Four years ago, Texas Instruments Inc. was the dinosaur of the semiconductor industry. The company that invented the integrated circuit in 1958 had failed to keep pace with the technological revolution sweeping the chip industry. During the 1980s, the company's share of the chip market plummeted to 5% from 30%. Worse, TI slipped into the red in 1989, with the company posting a loss of $39 million.

Nowadays, though, TI is looking a lot healthier. The Dallas-based company has spent the past few years drastically overhauling its strategy and remaking its culture. The inventor of the digital watch and an early entrant into personal computers, TI has scrapped its plans to build a billion-dollar consumer business. Instead, it's focusing on high-tech corporate clients. And this traditional supplier of cheap, off-the-shelf memory chips is moving upscale. The company now gets almost 50% of its revenues from high-margin products such as microprocessors and customized chips, compared with just 25% or so in 1988.

For a supposed dinosaur, TI is acting remarkably nimble. Despite the recent recession, the company managed to increase its manufacturing capacity by roughly a third by pursuing joint ventures in Europe and Asia. That in turn has enabled it to cash in big on the now-surging demand for chips. On July 16, TI said that its earnings jumped 55% in the second quarter, to a record $112 million, as revenues grew 13%, to $2.1 billion. For the year, analyst Millard Phelps of Hambrecht & Quist Inc. expects profits to rise by 41%, to $417 million, with revenues increasing 5%, to $7.8 billion.

HOT STOCK. Not that all TI's worries are behind it. More than half of its chip revenues still come from basic commodity chips. And analysts fear that an imminent slowdown in the PC market could depress chip demand and snuff out TI's turnaround before it captures a significant share of the market for more sophisticated chips. TI's stock, a hot performer this year, fell 6% the day it reported its record profits, largely because of concern about the chip business following a loss at Apple Computer Inc. and expected red ink at Dell Computer Corp. The shares have since recovered to about 73.

The tale of TI's transformation begins in June, 1989, when CEO Jerry R. Junkins called his top 20 deputies to an Austin (Tex.) hotel for a week-long strategic-planning exercise dubbed TI 2000. Many attendees expected another routine corporate thinkfest. Instead, as one attendee recalls, the genial Junkins, who has been CEO since 1985, voiced what most of the group already knew: "We've been infected by mediocrity, and we're not the company we thought we were."

That certainly was clear. By 1989, TI had dropped from a dominant first to a moribund seventh in chip sales. Why the slide? The company's preoccupation with building a consumer business distracted TI from its bread-and-butter chip business. First came its failed attempt to build a home computer. Two years later, in 1985, TI's personal computer flopped. Its technology wasn't so much to blame as a failure to read the market. TI's PC, for example, couldn't use the widely available software developed for the IBM PC, the industry standard.

One of Junkins' first tasks: revamp the tightly controlled culture that long prevailed at TI. His predecessor, Mark Shepherd Jr., had closely monitored every facet of TI. And executives often were reluctant to speak their minds. But Junkins wasn't interested in micromanaging. He believed less meddling from headquarters would lead to a freer exchange of ideas and ultimately to more innovative thinking. Junkins also encouraged senior managers to loosen up. "We outlawed the use of slides in the presentations and sometimes didn't even set agendas," recalls William F. Hayes, who runs TI's defense unit.

Junkins also knew TI had to redefine its basic business if it hoped ever to regain its past glory. After scaling back TI's ambitions in the consumer business, he took a hard look at the chip industry. With technology advancing and competition intensifying, customers were demanding more complex, customized chip products. And that was devastating news for a commodity-chip producer such as TI, which sold most of its chips through a catalog.

NEW FAITH. One way to overcome this market disadvantage, in Junkins' view, was to start jointly designing chips with key customers. TI had long felt its chip knowhow was the best in the business and paid little attention to outside technology. "We were so successful in the 1950s and 1960s that we became a real arrogant bunch of young SOBs," admits former CEO Shepherd. But by collaborating with clients, Junkins believed TI would not only start making chips that had guaranteed buyers but also become an indispensable partner. "We're looking for shared dependence," says Junkins. Moreover, joint development would expand TI's technology capabilities, allowing it to make complex chips that could be sold to a wider audience.

That was clearly the goal in 1989, when TI agreed to make microprocessors for Sun Microsystems Inc. At the time, Sun was trying to perfect its Sparc microprocessor for its next generation of computer workstations. A number of licensees had been trying without success to meet Sun's needs. But TI went further than other chipmakers. The company loaned Sun its own engineers and shared its own proprietary technology to help perfect Sparc. As a result, TI has emerged as one of the top suppliers of Sun microprocessors. "TI really understands how to form these partnerships and make them grow," says Mel Friedman, Sun's vice-president for purchasing. The companies are now making joint sales calls to market the chips and exploring ways to share distribution channels.

Pursuing a similar strategy, TI has become a leading chip supplier for Sony, General Motors, and Ericsson. Indeed, next year TI and the Swedish telecommunications giant plan to install a pilot production line to design chips. Once prototypes are developed, the designs would be transmitted electronically to TI fabrication sites.

A daring effort to expand TI's manufacturing capacity is also paying off. TI and other chipmakers were scrambling when the PC market exploded in the

1980s and demand for basic chips soared. U.S. chipmakers, such as Intel Corp., decided to focus on new technologies and surrender much of the commodity chip market to Japanese rivals. By contrast, Junkins believed the basic chip market was critical to TI's future.

It was a risky strategy. Junkins doubled TI's investment in new plants, to more than $1.7 billion between 1989 and 1990. But rather than go it alone, as TI had done since the 1960s, he stretched the investment by seeking out joint ventures in Europe and Asia. TI's first partner was the Italian government. Eager to create new jobs, Italy put up half of the $1.2 billion needed to build a state-of-the-art chip factory in the depressed mountain town of Avezzano in 1990. Three similar deals followed over the next two years with Taiwanese PC maker Acer, Japan's Kobe Steel, and, this past April, with a consortium composed of Hewlett-Packard, Canon, and the Singapore Economic Development Board. The plants in Italy and Taiwan are in full production; the other two should be operating in a year.

NO GUARANTEES. Despite its success, TI's long-term prosperity isn't assured. After growing by 16% in 1992, the semiconductor market's annual growth is expected to slow to 5% by 1995, says Dataquest Inc., a market research firm. And analysts wonder if TI's push into more complex chip markets will produce results fast enough to offset that slowdown. After all, rivals from Motorola Inc. to Toshiba Corp. are also forging customer links to expand their businesses.

And for all its efforts, TI remains dependent on commodity chips. TI Europe President Roberto Schisano says specialty chips should account for 60% of the Avezzano plant's output by 1995, vs. 35% in 1992. But the plant conversion has been postponed because of the demand for commodity chips. What's more, TI's joint-development programs with its high-tech customers have yet to result in sophisticated chips that have blockbuster sales appeal to the broader market, as Junkins had hoped.

For now, Junkins isn't too concerned. All of TI's new plants, he points out, will eventually be capable of making specialty chips. And despite the competition, TI's business relationships are growing. True, the company may not be the market leader it once was--but its fossilized years seem well behind it.

Before it's here, it's on the Bloomberg Terminal. LEARN MORE