Forget Computers. Here Comes the Sun.

SAN JOSE, Calif. — T. J. Rodgers is surrounded by a sea of silicon wafers on the roof of his company's headquarters in a Silicon Valley industrial park.

No, not the ones that Mr. Rodgers, who founded Cypress Semiconductor in 1982, used to make high-speed computer memories or the newer specialized chips that go into iPods and high-end Mercedes-Benzes. These wafers are soaking up the sun's rays and turning them into electricity.

On the roof, he fusses over the occasional weed that has grown up in the cracks between the panels and speculates about using robots to keep the glass surfaces of the panels clean. But it is nothing like the problems of manufacturing computer chips in superclean rooms.

"You don't have to do too much of anything," he said. "You just wash them down. All you want to do is get the stuff off."

Mr. Rodgers has plenty of motivation to keep an eye on his roof. The growth of his company may soon depend on SunPower, a small subsidiary that employs the six-inch-square silicon wafers to make a more efficient solar cell.

The contrast between the two uses of silicon could not be more pronounced. As it turns out, the fledgling solar-cell industry uses just about as many silicon wafers as the chip industry does, but the resemblance ends there.

Today, solar cells are a tiny niche in the energy business — rapidly expanding to be sure, but without the potential for exponential gains in performance and falling costs that are hallmarks of the computer world.

Indeed, the solar-cell industry is reliant upon government subsidies, to the consternation of Mr. Rodgers, an outspoken libertarian.

"The culture that got built is what I call a grant culture," he said. "They're all pitching to the U.S. government, looking for funding."

Such criticism aside, the subsidies are in place, both in the United States and Europe, and Mr. Rodgers is ideally positioned to capitalize on the government support he has long railed against. "I can make a good profit for my shareholders," he said, "and provide a lot of good eco-stuff to the world as well."

The paradox is that Mr. Rodgers, 58, who has long been a free-market iconoclast, even by the tough-guy standards of the valley's chip industry, may end up striking pay dirt by moving from the cutthroat world of computer processing power to the more sensitive realm of solar power.

At the same time, by marrying the silicon-based technology of computer chip making with the ability to produce photoelectric cells more effectively from the same raw material, he is infusing the solar industry with fresh energy.

"I think T. J. has found a lot of good things in SunPower," said Alan F. Shugart, a Silicon Valley disk-drive industry pioneer and a Cypress board member. "The tail could easily end up wagging the dog."

In Wall Street's eyes, the tail is already in motion.

Cypress owns 85 percent of SunPower, which went public in November. Cypress is valued near $2.5 billion, with its stock trading at $17.24. SunPower's capitalization is about $2.38 billion; since its offering, its stock has risen from $24.42 to a closing high of $44.07 on March 1. This suggests that much of the value of Cypress these days comes from SunPower.

"It appears that Cypress has been able to leverage its manufacturing process expertise," said Louis Gerhardy, a semiconductor analyst at Morgan Stanley. "T. J. found an industry that had been around for 30 years but now is showing new opportunities."

If he succeeds it will be because six years ago, Mr. Rodgers, who has the build of a former athlete to go with his straight blond hair and owlish glasses, decided to play a hunch when he ran into an old Stanford University graduate school mate at a coffee shop.

"How are you doing?" Mr. Rodgers recalled asking.

"Well, I'm about to go out of business," replied Richard Swanson, an electrical engineer who had founded SunPower to make a highly efficient solar-power cell. The company had some success in specialized applications, but with energy prices relatively low in the early 2000's, the consumer market had not developed as he had hoped.

"We've been on the edge, and I can't cut it anymore," Mr. Rodgers recalled Mr. Swanson telling him. He was about to lay off half his work force of 40 people.

"Why don't you have me over and I'll take a look," Mr. Rodgers responded.

At the time, his own chip business was not exactly shining. In the previous three years, he had pushed Cypress into niche markets for the communications industry. While those markets were still growing, the dot-com collapse in 2000 had undermined any hope that Cypress would become a major power in the data communications world.

On impulse, Mr. Rodgers wrote a $750,000 personal check to buy the company and then spent the next 15 months trying to overcome the skepticism of his board that SunPower wasn't "strategically aligned" with the rest of Cypress's business.

A scientist by training, Mr. Rodgers is also a gambler. And while the solar-cell industry appealed to his technical background, it also meshed well with his experience in the semiconductor market, where he has long tried to succeed by finding commodity chip markets that would allow Cypress to eke out a sustainable edge.

"One thing I will say about T. J. is that he has a nose for what's next," said Andrew Kessler, a longtime Silicon Valley financial analyst. "With his communications business he was late; here in solar he's early."

He has also, on occasion, been wrong. In the early 90's Mr. Rodgers bet big on the Sparc microprocessor chip from Sun Microsystems, a technology that has been unable to gain a market beyond Sun's computers.

A fixture on the Silicon Valley scene since he left Advanced Micro Devices to found Cypress Semiconductor in 1982, Mr. Rodgers can swing from bombastic to modest and back again in the space of adjoining sentences. But after three decades of warring with the vicious up and down cycles of the semiconductor industry, he is now a little more rueful about his past pronouncements.

"T. J. used to say that, 'real men have fabs,' " said Ashok Kumar, a financial analyst at Raymond James & Associates. The term fab refers to a semiconductor foundry, the type of manufacturing plant that was once the symbol of Silicon Valley's technology prowess.

But now, with the exception of the very largest chip companies, the valley's semiconductor businesses are generally focused on designing and marketing chips, outsourcing the actual manufacturing.

Mr. Rodgers should get credit for his willingness to rethink his business philosophy, Mr. Kumar said. "He has done a complete about-face," he said, noting that Cypress had found its biggest recent success in a line of chips that translate the user's thumb motion on Apple Computer's iPod MP3 digital music players.

But his ability to admit error in business judgments has done little to dampen Mr. Rodgers's enthusiasm for picking political quarrels with all comers. He is famous in these precincts for his war of words with a Franciscan nun, Sister Doris Gormley, and with the Rev. Jesse Jackson over the issue of social responsibility in the corporate boardroom.

Part of Mr. Rodgers's charm, though, is his unpredictability. Recently, he responded to reports that the Bush administration was using the National Security Agency to turn loose data-mining software on the phone calls and Internet communications of American citizens with an op-ed article in The San Jose Mercury News that might qualify him for membership in the American Civil Liberties Union.

"Our own government," he wrote, "is a bigger threat to our freedom than any possible menace posed by Al Qaeda."

Change the subject back to sources of power, and he again displays seemingly contradictory views.

"I would gladly live next to the Diablo Canyon nuclear plant," he said. "It wouldn't bother me a bit."

Nuclear power will inevitably be part of the nation's energy future, he reasons, but solar will play an increasingly significant role as well.

For Mr. Rodgers, that is the beauty of the six-inch squares of silicon that are colored black to absorb the sun's radiation. SunPower is on track to gain the ability to make about 35 million wafers a year by the end of 2006, enough to produce 100 million watts of solar power annually.

That should give SunPower an important stake in a market that is expanding at a 31 percent compound annual rate.

After years of promise, the market for solar power is finally taking off, with annual demand expected to increase to as much as 2,500 megawatts by the end of 2008, from about 1,000 megawatts now (which is the size of a large nuclear power plant).

Mr. Rodgers argues that his SunPower subsidiary has a crucial advantage over both larger and smaller competitors. While most of the industry has a conversion efficiency of around 14 percent, the SunPower photovoltaic cell will reach 21 percent, a 50 percent advantage that translates into both cost and performance leads for the company.

He will need that performance, because increasing demand will lead to more efficient solar cells from larger competitors like the Japanese manufacturing giant Sharp.

There are other hurdles to overcome as well. Producing 35 million silicon wafers requires more than 700 tons of silicon. Already, the simultaneous booms in the computer chip and solar-cell industries have combined to produce a global supply shortage of crystalline polysilicon, a material that is forged into tubular ingots and then sliced into thin wafers to make both fingernail-size silicon chips and palm-size wafers.

"We have contracts signed for 2006, but yes, we're worried," he said. "We expect the general market will loosen up in 2008, so we've got a couple of years when we've got to wheel and deal to make sure we get it."

But even as Mr. Rodgers has become enthusiastic about one of the environmental movement's favorite technologies, he remains his prickly self, independent of the cause.

He is in the process of building an ambitious vineyard he has named Clos de la Tech in La Honda, a small community in the hills west of Silicon Valley. Local residents have accused him of contaminating their water supply with the runoff from his grapes, which are planted on the steep hillsides overlooking town.

Both sides are now awaiting the results of an environmental study.

"They're hassling me about the water," Mr. Rodgers said. "The good thing about the environmental impact report is that you get professional environmentalists who are trained scientists, who get away from the little girl who gets up in one of the meetings and says she found three dead rabbits in her backyard and wants to know if it was caused by the vineyard."

Correction: April 21, 2006

 

An article in Business Day last Friday about T. J. Rodgers, a Silicon Valley computer chip entrepreneur whose company has put a focus on solar power and who is building a vineyard near La Honda, Calif., referred imprecisely to the views of some residents near the vineyard. They are concerned that proposed development of a particular hillside will allow herbicides and pesticides to run off into a watershed that supplies 85 percent of their drinking water. They did not accuse Mr. Rodgers of having already contaminated the water.