After two decades of false starts and of being stifled by international technology-transfer constraints, China's microelectronics industry turned a corner in 2000. In that year, the Chinese government issued a series of pragmatic policies to promote the development of the country's microelectronics industry. These policies are closely tied to information technology (IT) development in China and provide a bridge between these two critical technology areas. Adding to this policy momentum in 2000, foreign investors announced or began work on several large microelectronics projects in China worth a total of $7 billion.

Joining the world market

China has always regarded the microelectronics industry, especially the integrated circuit (IC) segment, as strategically important. The government listed microelectronics as a priority when it formally initiated the "four modernizations" program in the late 1970s. Nevertheless, it has been the weakest link in China's electronics industry. From 1980 to 1999, according to China Electronics News (Zhongguo Dianzi Bao), a total of ¥ 25.7 billion ($3.11 billion) was invested in the country's IC industry. Of this amount, foreign investors provided ¥12.6 billion ($1.52 billion). These levels lagged woefully behind investments made in Taiwan and South Korea over the same period. Today, officials from within the Chinese electronics industry recognize that this lack of investment — more than international export-control restrictions — is the key factor responsible for the slow modernization and growth of the sector. Accordingly, some key PRC officials have come to believe that foreign-invested ventures are the quickest way to improve the IC industry.

Indeed, some of the world's top foreign chipmakers have already committed to the China market. Advanced Micro Devices, Inc. (AMD) announced in January 2000 that it would set up its first wholly foreign-owned enterprise (WFOE) in Suzhou, Jiangsu Province. Total investment in the project is about $108 million, with annual production of 50 million units of IC products. Also in 2000, Motorola Inc. announced it would invest ¥16 billion ($1.9 billion) in Tianjin to build a new IC manufacturing center, including a chip plant and a semiconductor packaging and measurement factory, while Philips Semiconductors launched its first WFOE in Dongguan, Guangdong Province, a semiconductor components measurement and packaging facility.

Other large investments were also started in 2000 — in November, Shanghai Hongli Semiconductor Manufacturing Co. kicked off construction on a facility slated to produce 8- to 12-inch wafers (see Glossary). Beijing Xunchuang IC Co., a joint venture with investment from Beijing Electronics Holding Co., Beijing Oriental Electronics Group, Beijing Kingstron Electronics Factory, and Asia Pacific Technology Development Corp., began construction of an IC production line with total investment of $200 million in December. Also that month, Beijing Huaxia Semiconductor Manufacturing Co., Ltd. announced it would invest in an 8-inch, 0.25-micron production line. All of these projects reflect the growing sophistication and global integration of the Chinese electronics industry. The expanded foreign presence also reflects the changing dynamics of competition and cooperation in the global industry.

The so-called "Taiwan factor" is also playing a role in the transformation occurring in China's semiconductor industry. As Taiwan's economy experiences its own restructuring, a large number of key Taiwan-based microelectronics firms have decided to invest in the mainland. One of the most prominent projects is Semiconductor Manufacturing International Corp.'s contract manufacturing facility in Shanghai, which has the capacity to produce 480,000 8-inch wafers at 0.25 microns. Another is Shanghai Grace Semiconductor Manufacturing Corp.'s high-profile $1.6 billion wafer foundry joint venture involving Winston Wong, son of Wong Yunqing, former CEO of Formosa Plastics, and Jiang Mianheng, son of PRC President Jiang Zemin. Morris Chang, the high-technology guru who turned Taiwan Semiconductor Manufacturing Corp. into one of the world's most successful foundry operations, is reportedly seriously considering establishing a facility on the mainland. These key players from Taiwan bring both extensive expertise and huge capital resources to the table.

A long-term strategy

The output of the global semiconductor industry topped $200 billion in 2000, of which nearly 90 percent was from the IC sector. China's IC industry grew 42 percent in 2000, producing 5.88 billion units; total sales reached ¥20 billion ($2.4 billion). According to Cahners In-stat Group, a digital communications research firm, China has the potential to become the third-largest semiconductor producer after the United States and Japan by 2003 and the second-largest IC and semiconductor market after the United States by 2005. By 2005, China hopes to increase IC production to 20 billion units and attain sales revenues of ¥80 billion ($9.7 billion), or 2 percent of the global market. Such production will meet about 30 percent of domestic demand. By 2010, China anticipates its IC production will reach 50 billion units, with sales revenues of ¥200 billion ($24.2 billion). At current growth levels, this will constitute roughly 5 percent of the 2010 global market and will meet 50 percent of domestic needs.

For national security reasons, domestic IC enterprises will likely provide key IC products for national defense and related industries such as space and satellites. In fact, some government officials reportedly believe that China should rely more on domestic than foreign investment in developing the sector.

China will require huge amounts of investment, however, to realize these ambitious plans. According to the US-based Semiconductor Industry Association, it takes at least $200 million to build a 6-inch line; $1.2 billion to build an 8-inch line; and $2.5 billion to build a more advanced 12-inch line.

During the Ninth Five-Year Plan (FYP, 1996-2000) period, China invested over ¥10 billion ($1.2 billion) in the 909 Project, the Chinese government's most concerted initiative to date to develop an advanced semiconductor industry in China. The centerpiece of the 909 Project is Shanghai Huahong NEC Electronics Co., formed in October 1996 to design and produce memory and logic semiconductors using 0.35-0.5 micron technology. Because of the continued restrictions on technology transfer posed by the United States and several other countries, the PRC government picked Japan's NEC Corp. to form the joint venture. The project was finished in late 2000 and although there have been reports that the project is facing financial difficulties, it has significantly upgraded the technological levels of the industry in China. China still imports over 70 percent of the IC products it needs, however.

Imports and exports

Imports and exports of Chinese microelectronic products have developed quickly over the past five years. Most Chinese exports are lower-end components involving basic processing and manufacturing techniques. Imports, in general, are much more sophisticated. Imports of hybrid IC products grew at a compound annual rate of 91.7 percent from 1995 to 2000, while the level of exports over the same period rose by 59.9 percent yearly.

Because Chinese microelectronics exports are generally low value-added components, such as transistors, capacitors, and resistors, they mainly compete on price. Export volumes are therefore sensitive to price changes in the global market. The Asian financial crisis in 1997 caused significant currency devaluation in many Asian economies, which resulted in falling prices for microelectronic products and the subsequent sudden drop in Chinese hybrid IC exports in 1997-98.

In 2000, Chinese IC exports reached 4.67 billion units, up 34.8 percent over 1999, and total IC exports were valued at $2.1 billion, a 30 percent increase over 1999. State-owned enterprises (SOEs) accounted for $112 million, roughly 5.3 percent, and foreign-invested enterprises (FIEs) for $1.99 billion, about 94 percent, of total exports. Only $66 million fell under the normal, direct-trade category, while $938 million, or 44 percent, came under export-processing operations. About $1 billion, or 48 percent of units, were made mainly from imported materials.

A strategic model of competitiveness

The "diamond model," created in the early 1980s by Michael Porter, a Harvard Business School professor and co-founder of Monitor Group, can be used to determine the current competitiveness of the Chinese microelectronics industry and to suggest possible steps to enhance this competitiveness (see diagram). According to Porter's model, every nation or region must create an environment that enables and inspires companies to build competitive assets. The model points out four broad attributes or determinants that explain productivity and performance:

• Factor conditions
  Advanced and specialized human resources, technical infrastructure, and other factors of production needed in the industry and the ability to upgrade them continuously.
• Firm strategy and rivalry
  Capable, committed, and fiercely competing local rivals and a government that ensures competition and new investment.
• Demand conditions
  A sophisticated local customer base expecting the latest innovations and the highest quality standards. Local demand should anticipate global trends so that producers can learn, adjust, and ultimately upgrade to compete for the world's most sophisticated customers.
• Related and supporting businesses
  A group of strong local suppliers and distributors that can contribute to the innovation process. Related businesses can reinforce skills using advanced-process technologies or strategic marketing channels.

These four determinants are self-reinforcing: the effectiveness of the whole diamond depends on how the determinants work as a system, with government and chance influencing the outcome. Using this model, a firm can understand the dynamics in a specific industry. Sustained success results from innovation and improvement within this system. Improving one determinant can help to improve the entire system.

Factor conditions

China has no shortage of knowledge about the global microelectronics industry. Most PRC officials involved in this industry can provide an excellent discourse on current trends in both the business and technology sides of the sector. What the industry lacks, however, is enough qualified people to run it. According to one government estimate, the United States has roughly 400,000 IC designers, while China has fewer than 4,000. By 2010, China will require roughly 300,000 IC design engineers and specialists to support the industry.

Three major reasons account for the insufficient numbers: first, the microelectronics industry is international, and multinational companies recruit talented people worldwide. Because of the significant differences in compensation packages and the complex personnel system in China, many Chinese microelectronics experts have chosen to work abroad. After training at top research universities, such as the Massachusetts Institute of Technology or Stanford University, Chinese engineers are lured by exciting opportunities to work in Silicon Valley and other attractive destinations.

Second, Chinese universities are good at training students with strong technical skills, but are weak at teaching production and management skills required by the modern microelectronics industry. Third, China lacks managers qualified to handle the business and financial planning activities needed to drive the industry beyond its modest beginnings to world-class levels (see The CBR (members only), May-June 2001).

Strategy, structure, and rivalry analysis

At present, China has seven core IC production enterprises, a dozen important packaging factories, several dozen design companies, and several manufacturers of key materials and equipment. Last year, total IC production topped 5 billion units and sales revenue topped ¥15 billion ($1.8 billion). The Chinese IC industry has been growing more than 40 percent annually for 10 years, with the most significant improvements in product quality and sophistication occurring over the last two years. (Interestingly, this was nearly five times GDP growth over the same period.)

Initially, all Chinese IC enterprises tried to be self-reliant. In the jargon of the industry, they followed the integrated device manufacturer format, that is, one single factory performed the entire set of processes including design, manufacturing, measuring, and packaging. Examples of this type of enterprise include Wuxi Huajing Semiconductor Microelectronics Co., Ltd.; Huayue Microelectronics Co., Ltd.; and Shougang NEC Electronics Co., Ltd. Shanghai Beiling Corp. Ltd. does its own design, manufacturing, and measurement, but outsources packaging. A foundry line begun at Shanghai Philips Semiconductor Co. (now Shanghai Xianjin Semiconductor Manufacturing Co., Ltd.) only undertakes manufacturing and some measuring tasks and does not have design and packaging capacity. The first true foundry line was China Huajing Electronic Group Corp.'s metal-oxide semiconductor production line, which obtained technology from both Toshiba Corp. and Siemens AG and accepts orders from both domestic and overseas design and manufacturing companies. Shanghai Beiling was the first publicly listed company in the Chinese microelectronics industry, listing on the Shanghai Stock Exchange in October 1998.

At the end of 2000, China had 25 major IC production lines, but only two that produced more than 20,000 chips per month: Shanghai Beiling's 4-inch line and Shanghai Xianjin's 5-inch bipolar line. Worldwide, there are 246 4-inch lines, 164 5-inch lines, 287 6-inch lines, 252 8-inch lines, and a few dozen 12-inch lines. Most are 8-inch lines running at 0.18-0.25 microns. Table 1 shows the major Chinese IC production lines.

Competition in China's IC industry is very much driven by regional rivalries. The government clearly aims to have several key "national champions" that are flag-bearers for the Chinese microelectronics and IT industries in the global marketplace. Though several potential candidates are now on the scene, it is probably too soon to know which firms will become future leaders. One of the major issues China faces as it seeks to stimulate development in this sector is whether it can accept the outcomes of the global marketplace. The growing presence of key foreign players in the industry throughout the value chain suggests that China will only become further enmeshed in the competitive dynamics of the international microelectronics sector and that domestic competition for scarce resources will increase as Shanghai, Beijing, and other cities seek to enhance their attractiveness as sites for design, processing, and manufacture of key components and final products. So far, Shanghai appears to be taking the lead in the race to become the country's first serious microelectronics hub, with much of the new activity occurring in Shanghai's Zhangjiang High-Technology Park.

Domestic demand

Demand in China for IC products promises to keep growing as the economy develops (see Table 2). China ranks number one worldwide in the production of televisions, video cameras, VCD players, telephones, electronic watches, calculators, refrigerators, and air conditioners, among other electronics. China also is a major producer of switches, personal computers, mobile phones, disk drives, monitors, and computer boards. ICs are used in all of these products.

China's rapidly developing telecommunications industry, in particular, promises to be a significant source of demand for IC products in the near future. According to China Electronics News, the number of Chinese fixed-line telephone users reached 167 million by mid-2001 and is expected to grow 10 percent each year, while the number of mobile phone users, which is put at more than 120 million, will grow more than 20 percent each year. Internet usage growth, which is slowing down slightly, will nevertheless likely expand 40-50 percent each year for the next three to five years. And China will continue to restructure and upgrade traditional industries.

Another important area of IC demand will be smart cards, which include public telephone IC cards, mobile phone subscriber-identity-module (SIM) cards, and other IC cards used in banking and utilities payment. More than 230 million smart cards were sold in China in 2000, including 120 million phone cards and 42 million SIM cards. Demand will probably exceed 300 million in 2001. Domestic firms are attempting to enhance their design and production capabilities in the face of increasingly intense competition from foreign vendors such as Philips, Gemplus SA, Motorola, and ST Microelectronics. Shanghai Hua Hong Group, for example, sees the smart card as one of the main drivers behind its entry into design and processing, according to China Electronics News. The PRC government has certified more than 15 foreign suppliers via the State IC Card Registration Center.

Cluster analysis

In Porter's "diamond model," the cluster factor refers to the presence or absence in a nation of internationally competitive supplier and related industries. Clearly, the existence or absence of the full complement of microelectronics-related IC activities such as design, packaging, equipment, and component manufacturing is relevant in this regard.

For China, the most controversial aspect of the recent expansion of the country's microelectronics industry is the plethora of foreign companies that have started to invest in the supplier, vendor, and service networks of many of the Global Fortune 500-size microelectronics and telecommunications firms. Specialized companies from the United States and elsewhere that hope to generate business by supporting foreign and domestic customers onsite are filling many of the gaps in China's own domestic industry. One good example is Trident Microsystems Inc., which designs ICs for consumer electronics and multimedia chips for computers. Its Shanghai-based design unit now directly serves key Chinese customers such as Konka Group Co., Ltd.; Changhong Electric Co., Ltd.; and TCL Holdings Co., Ltd.

• IC design
  Chinese officials are well aware that most IC products used in consumer electronics are designed and produced abroad. Aside from the largest firms, most Chinese design companies are not doing well financially because they lack equipment, personnel, and technical expertise and cannot meet customer requirements. Chinese companies have a total annual design capacity of only about 300 types. The highest design capability is 0.25 microns, and mainstream capability is 0.8-1.5 microns. Despite these advances, Chinese IC design companies as a whole still lag significantly behind international standards.
  
  The first important IC design company in China, Huada IC Design Center (formerly Beijing IC Design Center), was formed in 1986. By 1998, fewer than 60 IC design companies had set up operations in China. Today, China has more than 120 IC design companies, centers, and institutes, with a total revenue of more than ¥1 billion ($121 million) last year. These new firms include WFOEs, joint ventures, private companies, and SOEs. In 2000, only four domestic design companies boasted sales revenue of more than ¥100 million ($12.1 million): Hangzhou Shilan (a private company), Beijing Huada (an SOE), Wuxi Semico Microelectronics (a shareholding company), and Datang Microelectronics (a listed company under Datang Telecom).
  
  In the 1990s, after Wuxi Shanghua, Shougang NEC, Shanghai Xianjin, Shanghai Beiling, and others began accepting production orders from non-affiliated companies, Chinese design companies developed rapidly and can now be classified into five categories: pure design companies, such as China Huada IC Design Center; design departments of IC enterprises or research institutes, such as Huajing and Shougang NEC; design departments of electronics or communication equipment manufacturers, such as Huawei Technologies, Shanghai Haixin Group, and Nanjing Panda Electronics; design departments of academic institutes, such as Qinghua University, Beijing University, and Fudan University; and foreign-invested IC design companies, such as Intel Corp., Motorola, and Seiko Epson Corp. The most recent addition to this group is Integrated Silicon Solution (ISSI), which has confirmed it will set up a $10.5 million design and R&D center in Shanghai's Zhangjiang High-Tech Park. In February 2000, the Ministry of Science and Technology established China's first formal industrial base for IC design in Shanghai. With investment from the New Huangpu Group in Shanghai, the base hopes to attract a broad array of design firms so that China can expand its ability to contribute to this section of the semiconductor value chain. By mid-2000, 16 IC design institutes had already moved into the park; some reports suggest as many as 50 design firms are already operating there.
  
  
• Packaging
  The sales revenues of Chinese IC packaging enterprises exceeded ¥13 billion ($1.6 billion) last year, with 14 enterprises registering sales revenues of more than ¥100 million ($12.1 million). Roughly 4.5 billion ICs were packaged last year, but only five enterprises packaged at least 100 million ICs each.
  
  In the Chinese IC packaging industry, domestic companies coexist with foreign ones, and advanced technologies coexist with outdated ones. Currently, China has about 60 major packaging companies and more than 100 small companies. A dozen or so are WFOEs, which generally utilize the latest technology and equipment. These WFOEs target the high end of the market, and include AMD (Suzhou), Dongguan Philips Semiconductor, Intel (Shanghai), Jinpeng Packaging (Shanghai), Matsushita (Suzhou), Motorola (Tianjin), Samsung Electronics Semiconductor (Suzhou), Sanyo Semiconductor (Shekou), Suzhou Shuangyue, and Yiheng Technology (Wuxi).
  
  Some joint ventures have technology from the mid-1990s. These companies target the middle of the market and include Beijing Mitsubishi Stone Integrated Circuits Co. Ltd., Changsha Qunli Integrated Circuit Co. Ltd., Leshan Phoenix Semiconductor Co. Ltd., Nantong Fujitsu Microelectronics Co. Ltd., Shanghai Alphatec Semiconductor Packaging Co. Ltd., Shanghai Kaihong Electronics Co. Ltd., Shanghai Matsushita Semiconductor Co. Ltd., Shanghai Xinkang Semiconductor, Shenzhen SGS-Thomson, Shougang NEC Electronics Co. Ltd., and Wuxi Huazhi Semiconductor Co. Ltd.
  
  WFOEs and joint ventures dominate the high-end product markets. The arrival of firms such as Amkor Technology, Inc. and ChipPac Inc. in Shanghai demonstrate the upgrading that is taking place in the packaging sector of the microelectronics industry. ChipPac, working with Qualcomm Inc., will operate a $25 million facility to bring ball-grid array and other advanced packaging technologies to China. Amkor, the world's largest merchant supplier of chip-packaging and testing services, will focus on providing packaged ICs for the PRC personal computer, laptop computer, and mobile phone markets. The chip packages will contain static random access memory and flash memories, power controllers, radio frequency power amplifiers, and graphic chip sets.
  
  More than 40 domestic enterprises produce on a small scale. As their technology level is comparable to the international standard of the 1980s, these enterprises target the low end of the market (see Tables 3, 4). Domestic packaging companies account for less than half of the chip packaging market and will continue to do so for the next two to three years. Even given predictions that the size of the overall Chinese semiconductor packaging market will double by 2002, FIEs will likely continue to dominate this aspect of the industry. Faced with fierce competition brought in by these multinational players, many Chinese microelectronics packaging enterprises are restructuring to enhance their competitiveness.

• Microelectronics industry equipment
  China is still incapable of producing most of the equipment used in an 8-inch IC production line, though it can produce some supplementary machinery. As for 6-inch lines, Chinese firms are technically capable of producing almost all required equipment but no one firm is manufacturing enough to be considered a world-class producer.
  
  In the past, China was not a major target market for the top equipment suppliers because of the low volume of chip production inside China as well as export-control restrictions. Lately, however, equipment producers have begun to pay much closer attention to the Chinese IC industry because of its growth potential and because many of the previous restrictions on technology transfer have been relaxed over the last two or three years, allowing 8-inch, 0.25 micron technology to come to China. These factors help to explain the design of Applied Materials, Inc., the world's largest semiconductor equipment and service supplier, to establish a new regional headquarters in China. According to the company's general manager, Applied Materials expects its sales to exceed $1 billion in three to five years.
  
  Since 2000, the construction of several new IC production lines has started, and several more are being contemplated or are already under contract. According to official planning documents, Beijing and Shanghai will construct a combined 15-20 production lines for wafers of 8 or 12 inches. The total investment required for these production lines will reach $20 billion. Of this huge investment, 70 to 80 percent will be used to purchase advanced equipment, most of which is unavailable in China.
  
  Industry experts have assessed the market demand for semiconductor equipment in China at roughly $4 billion. This is a significant jump over previous years, when the number and size of planned projects was much smaller. In 2003, demand will likely increase to $7 billion, and by 2010 China could be one of the biggest markets in the world for semiconductor and related equipment, with well over $10 billion in sales.
  
  In the meantime, China will have to import most of the equipment it needs. As noted, Chinese microelectronics equipment manufacturers produce on a very small scale. Even the best company had sales revenue of only ¥276 million ($33.4 million) in 2000. The total sales revenue of domestic manufacturers of semiconductor equipment in 2000 was a modest ¥170 million ($20.5 million). These sales met less than 10 percent of domestic demand. Domestic manufacturers provided only 10 percent of the equipment for technology improvement projects with government subsidies approved by SETC in 1999 and 2000.
  
  
• Electronics components
  Low technology levels, high unit production costs, and small-scale production have long characterized China's electronics components industry, which had total sales revenues of about ¥95.4 billion ($11.5 billion) in 2000. Production processes at China's 6,093 electronics components companies employ 1.3 million people and are still dependent on outdated equipment and technology. Advanced management systems are generally absent, and the ability to develop new products is limited. For example, 95 percent of Chinese light-emitting diode (LED) manufacturers are actually performing packaging work and must import the core chips. Because of the constraints of technology and equipment, 90 percent of the materials used for the production of 0.35-0.5 micron chips must be imported. Last year, Shanghai Feile Co. Ltd. ranked number one in terms of sales revenue at only ¥3.2 billion ($384.3 million).

The top 100 electronics components firms include all five different types of ownership structures, with shareholding enterprises and FIEs in the majority. The mean sales revenue of these 100 firms is ¥437 million ($52.8 million). For comparison, Intel's sales revenue in 2000 was $26.8 billion, according to a survey by Gartner Inc., a Connecticut-based research and consulting firm, and the top 20 semiconductor companies in the world had sales revenues of over $2 billion each. The remaining 6,000 or so companies had average sales revenues of only ¥8.6 million ($1 million) well below world-class levels. The average gross profit margin of the top 100 was about 14 percent, but SOEs and collective enterprises had only 9.9 and 10.6 percent gross profit margins, respectively. This is probably due to limited economies of scale and poor management. In contrast, FIEs, which account for more than 50 percent of all PRC electronics exports, excel because they have advanced technology, relationships with overseas parent companies, and an export-processing orientation. Finally, because the companies on this list include only the best performers each year, the operating results for the industry as a whole are likely to be much worse.

The impact of WTO entry

The Electronics Industry Alliance calls China "the most promising emerging market" for the US electronics industry. In 2000, US electronics exports to China totaled $4.2 billion, a 43 percent jump over 1999. US companies also increasingly rely on China as a site for sourcing electronics parts and components that are not produced in the United States; the electronics trade deficit reached $22 billion in 2000.

The entry of China into the World Trade Organization (WTO) has made US companies optimistic that they will have opportunities not only to export more but also to invest directly in some of the areas of most rapid Chinese economic growth. In particular, when the United States and China concluded their bilateral agreement for eventual WTO membership in November 1999, China agreed to six requirements that the American semiconductor industry had presented as top priorities: elimination of tariffs by 2002 on semiconductor imports (currently 6-10 percent); trading and distribution rights for foreign firms; de-politicization of buying decisions by SOEs; removal of technology-transfer and export requirements as a condition for foreign investment or market access; adequate protection of intellectual property rights; and continued permission for the United States to use antidumping methodology in trade disputes with China.

Foreign investors should be cautious in their optimism, however, as central government players' opinions about how best to develop China's semiconductor industry seem to be diverging. Though some still see technology import and foreign investment as the best way to promote the industry, others advocate national self-reliance and domestic development of technology.