Flash back ten years. Apple is a company bleeding money, thought by many to be in a death spiral. Its head of hardware development, a 41-year-old electrical engineer named Jon Rubinstein, is busy cutting projects and people and launching the G3 and iMac computers. He's living in Palo Alto, Calif., with a cat for companionship, working seven days a week, and subsisting on take-out food at the office.

Then came the iPod.

In 2001, on a business trip to Japan, Rubinstein visited Toshiba, where his hosts trotted out a 1.8-inch hard drive they were working on. They thought it was interesting but weren't really sure what they could do with it. Perhaps Jon had some ideas?

You know those hokey cartoons where the lightbulb goes off over somebody's head, indicating a brilliant idea? That was Jon, sitting there in that Toshiba office. ”I'm, like, ’Yeah, I know exactly what to do with that ,' ” he recalls.

Some four months before, Steve Jobs, his boss, told him to check out the possibility of designing an Apple music player. ”He told me to see what it would take to do it and see if I think we can do it,” Rubinstein recalls. ”My first reaction was ’Hey, we're really busy right now. I don't need more on my plate.' ” But, of course, that's not the sort of thing you say to Steve Jobs. So what Rubinstein ­actually said was, ”Okay, I'll go take a look at it.” He did the research and decided that a music player just didn't make sense for Apple at the time, because the device would have to be too big and clunky to be appealing or it simply wouldn't hold enough songs to be useful. He shelved the idea indefinitely.

Then he saw the Toshiba hard drive. The lightbulb went on. And the rest is consumer electronics history.

Suddenly, he was running a crash project to develop a little music player that became the iPod, which revolutionized the music industry, pushed Apple into the Fortune 200, launched a thousand business-school case histories, and will be described in books (or whatever succeeds them) a century from now.

These days, Rubinstein lives in a San Francisco high-rise with his wife of six years and two dogs. (The cat died.) He takes the train to Sunnyvale, where, as executive chairman of the struggling smart-phone maker Palm, he's tackling another turnaround. If all goes well, he'll soon be competing more or less head-to-head with his former mentor—Jobs.

But not everything has changed. Rubinstein still pads around the house barefoot, buys his jeans at budget retailer Mervyns, and gripes about the cost of running shoes. His income, place of residence, and age suggest he should be driving a Porsche Carrera or maybe a Mercedes SL 550. But he doesn't even own a car. ”I do drink better wine these days,” he says.

At Palm, Rubinstein is essentially expected to do what Steve Jobs did for Apple. That is, take a once-high-flying company heading for oblivion and, through product innovations, make it a player again. Palm helped create the smart-phone market with the Treo, then lost ground to the BlackBerry and, more recently, the iPhone. Rubinstein's job is to get Palm back into the race. ”Palm can be saved,” says Ken Dulaney, an analyst at Gartner's San Jose, Calif., office. ”They had a great product, though they stuck with it too long. Their name is known.”

But, says Dulaney, the next product is critical: ”If they don't come up with a good piece of hardware next, they'll be crippled.” The rumor mill says that product will be a Web-surfing phone that will blow the iPhone out of the water and that it'll be coming in the first half of 2009.

Rubinstein says he's always been ”a product guy.” It's just that the hardware keeps getting smaller. First, in the early 1980s, after he graduated with a master's in engineering from Cornell, he worked on test methods for computer production at Hewlett-Packard. Then he helped design subsystems for the HP 9836 workstation. In 1986, start-up Ardent Computer, headed by industry legend Gordon Bell, hired Rubinstein to help design a new kind of compact, graphics-oriented supercomputer.

Then, in 1990, Steve Jobs came calling. He wanted Rubinstein to help develop reduced-instruction-set computing workstations at NeXT Computer, the quirky but influential start-up Jobs launched during his years of exile from Apple. Alas, NeXT ran out of money before it could ship anything Rubinstein worked on.

With Jobs's blessing, Rubinstein took some of the technology he had been developing and started a company called Firepower Systems to build PowerPC-based personal computers. However, when IBM killed its own PowerPC line and stopped investing in the market, Firepower's business was pretty much destroyed. Still, he made a tidy profit six months later when he sold Firepower to Motorola. Weary of the high-tech merry-go-round, at age 40 he began what he intended to be an extended vacation.

But in January 1997 Jobs called again. He was selling NeXT to Apple and wanted Rubinstein as senior vice president of hardware engineering. How could he say no to the great persuader? Apple was at a low point, with annual sales of US $9.8 billion, down from more than $11 billion the ­previous year. Its Power Macintosh computers were running tired six-year-old System 7, and it had only 3 percent of the personal computer market, down from over 9 percent in 1993.

Rubinstein and the rest of Jobs's execu­tive team revamped the company's culture, product line, and engineering teams. By mid-1999 the company's computer business was becoming healthy again, although its market share, still hovering around 3 percent, was hardly more than a footnote in the PC marketplace.