(FORTUNE Small Business) – Bill Gross is a Geek's Geek. When he was a kid, he spent more time tinkering with inventions in his parents' Van Nuys, Calif., garage than playing sports. At 15, he built his first solar-powered product, a four-foot-wide, parabola-shaped dish that used the sun's rays to cook hot dogs. He sold plans for his solar hot-dog machine in the back of Popular Science for $4 each and made enough money to put himself through college.

Before long, he had launched a career as one of America's most prolific serial entrepreneurs. The founder of the startup incubator Idealab, Gross has created dozens of companies. Some, such as eToys and Myhome.com, have flopped. Others, including the online guide Citysearch and the digital-photo software firm Picasa, have done well. His biggest score: In 1999 he started pay-per-click software firm Overture with $200,000; he sold it to Yahoo in 2003 for $1.6 billion.

After all the heartburn—a lawsuit, angry investors—and all his successes, Gross, 46, is turning away from the Internet and back to the sun. Though he's still invested in 18 companies in fields ranging from robotics to wireless networking, he has become the CEO of a 30-person firm called Energy Innovations, started in 2001 with $10 million in funding. "I started thinking whether I could go back to the things I did as a kid, but with more scientific resources and more capital," says Gross, " to see if I could solve some of our energy problems." He sees solar power as a solution to our dependence on Middle Eastern oil and all its inherent political and economic costs. Says this crusader: "People have fought over energy for so long—if we could bring low-cost solar energy to the planet, I think it could reduce wars."

Gross figured that if he could dramatically cut the cost of photovoltaic cells—which convert sunlight directly into electricity—he could help transform a fledgling industry into one potentially worth trillions of dollars as people around the world convert to solar power. The problem, as Gross sees it, is that solar cells, which are made of expensive silicon, cost three times as much as traditional sources of power. Big companies that manufacture solar cells, such as BP, Kyocera, and Sharp, have slowly been making progress reducing the cost of photovoltaics, mainly through manufacturing efficiencies, by about 4% a year. At that rate, he estimates, it will take at least ten years for photovoltaics to become economical.

Gross doesn't have the patience to wait that long. His solution: a patented panel of mirrors called the Sunflower, which increases the amount of sun hitting each solar cell the same way turbocharging boosts power in a car engine, making each cell significantly more productive. Today's best solar cells translate only about 20% to 30% of the sunlight that hits them into electricity (see the diagram on page 40). What if, thought Gross, you could increase the amount of sunlight hitting the cell? The Sunflower, an array of 25 mirrors, tracks the sun as it moves across the sky and aims the sunlight directly onto the solar cell, increasing the number of photons hitting it by 25-fold. The trick is that each of the 25 mirrors moves individually, maximizing the light hitting the solar cells.

Gross, who plans initially to sell the Sunflower to owners of commercial buildings, will place six units in test sites in early 2005 and then roll them out in volume from his Pasadena factory by the end of the year. He calculates that the Sunflower—which will retail at roughly $400 each—will cut the cost of solar power in half. In California, which offers generous subsidies, that means solar power would be cheaper than fossil fuels—about 6 cents a kilowatt-hour, vs. 15 cents from a utility. However, in states with cheap electricity that don't offer subsidies, solar would cost roughly 50% more. "It won't be enough to compete with oil yet," says Gross, "but it will be enough to compete with other solar companies and really drive our business up."

Does Gross worry about the competition? Not at all. "The opportunity for solar is so large," he says, "that we don't even think of other solar companies. All of us combined could not make enough of this stuff to satisfy the planet's demand." Not a bad problem for any entrepreneur.