2900 Semiconductor Drive
National Semiconductor will put systems on a chip for our key trendsetting data highway partners, using our analog expertise as a starting point for forward integration.
National Semiconductor Corporation is one of the leading American manufacturers of semiconductors used in a broad range of electronics applications. During its rapid rise to prominence in the late 1970s, National Semiconductor gained a reputation as perhaps the most efficient producer of semiconductors in the world, turning out a wide array of standardized, reliable parts at very low cost. National's prosperity relied less on high-tech genius than on low-tech frugality and hard work, qualities instilled in the company by its longtime president and chief executive, Charles E. Sporck. In the increasingly crowded world of semiconductors, however, National suffered during the 1980s from Asian price competition, leading to a series of money-losing years. National staged a turnaround in the mid-1990s under Sporck's successor, Gilbert F. Amelio, and is now--under another new leader, Brian Halla--staking its future on sophisticated, higher-margin products, most notably the system-on-a-chip.
Founded in 1959
"Semiconductor" is the name given to a group of elements which under normal conditions do not conduct electricity, but which when slightly modified can be used as conductors with great precision and reliability. The development of the modern electronics industry, beginning with the 1949 invention of the transistor, has depended on the progressively more subtle use of semiconductors to control and direct electricity in very small packages known as integrated circuits or "chips." In 1959 Dr. Bernard Rothlein, formerly of Sperry Rand Corporation, joined the burgeoning semiconductor industry by creating National Semiconductor in Danbury, Connecticut (the company moved to Santa Clara, California, in 1968). The company was tiny by industry standards, with only $5.3 million in sales by 1965, but it offered a variety of fairly sophisticated semiconductors and was operating at a profit.
Dr. Rothlein's former employer filed a suit against National for patent infringement, however, and the case depressed the company's stock price when it reached the courts in the mid-1960s. The low stock price encouraged a substantial investment by East Coast financier Peter J. Sprague (son of the chairman of Sprague Electric Company), who became chairman of National in 1966 and set out to make the company a major player in the semiconductor industry. Sprague recognized that National needed an injection of strong management if it was to make the transition from small research lab to commercial manufacturer, and in the spring of 1967 he surprised the industry by hiring away five top executives from Fairchild Camera & Instrument Corporation, then the nation's second largest maker of semiconductors.
Sporck Hired in 1967
Chief among the new recruits was Charles E. Sporck, 39-year-old head of Fairchild's semiconductor division, who accepted a 50 percent cut in pay to become National's president (and owner of a chunk of its stock). From then until 1991 National Semiconductor would remain largely the creation of these two men, and especially of the hard-driving Charlie Sporck.
Sporck and National Semiconductor were ideally suited to each other. National had some excellent products but lacked management control, while Sporck was not a technical genius but knew how to run a tight ship, market his wares, and make money. With the full financial and moral backing of Peter Sprague and the board of directors, Sporck turned National upside down in the 12 months following his arrival. He marked down the value of National's inventory of transistors by $1.5 million (helping the company to a $2 million loss in fiscal 1967) and focused its energies on selling large quantities of standard semiconductors in three different market areas: linear, Transistor-to-Transistor Logic (TTL), and metal-oxide semiconductors (MOS). He also kept a tight lid on corporate overhead, using outside sales representatives whenever possible, farming out basic engineering and accounting work to independent contractors, and generally promoting a corporate ethic of spartan austerity. In an industry rapidly flooding with new competitors, Sporck's penny-pinching would prove the key to National's survival in the coming price wars. As he told Business Week in 1970, "We make money because we have to."
Though it might not have been apparent to the casual observer in 1967, National Semiconductor had assembled a trio of powerful business advantages. Most fundamentally, National was in an industry about to undergo tremendous growth, as the spread of computers made the semiconductor critical to every aspect of modern life; it could draw on the financial strength of its investors in order to raise the large amounts of money required for expansion; and the company was run by a man naturally inclined to efficiency and thrift. The combination of these elements allowed National to grow with amazing swiftness, 1965's sales of $5.3 million becoming $42 million in 1970 and $365 million in 1976. For other, less well-prepared companies, the same period was fatal, as bitter price wars erupting in 1969 and 1970 drove even giants such as General Electric and Westinghouse out of the semiconductor business and kept profit levels minuscule everywhere. Silicon Valley was suddenly very crowded, and from among its scores of visionary entrepreneurs only a few would survive to dominate the national scene.
Sporck early on brought a global awareness to National. National was one of the first semiconductor companies to move its assembly operations to the Far East, where labor was available at a fraction of its cost in the United States. The company also sold about 20 percent of its finished products overseas, much of it going to Europe at prices that stirred charges of unfair trade practices. On the other hand, Sporck appeared to have grossly underestimated the long-term potential for competition from the Far East, where the growth of an indigenous semiconductor industry drove prices ever lower on the kind of standardized semiconductors made by National. Sporck's preference for selling a high volume of standard items made possible National's prodigious growth, so long as its competition was limited to American firms operating on similar cost bases; when the Japanese made semiconductors a global business in the 1980s, the same formula brought consistent losses and a desperate appeal from Sporck for federal trade protection. National was at the forefront of political pressure leading to the 1986 Semiconductor Trade Agreement.
Failed 1970s Diversification Efforts
Sporck was aware of National's vulnerability, of course, and beginning in the early 1970s he made a number of attempts to diversify the company's sales base by "integrating forward," making consumer products as well as the semiconductors that went inside them. National leaped into the manufacture of calculators, digital watches, and video games, enjoying initial success as the public responded to the novelty of these high-tech gadgets. Within a few years, however, National's emphasis on low-price mass merchandising again left it vulnerable to the crush of competitors entering these markets. National had no experience in retail manufacturing, and before long its products were saddled with a reputation as low-end junk, without the style or cachet needed to survive in a maturing market. By the time everyone in America was wearing a digital watch National had been driven from the marketplace, suffering minor losses which were fortunately overshadowed by its roaring success in semiconductors.
Two other efforts by Sporck to widen National's product line had more complex histories, though both also ended in failure. In the mid-1970s, National became interested in the possibility of electronic point-of-sale terminals for use in supermarkets. In association with a group of California supermarketers, National developed the "Datachecker" system for the scanning and recording of sales, with which it built a substantial and modestly profitable business over the following decade. Of greater potential was National's decision around 1976 to enter the computer business, originally as a producer of mainframe computers for sale by Itel Corporation, a San Francisco-based marketing and finance company. As many others had tried before, Itel and National were hoping to cut into IBM's domination of the mainframe market by selling similar machines at a much reduced price. At first, National was satisfied simply to make computers for the Itel name; in the late 1970s, however, National tried to push into the market with its own line of mainframes and minicomputers, encouraged by the huge profits it was making on the Itel IBM-compatibles.
National's System 200 and 400 lines of large computers never got off the ground, due in part to renewed competition from the ever vigilant IBM. Itel had similar but more severe problems, forced by IBM price pressure in early 1979 to ask National for cheaper computers with which to compete. Sporck recognized a golden opportunity: he agreed to supply Itel with cheaper computers only if they would agree to buy more machines. Itel foolishly did so, and when the market softened later that year Itel was stuck with computers it could not sell and an obligation to buy a lot more of them. Faced with a complete disaster, Itel essentially gave its inventory and sales force to National in exchange for a release from its contracts; as one former Itel executive told Fortune, "National blackmailed us and then stole the business." Be that as it may, Sporck had "stolen" little more than a distraction and headache. National enjoyed sporadic success selling a line of mainframes made by Hitachi in Japan, but again its timing and approach were all wrong. Aside from the inherent difficulties of competing with IBM, the 1980s also witnessed a general decline in the mainframe segment of the computer industry that even IBM was not able to withstand successfully. The monolithic mainframe was being replaced by combinations of mini and microcomputers, and in a shrinking market National's chances of success against IBM fell from slim to none. The company sold its National Advanced Systems division in 1989.
Buffeted by Asian Competition in the 1980s
In the meantime National's semiconductor business continued to boom. Sales for 1981 reached $1.1 billion--tripling in four years--and the company employed 40,000 workers, two-thirds of them in Southeast Asian assembly plants. Its strength continued to be the manufacture of linear and bipolar logic integrated circuits in large quantities and at low cost; as a competing executive told Fortune, National was the "sweatshop of our industry--a pipe-rack, low-cost, survival-oriented company." But National was soon faced with a number of grave problems. In 1981, a handful of key National executives left the company to accept more lucrative offers elsewhere, among them Pierre Lamond, for years National's chief designer and engineer. Job-hopping is endemic to the electronics industry, but the defection of Lamond was still a significant blow for National: never strong on technical ingenuity, the company had now lost its brightest designer. Worse yet, competition from the Far East was now at flood tide, and the first to feel its impact would be companies like National who lived by volume and price appeal. National's line of staple items could be reproduced anywhere, and even better than having two-thirds of one's employees in the Far East was to have all of them there. Although the cheapest maker of semiconductors in the United States, National was not the cheapest worldwide.
The recession of 1981-82 plunged National into the red. Its losses totaled only $25 million, but they set the tone for the coming decade. The company was solid in the following few years, suffered a sharp downturn in 1985 and 1986, and then recovered again in late 1987. At this point, Sporck made a purchase that must have given him great personal satisfaction, whatever its permanent value to National Semiconductor. For $122 million National bought the semiconductor division of Fairchild Camera, the same Fairchild from which Sporck and his management team had emigrated back in 1967. Fairchild had been one of the pioneers in semiconductors, but since the departure of Sporck it had staggered through many losing years and was now available at what some observers thought was a bargain price. Sporck felt that Fairchild's strengths in chips for mainframes and supercomputers and its excellent military ties would complement National's relative weakness in those segments of the market. Once again, his timing was poor; the mainframe market continued to shrink and military spending declined from its peak during the Reagan administration. Moreover, Fairchild Semiconductor had lost $265 million in the two years prior to its purchase by National, and the latter already had plenty of its own problems.
From fiscal 1987 through fiscal 1992, National posted an aggregate loss exceeding $500 million. President Sporck and Chairman Sprague took vigorous measures to right their floundering ship, getting out of computers and point-of-sale equipment, dropping the two least profitable semiconductor lines, and in early 1991 replacing Sporck himself with Gilbert F. Amelio. Under the leadership of Amelio, a former Rockwell International executive who had a Ph.D. in physics from Georgia Institute of Technology, National turned more of its energies toward the market for analog semiconductors, used increasingly by the telecommunications industry. National was already the number one producer of analog chips, which for years had been all but lost in the excitement over digital chips for computer applications, and Amelio hoped that the changing demands of the market would make National's analog expertise far more valuable in the 1990s. The immediate signs were not encouraging. Even after writing off $144 million in the massive reorganization of the company in 1991, National came up with a $120 million loss in the following year.
Shift to Value-Added Chips in the Mid-1990s
Nonetheless, for the next three years, as the restructuring reached its culmination, National Semiconductor was on the rebound, posting healthy profits and seeing revenues increase each year. In addition to selling off nonstrategic assets, Amelio also divided the company's chip lines into two areas: lower-margin, more cyclical logic and memory chips--the so-called commodity chips upon which National gained prominence; and higher-margin, value-added analog and mixed-signal chips. The division seemed to point toward the eventual divestment of the commodity chips, but before he could make this dramatic move, Amelio left National Semiconductor in early 1996 to become CEO of Apple Computer Inc. and attempt another turnaround.
Brought on board in May 1996 as new CEO and president (and soon chairman as well) was Brian Halla, a former executive vice-president at LSI Logic Corp. who had also spent 14 years in marketing at Intel Corporation. The month after his arrival, National announced that it had consolidated its commodity chip lines within a new unit, Fairchild Semiconductor, resurrecting the name of the pioneering Silicon Valley chip company that National had bought in 1987. In March 1997 National completed the divestment of Fairchild whereby it sold the unit to an investment company for $550 million, with National retaining a 15 percent stake in Fairchild.
Halla also staked the company's future on the so-called system-on-a-chip, a new type of technology in development that would include all functions required for one device on a single chip. For instance, a PC would require only one chip rather than the several typically required (though in both cases, memory chips are also needed). This would result in cheaper, more reliable, and smaller systems, and the system-on-a-chip was therefore ideally suited to such proposed consumer products as cellular phones with Internet access built in and PCs small enough to fit into a wallet.
National Semiconductor was not the only company pursuing this technology but it seemed to be the most aggressive. The company made a number of acquisitions designed to gain the combination of technologies needed to make the system-on-a-chip a reality. In March 1997 National paid $74.5 million for Mediamatics, Inc., a Fremont, California-based maker of MPEG audiovisual chips for the PC market. Cyrix Corporation was acquired in November 1997 for about $540 million. Cyrix specialized in lower-price Intel-compatible microprocessors used in sub-$1,000 personal computers. Halla planned to use Cyrix chips as a base for the system-on-a-chip, which could be the heart of PCs costing less than $400. In May 1998 National paid $104.8 million to acquire ComCore Semiconductor, Inc., a maker of integrated circuits for computer networking and communications.
The acquisition of Cyrix was perhaps the final piece needed to complete the system-on-a-chip puzzle. National announced in April 1998 that the first product using the new technology would be ready in mid-1999. The company also announced that month that it would close some plants and lay off 10 percent of its workforce, or about 1,300 people, because of slackening demand worldwide. Charges related to these moves and to the acquisition of ComCore led to a full-year fiscal 1998 loss of $98.6 million. Revenues declined to $2.54 billion from $2.68 billion in 1997.
Principal Subsidiaries: ComCore Semiconductor, Inc.; Future Integrated Systems, Inc.; Cyrix Corporation; Cyrix Manufacturing, Inc.; Cyrix International, Inc.; Mediamatics, Inc.; Dyna-Craft, Inc.; National Semiconductor International, Inc.; DTS Caribe, Inc.; National Semiconductor Netsales, Inc.; National Semiconductor (Maine), Inc.; Comlinear Corporation; ASIC II Limited; National Semiconductor B.V. Corporation; National Semiconductor France S.A.R.L.; National Semiconductor GmbH (Germany); National Semiconductor (I.C.) Ltd. (Israel); National Semiconductor S.r.l. (Italy); National Semiconductor A.B. (Sweden); National Semiconductor (U.K.) Ltd.; National Semiconductor (U.K.) Pension Trust Company; Cyrix International Limited (U.K.); National Semiconductor Benelux B.V. (Netherlands); National Semiconductor B.V. (Netherlands); National Semiconductor International Finance S.A. (Switzerland); Natsem India Designs Pvt. Ltd. (India); National Semiconductor (Australia) Pty. Ltd.; National Semiconductor (Hong Kong) Limited; National Semiconductor (Far East) Limited (Hong Kong); NSM International Limited (Hong Kong; 51%); National Semiconductor Sunrise Hong Kong Limited; National Semiconductor Japan Ltd.; Cyrix K.K. (Japan); National Semiconductor SDN. BHD. (Malaysia); National Semiconductor Technology SDN. BHD. (Malaysia); DynaCraft SDN. BHD. (Malaysia); DynaCraft Asia Pacific SDN. BHD. (Malaysia); National Semiconductor Pte. Ltd. (Singapore); National Semiconductor Asia Pacific Pte. Ltd. (Singapore); National Semiconductor Manufacturer Singapore Pte. Ltd.; Shanghai National Semiconductor Technology Limited (China; 95%); National Semiconductor Korea Limited; National Semiconductor Canada Inc.; National Semicondutores do Brazil Ltda.; National Semicondutores da America do Sul (Brazil); Electronica NSC de Mexico, S.A. de C.V.; ASIC Limited (Bermuda); National Semiconductor (Barbados) Limited; Cyrix Export Sales Corporation (Barbados).