued Mr. Borrus, increase the rate of innovation while reinforcing national ad- vantages in areas of specialization. Standard Setting and Local Reinvestment Mr. Borrus argued that the launch market for an innovative new product is typically local, and that if such a product sets an industry-wide standard, the local economy benefits. When a product becomes a standard, additional investment in a specific region results as production volumes grow and as a network of support- ing industries and suppliers grows. The products of companies such as Microsoft and Intel in the United States and SAP, Nokia, and Ericsson in Europe are ex- amples. With half of the world's economy in North America and Europe and with an even greater share of high-technology markets, U.S.-European cooperation is a way to ensure that standards for new high-technology goods are set by Europe and the United States. The economic benefits of standards setting will, said Mr. Borrus, then accrue to the United States and Europe. In summing up, Mr. Borrus reiterated the essential role that U.S.-European collaboration must play to commercialize products and reap the fruits of R&D and innovation. There is, however, political resistance to this idea in some quar- ters in the United States. He noted the opposition among some members of Con- gress to the participation of foreign companies in the extreme ultraviolet lithogra- phy consortium. If there are balanced contributions from both sides of the Atlantic in cooperative projects, Mr. Borrus said that such political problems can be avoided. In the end, U.S.-European cooperation constitutes a huge opportunity for both regions to capture substantial economic benefits in high-technology mar- kets.
OCR for page 86
86 NEW VISTAS IN T^NSAT~IC SCIENCE AND TECHNOLOGY COOPERATION COMMENTS FROM PANELISTS AND THE AUDIENCE William Spencer commented on John Shamaly's account of the Hector 300- mm Project by observing that there is a substantial amount of government sup- port for small semiconductor equipment companies outside the United States. He noted that such governmental financial support for small equipment companies does not exist in the United States, and this could eventually hurt the U.S. market share in this sector. Kenneth Flamm responded to Mr. Borrus's points about a tradeoff between global technological benefits and national economic advantage by agreeing that there were cases in which there was no such tradeoff. Dr. Flamm argued, how- ever, that it is an empirical question as to whether the tradeoff generally does or does not exist. Efforts to convert to 300-mm wafers are an example. Japan has decided to have an exclusive 300-mm conversion project, apparently hoping to reap the economic benefits in the manufacturing equipment sector. Semiconduc- tor device makers in the United States and Europe have agreed in I300I to move equipment development forward based on common interests and to compete in the market for devices. Dr. Flamm also raised a question about intellectual property. Even though Europe has accelerated its economic and political unification and even though the European Union (KU) has a common patent agency, Europe still does not have a single forum for litigation of patent disputes. Dr. Flamm solicited input from the audience as to whether his understanding was correct about the status of dealing with patent disputes in Europe. In response, Dr. Jorma Routti, Director General of DO XII of the European Commission, acknowledged a need for harmonization between the United States and Europe on patent law in the context of the World Intellectual Property Orga- nization. This, however, was outside the scope of the U.S.-EU science and tech- nology agreement. Dr. Routti expressed the hope that patent law would encour- age the publication and dissemination of scientific research and preserve the ethic ~ . . . A. . . Ot openness In scenic Inquiry. Dr. Routti also inquired about the balance of software and hardware develop- ment in the development of semiconductor devices such as integrated circuits. Specifically, he wondered whether there was a software or hardware tradeoff or a proper balance between the two that researchers or policymakers should explore. Dr. Spencer responded by describing the resource tradeoffs that semiconduc- tor makers face in developing products. They can put money into the develop- ment of competitive products or into manufacturing processes and tool develop- ment. Companies must, of course, fund some of both activities, but the purpose of projects such as SEMATECH and I300I is to allow companies to minimize funds for the latter. This is why process and tool development efforts are known as "precompetitive" projects. The result is relatively more funds available for chip design, which is the main area in which semiconductor companies compete. Such
OCR for page 87
R&D IN THE FRAMEWORK OF THE NEW TRANSATLANTIC AGENDA 87 design activity encompasses development of software and hardware applications for the chip, and individual companies will place different emphases on each element, depending on their design approaches. This, in fact, is very much a competitive and proprietary issue for semiconductor companies. Alan Tonelson of the U.S. Business and Industrial Council Educational Foun- dation asked Mr. Borrus about the tradeoff between global technology dissemina- tion and national economic advantage. Mr. Tonelson wondered in particular whether this work contradicted earlier work of Mr. Borrus and his colleagues at the Berkeley Roundtable, which argued that there has been a convergence of manufacturing capabilities in Asia. Mr. Tonelson suggested that there has been an "electronics traffic jam" in Asia whereby too many countries have been mak- ing too many of the same electronic goods such as DRAM chips. Such overca- pacity in electronics manufacturing, said Mr. Tonelson, may be at the root of the · · ^- . . . Aslan ~lnancla. . cnsls. Mr. Borrus responded by noting that, whereas collapsing DRAM prices may have contributed to financial problems in Korea, it would be very difficult to assign a causal link between falling DRAM prices and the general economic crisis in Asia. Moreover, Asia does have a well-defined division of labor that supports the notion that a homogenization of technical tasks is not occurring. Mr. Borrus cited software development in Bangalore, process know-how in Singapore, and digital design in Taiwan as examples of Asia's division of labor.
Representative terms from entire chapter: