THE BAD BOY OF SILICON VALLEY
When T. J. Rodgers speaks, people bristle. The chief executive of Cypress Semiconductor Corp. calls large chip companies "dinosaurs." He accuses their executives of whining for "political protection" rather than innovating and investing. He complains that current government policies are designed to "prop up sagging companies." And he derides Sematech, the government- and industry-funded consortium intended to restore America's edge in semiconductors, as a "corporate country club" for big business.
Who is Thurman John Rodgers, and why does he keep saying those terrible things about his industry? Well-known around Silicon Valley as just "T. J.," Rodgers certainly has a taste for controversy. His insult-peppered insistence that mediocre management is the cause of most of the chip industry's problems, not unfair competition from Japan, has earned him some staunch adherents and plenty of vociferous critics. Depending on whom you ask, he's either an arrogant, publicity-seeking braggart, or a brilliant and innovative entrepreneur. W. J. Sanders III, CEO of Advanced Micro Devices Inc.--and Rodgers' former boss--has compared him to Joseph Goebbels, calling him "the master of the half-truth." Management guru Tom Peters considers him "a national treasure."
ACID TEST. A boyish 43, Rodgers is often dismissed by his elder competitors as a naive upstart. But to his boosters, the founder of this fiercely competitive chip company run with bowstring tightness has combined technical brilliance with business acumen to become an innovative manager who represents a new generation of Silicon Valley executive, redefining entrepreneurship in a tougher, globally competitive world. "Cypress has been a strong rising star almost since its founding eight years ago," says Daniel L. Klesken of Prudential Securities Inc. "As we enter the '90s, it is clearly a force to be reckoned with."
As Cypress becomes more of an industry force, competitors say, Rodgers and his novel management theories will finally be put to a real test. It has been easy, they say, for Rodgers to sing the praises of fast-moving entrepreneurship, dismiss the competitive threat of Japan, and decry the notion that government should have a hand in guiding investment in the semiconductor business. Cypress can be nimble because it's tiny. It earned $33 million on just $225 million in revenues last year, making it a pipsqueak compared with industry behemoths Advanced Micro Devices (AMD), Intel, Motorola, National Semiconductor, and Texas Instruments, with collective revenues of $24 billion. "If Cypress were to disappear tomorrow, nobody would notice," says Craig R. Barrett, a senior vice-president at Intel Corp., the premier U. S. chip company.
And Cypress can shrug off global rivals because it has been a niche player, specializing in a fast memory chip known as the static random-access memory, or SRAM. It is only now, as its volume and product lines continue to expand, that Cypress and its founder will draw real competitive fire from Japanese and U. S. rivals.
It's certainly hard to argue with Rodgers' success so far. Analysts expect Cypress' earnings to rise nearly 16% this year, to $38.5 million, on revenues of $295 million, up 31%--this at a time when U. S. industry growth is around 5%. Aside from a growth rate that matches that of industry pioneer Intel at the same age, investors are impressed with the steady performance of Cypress. Even in tough times, its gross margins hover between 57% and 60% of revenues, and pretax earnings remain above 20%, outperforming virtually every chipmaker except Intel.
The company's long-term outlook looks promising, too. A group of subsidiaries now puts Cypress into new growth markets such as high-speed microprocessors and memory chips capable of powering supercomputers. And it just finished the conversion of a modern manufacturing plant, bought last year from Control Data Corp., that will hike the company's capacity by 75%.
Those accomplishments are the result of Rodgers' tough, frugal, and highly aggressive management style. An electrical engineer with a PhD from Stanford University, he runs his company with an engineer's precision. He has designed software programs to help manage everything from sniffing out minor production glitches to keeping people within deadlines (box, page 70). Every Wednesday, he combs through a dozen printouts on the status of his pet projects, noting the names of managers whose goals are more than five weeks overdue. The offending manager is likely to get a handwritten memo from a pad inscribed "From the Desk of God."
He competes with Japanese giants by using a flexible manufacturing process that enables him to make hundreds of specialized types of chips. The secret is to move fast, making small amounts of the models customers want and changing quickly from one product to another. That sums up Rodgers' basic theory of management: Big companies are bad, small companies are good.
Similarly, he argues, U. S. companies should not build huge plants in an effort to go head-to-head with the Japanese, who have cheap capital and government support. Rather, American companies should stay small, exploiting their unique design technology and unsurpassed system of venture-capital-backed entrepreneurialism to maneuver around the slower-moving Japanese. "If we have to rely on juggernauts to win," he says, "the Japanese will do a better job."
NO FREE MONEY. Cypress, he vows, will never be a huge, unwieldy company. Instead, it will be a conglomeration of small ones. Other companies are expanding by creating or buying new subsidiaries, but management consultant Peters says none has adopted the idea of "perpetual entrepreneurship" as completely as Cypress.
As he expands into new businesses, Rodgers spins new product lines or manufacturing plants into separate companies, each with a president, a board of directors (including Rodgers), and venture-capital-style funding from the parent. If the plants need more funding before creating positive cash flow, they must sell more equity to the parent. That keeps them extraordinarily frugal. Three subsidiaries that design new products now account for about 28% of Cypress' revenues, up from less than 12% only 18 months ago. The manufacturing subsidiaries sell products back to Cypress for resale by its centralized sales force and can find other customers if they wish.
Eventually, the managers of these units will be able to cash out by selling stock back to the parent, to private investors, or even on the public market. Cypress has the right of first refusal, just like any venture capitalist. In April, the three-year-old Round Rock (Tex.) plant sold its employee-held stock, 9% of the total, to Silicon Valley chip company Altera Corp., which also has its own chips manufactured at the plant. That brought some 190 Round Rock employees $7.4 million, including $2 million for the company's president, who then retired.
Just as he has tried to keep subsidiaries entrepreneurial, Rodgers has set up virtually every internal company function as its own profit-and-loss center. A group that tests parts before shipping them, for example, sells its services to the product lines. The upshot: Product managers who pay for the service aren't going to demand exotic tests unless absolutely necessary, since they'll have to foot the bill. "Free money is one of the biggest problems of a large company," Rodgers says. "We've gotten rid of socialism in the organization."
Not everyone agrees that the system gives managers the responsibility and freedom that Rodgers claims. Three years ago, for instance, he fired the founder and president of Aspen Semiconductor Corp., a Cypress subsidiary founded in 1987 to make ultrafast memory chips for supercomputers. The former executive, Narpat Bhandari, sued Rodgers, alleging that Cypress used his technological knowhow to create new products and then tossed him out. Rodgers says Bhandari was a poor manager whose technology was worthless. Retorts Bhandari's son, Ravi, who was involved in his father's court case: "Every decision he made was undermined by T. J. Rodgers from the start. You had to get his approval to buy a pen." Bhandari, who is traveling and could not be reached for comment, recently signed an out-of-court settlement with Cypress.
No one denies that Rodgers can be an exacting, sometimes brutal boss. Chief Financial Officer Kenneth A. Goldman recalls at least four times in the past two years when Rodgers pounded on the desk in front of him, demanding the near-impossible. In one instance, Goldman was late in creating a new computerized system to track the financial results of each product line, and Rodgers dressed him down in front of several vice-presidents. Rodgers yelled and "got red in the face and huffy and puffy," says Goldman. "Usually when he gets upset about something, though, he's right on."
Valletta Massey, Rodgers' former secretary and now his girlfriend, recalls the time he handed her a stack of hand-drawn charts and asked her to find some software to turn them into slides for a presentation he was about to give. She spent five days with virtually no sleep to shop for the program, learn to use it, and create the slides. Exhausted and on the verge of tears, she finally told him she couldn't complete the project. His response: "I don't need a crybaby, I need a secretary." She got it done. "People think they know how demanding he is," says Massey. "They don't." Since January, Rodgers has fired three other secretaries. Cypress says its annual turnover rate is about 12% a year--not all that bad for any company.
DIVING WITH SHARKS. Rodgers drives himself hardest of all. He goes through his 13-hour business day fueled by little more than black coffee and vitamin C, sometimes making room for a dinner meeting. He takes a single break for a fast three-mile lunchtime run and thinks nothing of sweating through afternoon meetings wearing his running shorts, or greeting visitors with a damp post-run handshake. His schedule of meetings alone--from 8 a.m. to 5 p.m. or 7 p.m. four days a week--would be a full-time job for many people. He settles down in his office every evening for paperwork until about 9 p.m., and takes his "fun reading," letters and trade magazines, home for the evenings.
For all the publicity he generates, Rodgers is described by friends as a private person. He tries to keep weekends free, and is apt to spend them cooking an Italian meal or frequenting the valley's cafes, restaurants, and movie theaters--he's a fan of Arnold Schwarzenegger movies. Rodgers drives himself to the edge even on vacations. When he goes snorkeling in Hawaii, he prefers areas with sharks. "He loves living on the edge of danger," says Massey.
His competitive nature is firmly rooted in his Midwest upbringing. Rodgers was raised in Oshkosh, Wis., the elder son of a car salesman and a schoolteacher. Thurman John got the first names of his father and his grandfather, but even his baby pictures have T. J. scribbled on the back. It was Rodgers' mother who sparked his interest in technology. She taught basic electronics to soldiers in World War II, and started teaching her son from her textbooks before he entered kindergarten. He tinkered with appliances and radios through his childhood, surviving several shocks and a few nasty burns.
He was an exceptionally bright student, at 5 ft., 8 in. one of the smallest players on his high school and Dartmouth football teams, and a sharp dresser who "always dated the prettiest girls," recalls his brother James, a dentist in Denver. Still, James adds: "He was always the loner, the superintelligent guy all the way through school." He married his high-school sweetheart but divorced after 15 years.
Armed with an undergraduate degree in physics and chemistry from Dartmouth College, he went on to graduate school at Stanford in 1970. He studied under William B. Shockley, the inventor of the transistor, whom he has described as one of the few people he has met smarter than himself--although he says he never paid any attention to the racist theories Shockley espoused late in life. Shockley's company, Shockley Electronics, folded when his brightest engineers left to found Fairchild Semiconductor Corp., which became the incubator of Silicon Valley, spinning off Intel, National, and AMD--the "dinosaurs" that Rodgers derides.
In 1975, fresh out of Stanford with both a masters' and PhD in electrical engineering, he turned down a job offer from Intel, deciding that CEO Andrew S. Grove was unlikely to give him the freedom to pursue his own projects. Rodgers accepted a job at Advanced Microsystems Inc. (AMI). His project, a special chip technology he invented at Stanford, was a financial failure, and he was fired.
Once again, Rodgers turned down an offer from Intel, instead accepting a position at AMD. From CEO Sanders, Silicon Valley's dapper, flamboyant marketeer, Rodgers learned the importance of image and marketing. AMD was selling a memory chip similar to the unsuccessful one Rodgers had created at AMI, but AMD managed to charge nearly twice the price. In 1983, Rodgers found venture capital for Cypress and took Sanders' marketing vice-president, Lowell Turriff, with him. That inspired a lawsuit from AMD, starting a feud between Rodgers and Sanders that continues to this day.
LOSERS CLUB. Associates at AMD were surprised when Turriff left to co-found Cypress with Rodgers because the two had spent so much time in loud arguments. But Rodgers insists on hiring people as tough as he is. His hiring process is a strict regimen requiring at least 10 interviews, including one called the "pack of wolves" session, in which several people ask the candidates technical questions that they are unlikely to be able to answer just to see how they handle pressure.
Executives who survive that process are willing to stand up to the boss. Manufacturing Vice-President Mark K. Allen's idea of a vacation is to boat down a river in Ecuador in search of the Auca Indians, a seldom-seen and sometimes violent tribe. "Rodgers develops a winning team," confirms former CFO Stanley J. Meresman, who left Cypress to join a computer company. "They have a lot of respect for him."
Rodgers' abrasive attitude has made him a lightning rod in the most heated debate in Silicon Valley: How to save the U. S. semiconductor industry--and by extension, all of high technology--from the onslaught of competition from the Far East. His answer? Do nothing. Sure, he acknowledges some unfair business practices from Japan. But when it comes to direct government aid for his own industry, he's willing to just say no. Companies seeking such help are "losers," he insists, and "the remedies they are proposing are preposterous."
The list of proposals he considers "preposterous" keeps growing. A pet complaint is the five-year-old industry consortium Sematech, partially funded with $100 million a year from the federal government. Its purpose--to conduct joint research into manufacturing techniques that can be parceled out to members--is "like General Motors trying to become more efficient by having a centralized fin-design department." The 1986 Trade Agreement, which punished Japanese companies that were dumping chips in the U. S. below cost by forcing them to raise prices was tantamount to "a jury awarding damages from the victim to the criminal."
Such talk angers many of his competitors concerned about the American chip industry's dangerous slide in market share. The Semiconductor Industry Assn. (SIA) points worriedly to figures showing that the U. S. share of the world's chip business declined to 37% in 1989, from 57% in 1982. Japan, at the same time, increased its share from 33% to 50%. Intel and others argue that the chip business has grown so expensive, with a single big chip plant costing close to $1 billion to build, that they must now emulate the Japanese and pool their resources. That's why 14 chip companies, representing some 90% of the U. S. industry's revenues, joined Sematech.
JUST PLAIN WRONG? But to Rodgers, Sematech is the great American cop-out. He believes most of the U. S. market-share erosion is an illusion created by currency fluctuations between the dollar and the yen. The SIA says only 30% of the decline is attributable to currency. Rodgers also believes improved U. S. quality from the best companies has virtually stopped any real erosion in the last few years. He cites his experience: Last year, Cypress offered to let Toshiba Corp. make its own version of a chip it now buys from Cypress. After evaluating the part, Toshiba said it could make the chip only by building it 30% larger, and with a higher defect rate. "When we called their bluff we found out they didn't have the technology," says Vice-President Allen. A Toshiba official says the chip wouldn't have made economic sense.
Competitors argue that the facts Rodgers uses to support his arguments are sometimes questionable, or just plain wrong. In a recent speech to Congress, for example, Rodgers complained that he had trouble buying a piece of chipmaking equipment from Westech Systems Inc., a company with Sematech contracts, because the consortium had ordered it not to sell the machine to nonmembers for a year. Thomas N. Tucker, president of Westech, denies the charge and has met with Rodgers to tell him he is wrong. "It seems no amount of data or discussions will change his mind," Tucker says. Still, an agreement with Sematech, signed by Tucker and provided to BUSINESS WEEK by Rodgers, does contain a one-year restriction on the sale to nonmembers of technology developed in connection with Sematech contracts.
Rodgers also irritates people because he's so fond of denigrating his adversaries. Says Sanford L. Kane, a former IBM executive who tried to start another chip consortium: "It's one thing to say, 'Here's what I think.' But he says, 'Everyone else is an idiot.' " And William J. Spencer, president of Sematech, says Rodgers is more talk than action. "Cypress has probably appeared in print more because of his bashing of Sematech than because of the success of his company," he says.
'ICONOCLASTIC.' It's true that Cypress--and Rodgers--still have a lot to prove. For one, it's hardly clear that the future in semiconductors belongs to the small and nimble. Rodgers is even betting that his own microprocessor subsidiary can one day overtake Intel. Ross Technology Inc., a Cypress unit headed by former Motorola chip designer Roger M. Ross, is making chips for Sun Microsystems. Rodgers claims Sun's workstations will eventually supplant PCs--and Ross will effectively replace Intel. That's quite a claim, considering Intel's dominant market position and its $3.9 billion revenues. For now, the vast majority of business software is still being developed for Intel-based machines.
One thing is certain: When T. J. Rodgers talks, Silicon Valley listens. "He's a breath of fresh air," says Representative Tom J. Campbell (R-Calif.), the congressman from Silicon Valley who has seen him in action testifying on Capitol Hill on competitiveness issues. "He's bright, iconoclastic, he shakes things up, and he brings things to the debate." And Rodgers may even have sparked some changes at Sematech. For example, it is considering whether to lower its minimum dues to make itself more accessible to smaller chipmakers.
There are lots of less vocal executives who agree with many of Rodgers' views. Says George Gilder of the Hudson Institute and author of a book about Silicon Valley: "He's the spokesman for the silent majority of Silicon Valley." As Cypress grows in stature and the debates on U. S. competitiveness increase, the insistent voice of T. J. Rodgers will be even more difficult to ignore.Richard Brandt in San Jose, Calif.