Flash memory chips, which store data in cell phones, digital cameras, and music players, didn't exactly arrive in a flash. They were invented more than two decades ago by Fujio Masuoka, a midlevel factory manager at Toshiba (TOSBF ). He was supposed to be working on chips called DRAMs (for dynamic random-access memory), which became the workhorses of the personal-computer revolution. But such semiconductors retain data only when the computer is switched on. That wasn't good enough for Masuoka. His heart was set on memory chips that would preserve the user's data whether the power was on or off, thus eliminating the need for fragile hard-disk drives.
Few engineers had faith in Masuoka's ideas at the time. But the intervening years have provided sweet vindication to the circuit designer, who is now 62 and a professor at Tohoku University in the northern Japanese city of Sendai. Global demand for flash memory is poised to hit $25 billion this year, according to research from CLSA Asia-Pacific Markets, thanks mostly to fast-proliferating handheld gadgets. And on Mar. 21, Samsung Electronics unveiled the world's first 32-gigabyte drive built entirely of flash. This drive should lead to a new breed of futuristic -- albeit expensive -- notebooks. They'll be shock-resistant; they should boot up in seconds rather than minutes; and they could stay powered-up 50% longer than other notebook PCs between battery charges.
Nobody anticipated such a development in the early 1980s, when Masuoka began bucking conventional wisdom on data storage. "Simply put, I wanted to make a chip that could one day replace all other memory technologies on the market," he explains. To pull it off, Masuoka drew four other engineers into an ad hoc team and came up with a blueprint. In 1981 he filed a patent for EEPROM -- short for electrically erasable, programmable read-only memory. One of Masuoka's colleagues simplified that to flash memory, because the transistors' ability to erase data in a split second reminded him of a camera flash.
Despite Samsung's recent announcement, many industry experts continue to question the long-term implications of Masuoka's invention. "If you're talking about servers or desktop PCs, flash memory will never replace hard-disk drives," says Nam Hyung Kim, an analyst at market research firm iSuppli. "Hard drives are 100 times cheaper."
Price, however, depends on volume, and production of flash memory is poised for an historic growth spurt. Intel (INTC ), the world's biggest chipmaker and an early proponent of one type of flash, has inked a $2.4 billion joint venture with Micron Technology (MU ) to produce such chips for consumer products and for PCs.
Toshiba, meanwhile, ranks as the world's second-largest flash memory maker, after Samsung. Masuoka says he is glad the company is profiting from his brainchild, but their relations have been strained for many years. In the '80s, Toshiba was slow to realize that it was sitting on a breakthrough, Masuoka says. The truth didn't sink in until Masuoka made a presentation at an international chip conference in San Jose, Calif., in 1984. Four years later, Intel came up with a competing version that was easier to manufacture.
Intel's move finally galvanized Toshiba, and Masuoka was promoted to head one of its advanced research labs. But his bluntness made him enemies in a culture where compromise and consensus are prized. Masuoka left Toshiba in 1994, after a run-in with his boss got him transferred to a position without staff or funds. In 2004 he sued the company for what he considered his rightful cut of its profits on flash: about $9 million. The case is pending in a Tokyo court.
Masuoka now predicts that current versions of flash may reach their technological limits in a few years. But he's convinced that some type of low-cost, 3D chip will eventually chase disk-drive makers out of business. There will always be another barrier, he concedes, but "someone will come up with a breakthrough technology to get around it."
By Kenji Hall