SIC 3674
SEMICONDUCTORS AND RELATED DEVICES



This category covers establishments primarily engaged in manufacturing semiconductors and related solid-state devices. Important products of this industry are semiconductor diodes and stacks, including rectifiers, integrated microcircuits (semiconductor networks), transistors, solar cells, and light sensing and emitting semiconductor (solid-state) devices.

NAICS Code(s)

334413 (Semiconductor and Related Device Manufacturing)

Industry Snapshot

No longer wholly dependent on personal computer sales, the U.S. semiconductor industry in the early 2000s provided components for a wide range of consumer and industrial electronics. Computers still accounted for more than 50 percent of the industry's sales. However, the growing field of data and telecommunications accounted for more than 25 percent. The industry was notoriously cyclical. After worldwide semiconductor sales grew 40 percent in 1995 to nearly $150.0 billion, sales dropped in 1996 and were flat through 1998. Sales reached record levels in 2000, totaling $204 billion. However, they fell sharply the following year, dropping to $139 billion as the industry experienced the worst year in its history. In the wake of 2001's dramatic decline, tens of thousands of industry workers lost their jobs.

The semiconductor industry was one of the fastest growing sectors in the U.S. economy between 1987 and 1996, when it grew from the seventeenth largest industry in the United States to the largest as measured by its contribution to the U.S. gross domestic product. Entering the twenty-first century, the information technology sector accounted for 11 percent of the U.S. gross domestic product and one-fourth of U.S. manufacturing output. According to the Semiconductor Industry Association (SIA), in 2001 sales from U.S. semiconductor manufacturers totaled $72 billion, representing 51 percent of the global market. That year, the industry employed 283,875 U.S. workers, invested $18 billion in capital, and devoted more than $13 billion to research and development.

Organization and Structure

Sometimes referred to as "the crude oil of the information age," semiconductors are a pervasive but generally unseen aspect of everyday life. The tiny electronic circuits etched on chips of silicon are critical to the operation of virtually all electronics, from automatic coffee makers and antilock braking systems to cellular phones and supercomputers.

The computer industry is by far the largest market for semiconductors. In the early 2000s, sales to computer manufacturers and related enterprises accounted for about 50 percent of overall U.S. sales of semiconductors. Consumer electronics and the automotive industry were also important users of semiconductors and related products. The fastest growing market for semiconductors was related to communications, which accounted for more than 25 percent of sales in 2001, reflecting the rise of a global networked economy that relied on the electronic transfer of data.

Semiconductor chips are manufactured in "clean rooms," free of contaminating dust. In those facilities, thin, round silicon wafers are processed in batches. Chipmakers buy polished blank wafers from companies that specialize in growing silicon crystals, from which the wafers are cut. Each wafer is about half a millimeter thick. Microelectronics circuits are built up on the wafer layer by layer.

Circuit patterns—the collection of transistors, capacitors, and associated components and their interconnections—are inscribed on large glass plates called photomasks. The photomasks are later reduced and photolithographically projected onto the silicon wafers. Each mask comprises a total integrated circuit design.

Semiconductor companies design and manufacture primarily two types of products: integrated circuits (ICs) and discrete devices. A discrete semiconductor is an individual circuit that performs a single function affecting the flow of electrical current. For example, a transistor, one of the most common types of discrete devices, amplifies electrical signals; rectifiers and diodes generally convert alternating current into direct current; capacitors block the flow of alternating current at controlled levels; and resistors limit current flow and divide or drop current.

Integrated Circuits. Also called chips, integrated circuits are a collection of microminiaturized electronic components, such as transistors and capacitors, placed on a tiny rectangle of silicon. A single integrated circuit can perform the functions of thousands of discrete transistors, diodes, capacitors, and resistors. There are three basic types of integrated circuits currently produced by American semiconductor manufacturers: memory components, which are used to store data or computer programs; logic devices, which perform such operations as mathematical calculations; and components that combine the two. This latter category of integrated circuit is the most sophisticated and includes microprocessors, the computer "brain" that manipulates a wide range of data, and micro controllers, which perform repetitive tasks.

The two largest selling types of memory integrated circuits are DRAMs and SRAMs. A DRAM (dynamic random access memories; pronounced DEE-ram) stores digital information and provides high-speed storage and retrieval of data. It is called a "dynamic" circuit because the data is stored in a temporary medium that allows it to fade, and so must be constantly refreshed electronically.

SRAMs (static random access memories; pronounced ESS-rams) perform many of the same functions as DRAMs, but at higher speeds. Unlike DRAMs, they do not require constant electronic refreshing, hence the term "static." They also contain more electronic circuitry and are more expensive to produce than DRAMs.

Both of these integrated circuit products are manufactured in large quantities and so are considered to be "process drivers." That is, the manufacturing processes used to produce them are constantly being refined, and those refinements often affect manufacturing processes of other products.

Two other important semiconductor memory products are EPROMs (erasable programmable read-only memories) and EEPROMs (electrically erasable read-only memories). EPROMs are used to store computer programs. Unlike older read-only memories (ROMs) that carried fixed programs, EPROMs are programmed by the customer. EEPROMs are easier and faster to update than EPROMs because they are programmed using electricity. While EPROMs are usually programmed only once, EEPROMs can be reprogrammed without removing them from their applications, so they can be updated virtually anytime.

ASICS. Most logic semiconductors are now customized products tailored to the specific needs of each customer. In fact, ASICS (application-specific integrated circuits) have become the most commonly manufactured nonmicrocomponent logic semiconductors.

There are four basic classes of ASICs; each class has a different degree of customization of the chip. Full-custom ASICs are designed from scratch; standard cells are designed by combining modular cells from a cell library; semi-custom chips are customized in only one or two areas; and programmable logic devices are programmed by blowing fuses in a device to alter the logic function. Because of high design costs and the often-limited quantities produced, ASICs tend to be more expensive than integrated circuits built from off-the-shelf components. But because they combine several specialized functions on a single chip, they offer some important advantages: they are smaller, simpler, and fewer of them are needed; they allow for a greater degree of integration, which leads to more efficient use of circuitry; and, since they contain less circuitry, fewer interconnections are needed and overall performance is enhanced.

Microprocessors and Controllers. Microprocessors (MPUs) are the central processing units in all microcomputer based systems. These products perform a variety of tasks by manipulating data within a system and controlling input, output, peripherals, and memory devices.

The two major types of MPUs are CISCs (complex instruction set computing) and RISCs (reduced instruction set computing). Though CISCs used to be the basis for all MPU operations, RISCs became increasingly popular in the 1990s because of their faster operating speeds, their ability to run more sophisticated software, and their ability to deliver better graphics. MPUs are used in local area networks (linked personal computers and workstations; called LANs) and satellites. The latest generation of these circuits operates at speeds of from 40 to 50 million cycles per second.

Microcontrollers (MCUs), which combine a micro-processor, memory circuits, and input/output circuitry, are used as embedded controllers in virtually every electronic product. They perform such repetitive tasks as controlling the antilock brake systems in automobiles.

Background and Development

Semiconductors were invented in the United States in the late 1950s, but the invention that truly began the electronics revolution appeared nearly 50 years earlier. The three-element vacuum tube was invented by Lee de Forest in Palo Alto, California, in 1906. Called the audion, the tube was used as a sound amplifier and generator of electromagnetic waves; its invention laid the foundation for the development of radio, television, radar, computers, and many other groundbreaking electronic devices. These early tubes, however, were bulky and fragile. For example, ENIAC (Electronic Numerical Integrator and Computer), the world's first large electronic computer, ran on 18,000 vacuum tubes and was the size of a house.

The tubes also played a vital role in the development of early telephone communications networks. But as those networks expanded across the United States, the unreliability of the tubes became intolerable. Consequently, the main push for a replacement for the vacuum tube came from researchers at AT&T Bell Laboratories in New Jersey.

For a number of years, the company had been studying potential uses of solid materials that were poor conductors of electricity, primarily silicon and germanium. Silicon, one of the world's most plentiful elements, is found in the earth's crust as silica and silicate and is the principal component of sand, quartz, and glass. In its pure form, silicon is a very poor conductor, but Bell Lab researchers found that it could be treated, or "doped," with other materials to act as a conductor under some conditions and as an insulator under others.

These new "semiconductors" allowed for the development in 1947 of the transistor, which marked the beginning of the age of solid-state electronics. In 1956, William Shockley, John Bardeen, and Walter H. Brattain—the Bell Labs research team responsible for the development and refinement of the transistor—received the Nobel Prize for their invention. The same year he was awarded the Nobel Prize, Shockley returned to his boyhood home of Palo Alto, California, and established his own semiconductor manufacturing operation. To staff his new company, Shockley recruited many of the country's brightest young scientists and engineers.

Disagreements eventually led seven of Shockley's recruits to set out on their own. The company they founded, Fairchild Semiconductor, would become "the mother of semiconductor companies." According to the Semiconductor Industry Association, more than 23 semiconductor and related enterprises can trace their origins back to Fairchild. Among them were such important and well-known companies as Intel, Advanced Micro Devices, and National Semiconductor.

Probably the most important technological development to come out of Fairchild was the integrated circuit or "chip." The head of Fairchild, MIT graduate RobertN. Noyce and Texas Instruments researcher Jack Kilby are credited with inventing the integrated circuit almost simultaneously in 1958. The original Texas Instruments version of the chip required the soldering of tiny gold wires on the outside to connect the components. The Fairchild version, on the other hand, relied on a thin layer of metal conducting film, which was sprayed onto the chip like paint. Roadways were then cut by lithography into this metallic layer to create the desired pattern of connections between elements of the circuit. This version of the chip was more readily manufacturable, and Fairchild soon emerged as the early leader of the semiconductor industry.

Noyce left Fairchild in 1968, along with Gordon E. Moore, a respected physical chemist. Together, they formed Intel Corporation and set out to manufacture a computer memory chip. Intel eventually came to dominate the industry as the undisputed leader in semiconductor technology. In addition to the first memory chips, Intel was responsible for pioneering the development of the microprocessor, the so-called "computer-on-a-chip."

U.S. manufacturers continued to dominate the semiconductor industry until the 1980s, when foreign industrial targeting and illegal dumping practices combined to erode U.S. worldwide market share. This "blood bath," as it was referred to in industry publications at the time, drove Intel, Motorola, National Semiconductor, Advanced Micro Devices, and Mostek out of the dynamic random access memory (DRAM) market altogether. Japanese manufacturers, however, who utilized investment cost advantages to conquer the DRAM market saw that market plunge at the onset of the 1990s.

Consequently, U.S. semiconductor manufacturers began to refocus their efforts on proprietary products during the early 1990s, capitalizing on their well-known strengths in design and innovation and moving away from commodity products. According to industry observers, two Congressional actions were instrumental in paving the way for this development.

The first was the establishment in 1982 of the U.S. Court of Appeals for the Federal Circuit in Washington, D.C., a court specifically formed to hear patent cases. Previously, patent cases had been tried in federal district courts, where an estimated 70 percent of patents were successfully challenged. With the new court, however, that statistic was reversed, with about two-thirds of patents upheld.

The second was the Semiconductor Chip Protection Act, passed by Congress in 1984. The new law specifically protected semiconductor design, or "mask work," for up to 10 years. As electronics firms began to exercise their rights, the courts continued to provide stronger legal protection for proprietary chip designs. In 1991, Congress extended the Act through 1995.

The U.S. semiconductor industry experienced generally sluggish conditions during the mid-1980s but entered a period of renewed growth in the early 1990s. Worldwide sales of semiconductors and semiconductor products grew dramatically from 1991 through 1995, from around $50.0 billion to $150.0 billion. The health of the semiconductor industry, though, is dependent on other historically cyclical industries, notably computers, automobiles, and consumer electronics. Consequently, the industry has a history of erratic earnings. As an international industry, it is also affected by economic conditions around the world.

The early 1990s also saw the continuation of the trend toward strategic alliances and corporate partnering among semiconductor companies. This trend was fast becoming an important competitive tool, allowing individual firms to share the ever-increasing costs of production. Entering the mid-1990s, U.S. semiconductor manufacturers were shifting their attention from commodity products to the development of innovative proprietary products, which they had begun vigorously protecting with the help of new patent legislation.

Two additional factors were expected to contribute to the continued growth of the semiconductor industry: overall increases in worldwide sales of electronic equipment and the increasing semiconductor content of electronic products. This growth was driven by the increasingly sophisticated nature of consumer electronics. Manufacturers of fax machines, notebook computers, and camcorders, for example, used semiconductors in these products to perform increasingly complex operations.

The continuing development of Integrated Services Digital Network (ISDN) technology was expected to provide an important new market for chipmakers in the future. The ISDN is a high-speed digital communications network capable of carrying voice, data, and video signals simultaneously over existing telephone lines. The network, which was first commercially introduced in the early 1990s, requires large numbers of semiconductors.

Another factor in the industry was the shrinking number of production options available to players in the field. Many companies that emerged in the 1990s in this realm farmed out production to other facilities with spare capacity in their wafer fabrication plants. As Business Week noted, by using these facilities, "U.S. entrepreneurs avoided the main hurdle for a chip start-up: the tens or hundreds of millions in wafer-fabrication costs. A new venture could thus devote its resources to innovative designs … By pioneering these cutting-edge products, fabless companies grew faster and earned higher returns than established chipmakers. However, excess capacity disappears in times of high demand, and companies without their own production facilities faced possibly substantial investment to secure guaranteed access to production facilities."

By the mid-1990s the semiconductor industry had become one of the most explosive segments of the economy as worldwide sales surged from around $100.0 billion in 1994 to nearly $150.0 billion in 1995. Historically the semiconductor industry was cyclical, though, with semiconductor products having short life cycles caused primarily by rapid technological innovations and resulting in pricing pressures. Over-expansion of fabrication facilities in times of strong demand also contributed to the cyclical nature of the industry.

The demand for chips was driven not only by the increasing sales of PCs but also by the use of chips in consumer electronics, telecommunications, and networking. As inventory exceeded demand in late 1995, DRAM prices started plummeting, creating an overall impact on the global chip market. Worldwide sales declined in 1996, with DRAM prices remaining low and worldwide sales staying flat through 1998.

Worldwide semiconductor sales for 1999 were projected to rise to $144 billion, according to the Semiconductor Industry Association (SIA). Analysts hoped a strong 1999 would signal a growth phase for the industry, as the Asia/Pacific region and Japan began to show signs of strength following the Asian financial crisis of 1998.

Following three years of flat sales and declining prices for DRAM chips, semiconductor manufacturers cut their capital spending budgets in 1998 by 21 percent. As demand began to rebound in 1999, chipmakers began utilizing more capacity, causing a tightening supply situation. That prompted some of the major manufacturers, such as Texas Instruments in the United States and others in the Pacific Rim, to expand their capital budgets.

Current Conditions

In 2001, the semiconductor industry came crashing down from a record year in 2000. Sales fell from $204 billion to $139 billion. In the February 4, 2003 issue of Electronic News , IC Insights President Bill McClean provided information about the conditions that led to the industry's woes in 2001. The publication explained: "For the first time ever the semiconductor industry found itself simultaneously facing each of the four major causes of a downturn: global recession, inventory surplus, overcapacity issues, and a decline in electronic systems sales."

A period of recovery began in the last quarter of 2001 and continued throughout 2002. Subsequently, the industry achieved a modest 1.3 percent improvement in 2002 as worldwide sales grew to nearly $141 billion. In terms of the future, Gartner Dataquest forecast the industry would grow just over 12 percent in 2003, according to EBN . The SIA provided a more optimistic outlook, forecasting an annual growth rate of about 20 percent for 2003 and 2004, with sales reaching new record levels in 2004.

The growth of a global, networked economy—and the resulting demand from data and telecommunications markets—has been an important growth factor for the semiconductor industry. Demand in the communications market stems from the need for greater bandwidth and faster transmission of data, as well as from the growth of the Internet and wireless communication and the resulting buildup of a communications network infrastructure. In 2002 and 2003, wireless local access networks (WLAN) were an especially strong market for the semiconductor industry, with growth rates greater than 35 percent forecast through 2005, according to the SIA.

Along with a proliferation of wired and wireless information appliances, the cellular handset market also was an important growth category for semiconductors in the early 2000s. This category achieved double-digit growth in the fourth quarter of 2002, fueled in part by new subscribers in Asian nations, especially China where the SIA reported that some 5 million new wireless users were being added monthly.

Despite the growth afforded by demand from the communications sector, PC sales remained in the early 2000s an important factor in the growth of the semiconductor industry. While PC sales showed strong growth in the late 1990s, by the early 2000s the PC industry was struggling amid a weak economic climate that presented challenges in business and consumer markets alike. This factor contributed to the semiconductor industry's woes.

Industry Leaders

The semiconductor industry is truly international, with major manufacturers located in Japan, Korea, and Europe, as well as the United States. The top three U.S. manufacturers in 2002 were Intel Corporation, with sales of $26.8 billion and status as the world's leading chip manufacturer; Motorola, Inc., with sales of $26.7 billion; and Texas Instruments, Inc. with sales of $8.4 billion.

The leader in electronics and electrical equipment industry, Intel was also ranked as one of the 10 most admired U.S. companies by Fortune . Intel holds a more than 80 percent share of the microprocessors market because of the success of its Pentium chip. Intel was founded in what would become California's Silicon Valley in 1968 by industry pioneers Robert N. Noyce, Gordon E. Moore, and Andrew S. Grove. Starting with 12 employees, Intel pursued research that led to the development of the first computer chip. The company also played an instrumental role in the development of metal oxide semiconductor (MOS) technology.

Originally a supplier of semiconductor memory for mainframe computers and mini-computers, Intel eventually became a leading supplier of microcomputers. The company sells its microcomputer components, modules, and systems directly to companies that incorporate them into their products. These buyers are primarily computer systems manufacturers, but they also include makers of automobiles and a wide range of industrial and telecommunications equipment. The company also sells personal computer enhancements and networking products through distributors, resellers, and retail stores worldwide. Intel has design, development, production, and administration facilities throughout the Western United States, Europe, and Asia.

Texas Instruments ranked fourth in the world semiconductor industry during the early 2000s. Headquartered in Dallas, Texas, the company has manufacturing, sales, or engineering services in more than 25 countries. In 2002, semiconductor sales represented 82 percent ($6.9 billion) of TI's total revenues.

The company was founded in 1930 as the "Geophysical Service" by J. Clarence "Doc" Karcher and Eugene McDermott. It was the first independent contractor to specialize in reflection seismograph methods of exploration. The firm's name was changed to Texas Instruments, better known as TI, in 1951. The company entered the semiconductor business in 1952 with the purchase of a license from Western Electric Company to manufacture transistors. In addition to semiconductors, TI products and services include software productivity tools, computer and peripheral products, electrical controls, and consumer electronics products.

Motorola Corporation ranked eighth among the world's semiconductor companies in the early 2000s. Paul V. Galvin founded the company in 1928 in Chicago, Illinois. As the Galvin Manufacturing Corp., the company's first product was a "battery eliminator" that allowed consumers to operate radios directly from household current instead of the batteries supplied with early models. In the 1930s the company successfully commercialized car radios under the brand name "Motorola." The company's name was changed to Motorola, Inc. in 1947, the same year it began research into solid-state electronics.

Motorola's semiconductor division designs and produces a broad line of discrete semiconductors and integrated circuits, including microprocessors, microcomputers, and memory products. These products are sold to computer, consumer, automotive, industrial, federal government/military, and telecommunications markets. By 2003, the company had developed expertise with specialized embedded semiconductors used in the wireless communications market, as well as the networking and transportation industries. In addition to being one of the world's leading providers of semiconductor technology, Motorola also provides wireless communication and advanced electronics equipment and services to worldwide markets. The company maintains sales and service offices around the world.

Workforce

Semiconductor jobs more than doubled from 115,200 workers in 1972 to 258,500 workers in 1996. According to the U.S. Department of Labor, the all-time high of almost 300,000 semiconductor workers in 1985 was reached in a robust economy. However, from 1985 to 1993 employment levels declined, in spite of a brief upsurge in 1988. U.S. firms employed about 284,000 workers in 2001, according to the Semiconductor Industry Association.

Within the United States, average hourly earnings for production workers in the semiconductor industry were $18.21 in 2000, significantly higher than the average for all manufacturing jobs. According to the U.S. Department of Labor, this partially reflected the higher skill and value added associated with the capital-intensive processing steps and research and development activities performed by U.S. semiconductor workers.

America and the World

In terms of semiconductor consumption, for many years the Americas formed the world's largest market. However, in 2001 the SIA reported that the Asia/Pacific region had become the world leader in this category, according to Standard & Poor's. With a 28.7 percent share, the region was only slightly ahead of the Americas (25.7 percent). Following the Americas were Japan (23.9 percent) and Europe (21.7 percent). Asia/Pacific was expected to maintain its new status well into the 2000s, supported by positive growth projections in the region.

The United States had a trade deficit in semiconductors throughout the 1990s and into the early 2000s. In 1995 the deficit amounted to some $16.0 billion, with U.S. imports valued at $38.0 billion and exports at $22.0 billion. By 2000, the deficit stood at about $3.5 billion, with U.S. imports valued at $48.2 billion and exports at $44.7 billion. In 2001 the industry achieved a trade surplus of $2.2 billion, with imports valued at $30.8 billion and exports at $33.0 billion.

In 1986, the United States and Japan entered into an agreement called the U.S.-Japan Semiconductor Arrangement, which was designed to eliminate dumping of Japanese products in world markets and to increase market access for foreign semiconductor manufacturers in the Japanese market. In 1991 the follow-on agreement to the 1986 agreement went into effect. The new agreement reflected U.S. expectations that more than 20 percent of the Japanese market could be captured by foreign suppliers by 1992 through the efforts of government and industry. By 1999 foreign chips accounted for about one-third of the Japanese market, up from only 8 percent in 1986.

Research and Technology

According to the SIA, about 18.0 percent of U.S. semiconductor industry revenues is spent on capital, and 18.5 percent of revenues is spent on research and development. This totaled more than $31 billion in 2001 alone. Costs for new semiconductor fabrication facilities are a major capital consideration for many companies. High-end DRAM wafer-fabrication facilities, for example, can cost more than $1.5 billion. The SIA revealed that, in 2001, the industry was expected to produce roughly 60 million transistors "for every man, woman and child on Earth." The association projected that number would climb to 1 billion by 2010.

High Definition Television. One emerging technology that could trigger a boom in semiconductor sales is high definition television (HDTV). HDTV produces pictures that are four to five times clearer than the standard television picture. In addition to commercial broadcast television, HDTV technology also could find applications in areas such as medical imaging and computer graphics. Since the sets require a huge number of semiconductors, they are expected to be a major new market for chipmakers.

Fuzzy Logic. Another emerging semiconductor technology expected to create important future markets was called "fuzzy logic." As Standard & Poor's Industry Surveys noted, "fuzzy logic allows microcontrollers to create gray areas between the yes/no, on/off choices of the binary world. The result is that engineers can design microprocessors that allow machinery to operate with gradual refinements."

With the unprecedented growth in the mid- to late 1990s, some believe the industry faces a slowdown due to the eventual breakdown of Moore's Law, according to The Wall Street Journal. Moore's Law, named after Intel's co-founder Gordon Moore, refers to the accumulation of transistors on microchips, doubling the computing capabilities on a single chip the same size every 18 months. Moore's Law was seen to reach its limits due to barriers imposed by quantum physics. According to The Wall Street Journal, the end of continued progress within the semiconductor industry was slated for the year 2010 by the SIA. This prediction prompted American companies to initiate new research projects.

Further Reading

Ascierto, Jerry. "A New Rising Sun?: Japan's Chipmakers Seek the Dawn." Electronic News, 16 August 1999.

——. "SIA: Sustained Growth into the Next Millennium." Electronic News, 1 November 1999.

Business Rankings Annual 2000. Farmington Hills, MI: Gale Group, 1999.

Cassell, Jonathan. "Dealing with the ODT." Electronic News, 1 November 1999.

Chappell, Jeff. "Don't Hold Your Breath: Capital Equipment Industry Will See Very Modest Growth at Best in 2002, No Capacity Buying Until 2003." Electronic News, 1 January 2002.

Delano, Daryl. "Economic Growth Exceeds Expectations." Electronic Business, July 1999.

"The Dragons Bulk Up." Business Week, 12 July 1999.

"Expect Renewal of Semiconductor Deal with Japan by End of July." Purchasing, 17 June 1999.

Gawel, Richard. "Load Up! Semiconductors Moving up in 1999." Electronic Design, April 1999.

Grossman, Steven. "Today's Global Chip Supplier." Electronic News, 29 November 1999.

Harbert, Tam. "Busted Barriers." Electronic Business, June 1999.

——. "High-Tech Runs into an Export Control Wall." Electronic News, 29 March 1999.

——. "The Little Company that Could." Electronic Business, April 1999.

"IC Industry is Mending." Purchasing, 16 September 1999.

MacLellan, Andrew. "New Forecasts Paint a Divergent Picture of the Chip Market's Future." EBN, 18 November 2002.

Morrison, Gale and Tam Harbert. "SIA Lauds China Ruling." Electronic News, 7 June 1999.

Port, Otis. "For Chipmakers, Less of Moore?" Business Week Online, 18 November 2002.

Romanelli, Alex. "2001: What Didn't Go Wrong? Researcher Details What Caused the Chip Industry's Worst Year Ever." Electronic News, 4 February 2002.

Semiconductor Industry Association. "Global Chip Sales Up 1.3 Percent in Recovery Year," 3 February 2003. Available from http://www.sia-online.org .

——. "Industry Facts & Figures," 22 February 2003. Available from http://www.sia-online.org .

——. "Semiconductor Forecast Summary, 2002-2005," November 2002. Available from http://www.sia-online.org .

Standard & Poor's. Standard & Poor's Industry Surveys: Semiconductors. New York: Standard & Poor's Corporation, 14 October 1999.

——. New York: Standard & Poor's Corporation, 18 July 2002.

U.S. Census Bureau. Annual Survey of Manufactures. Washington, D.C.: U.S. Department of Commerce, Economics and Statistics Administration, U.S. Census Bureau, February 2002. Available from http://www.census.gov .

U.S. Department of Commerce: Bureau of the Census; International Trade Administration (ITA). "Trade and Economy: Data and Analysis," 23 February 2003. Available from http://www.ita.doc.gov .

"Worldwide Chip Sales Continue to Expand." World Trade, February 2003.



User Contributions:

Comment about this article, ask questions, or add new information about this topic: