Don't Mess with Texas Instruments The Dallas company conquered the market for specialized cell-phone chips. Now it's doing it again with chips for big-screen TVs.
(Business 2.0) – Only a decade ago, Texas Instruments, the once mighty company where engineer Jack Kilby invented the integrated circuit and essentially launched the modern electronics business, had the look of a fading power. It was bloated, adrift, and increasingly irrelevant. Dabbling in so many diverse industries, from calculators to weaponry, TI ended up ceding the vast market for PC chips to Intel.
But now, slimmer and refocused, TI has fought its way back: Its market value has tripled since 1996 to $36 billion, and it generates more than $1.35 billion a year in free cash flow. How? The Dallas-based company essentially conquered the market for microprocessors that transform a humble cell phone into a multifeatured digital camera and jukebox. And TI has emerged as the front-runner in a new and potentially enormous market--chips that make big-screen projection TVs brighter, thinner, and ultimately cheaper. Analysts predict that the TV-chip business will grow at a rate of 40 percent or more for the foreseeable future.
TI's turnaround is all the more remarkable considering the competition in the market for cell-phone chips: industry goliath Intel and wireless pioneer Qualcomm. Says Will Strauss, president and principal analyst at electronics market research firm Forward Concepts, "TI's dominance is really unchallenged at this point."
A Clearer Focus
Such a statement would likely have shocked the company's leadership in the early 1990s, when TI dominated nothing but bad news. It had stuck its silicon and wires into a hodgepodge of business segments, from oil and gas exploration to computers, educational toys, and radar and missile systems. By the middle of the decade, TI was so unwieldy and scattered that it was losing its grip in semiconductors too. Intel had cornered the market for microprocessors as the personal computer industry took off, while TI's mainstay--memory chips--had become a low-margin commodity. Making matters worse, in 1996, just as then-CEO Jerry Junkins was attempting to reinvigorate the company, he died of a heart attack.
Richard Templeton, then TI's semiconductor chief, took the first step in winning the cell-phone market by looking for ways to exploit the company's underused digital signal-processing (DSP) technology--chips that translate sounds and images into digital bits. Until he came along, the best identified use of the chips was in the Speak & Spell toy that ET hacked to phone home. Templeton realized that the emerging cell-phone market was a perfect opportunity for the technology. Focusing on cell-phone chips was a bold move for the new CEO, Tom Engibous, because he was betting the company on the potential of a very narrow market segment. (Both Engibous and Templeton got their reward. Engibous is now chairman and Templeton is CEO.) The next step, though, was key: finding a partner that wanted a win as much as TI did.
TI began forging a stronger relationship with Nokia, the Finnish telecom giant. (Back in the early '90s, Nokia was a nearly bankrupt also-ran.) To win Nokia's confidence, TI said it would make programmable DSP chips to Nokia's specifications, so Nokia could put its own software on the chips and customize them. The agreement made it much more difficult for Nokia to switch to another supplier, but it gave Nokia a leg up on its rivals. Once the technology proved popular and profitable, every other cell-phone maker wanted in. Today about half of the world's 460 million cell phones have TI's DSP chips inside. TI's closest competitor, Qualcomm, has only about 18 percent of the market. And Intel, despite spending more than $1.6 billion to establish its presence, has yet to make a dent.
DSP chips are now also used in five of the 10 top-selling point-and-shoot digital cameras, and Templeton is urging TI scientists to integrate even more functions on a single chip so cell phones can be used as digital camcorders, digital TVs, and 3-D gaming platforms.
What's in the Basement?
But with demand for phones tapering off, TI needs another big win--and big-screen TVs could be it. So TI again rummaged through its technology basement. It found a microchip that was being used in conference room projectors and airline ticket printers. Known as a digital light-processing (DLP) chip, it uses a million or more tiny hinged mirrors to magnify images. TI engineers tweaked it for TV so that when a powerful light is shined onto the mirrors, each mirror tilts, as often as 5,000 times a second, creating an image that is then magnified and projected on the rear of a screen. The result, TI's designers say, is a picture with higher contrast, sharper images, and better motion reproduction than that provided by plasma or liquid crystal display televisions.
TI was convinced that TVs with its chip would cost less than any competing technology. But it wasn't a slam dunk. Even the chip's inventor, Larry Hornbeck, was skeptical. Who would mass-produce such a TV to make it affordable? "I said, 'Let's see if there's ever one of those TVs at Best Buy--that'll never happen,'" he recalls.
Of course, TI had to find its Nokia for this project. That happened at the big Consumer Electronics Show in January 2001 when John Van Scoter, general manager of the DLP group, was impressed by the technology he saw at the Samsung Electronics display and told his execs, "Don't leave this show without getting a meeting." While the Korean manufacturer made high-quality consumer products, it didn't have the cachet in the U.S. market of a Sony or a Panasonic. It needed an attention-getting product as much as TI needed a strong follow-up to DSP--the same symbiosis that helped produce TI's win with Nokia. A few weeks later, Samsung sent an army of engineers to Texas. After a few more trips across the Pacific, the two companies agreed that they could build TVs that would sell for less than $3,000. The price was low enough to generate sufficient sales volume to drive down costs and create great consumer demand. "Since we made the decision to work on this project with TI, it's been a nonstop move to market," says Steve Panosian, Samsung's director of marketing for digital television, who praises his partner for providing much-needed help in setting up training programs for retailers across the country. "It's the classic right product at the right time at the right place."
Last year Samsung sold more than 100,000 sets in its DLP-based product line. (A further dividend: Samsung's digital-TV market share rose from 5 to 15 percent.) Overall, analysts expect consumers to buy 750,000 DLP TVs this year, three times more than last year. The TVs, now less than 18 inches deep, are getting slimmer. Other companies are making them; by Christmas more than 50 models will sport the technology, including an RCA that is a mere 7 inches deep. And right now at Best Buy, a 61-inch Samsung DLP TV goes for $4,500, roughly half as much as a comparable LCD or plasma TV. "I was sweating it out in the first half of 2001," says Dale Zimmerman, general manager of TI's home entertainment unit, "but working with Samsung was the best decision I ever made." Indeed, in July, TI announced a second-quarter net profit more than triple that of a year earlier, fueled in part by a 90 percent rise in DLP sales.
Naturally, Intel is fighting back; it recently announced an effort to improve and produce liquid crystal on silicon panels--known as LCOS--an alternative to DLP. Intel says this technology could allow a manufacturer to produce 50-inch high-definition TV sets priced at less than $1,800 by 2005.
TI is certain that as confused TV buyers struggle to make their choices--DLP, LCOS, plasma, LCD, old-fashioned cathode-ray tube--its technology can beat or match the competition. The unknown is whether years of consumers seeing "Intel Inside" on PCs will trump the Lone Star chip.
Kathryn Jones is a contributing editor at Texas Monthly. Additional reporting by Business 2.0 senior reporter Brian Caulfield.