By Stephen Baker In about a decade, the semiconductor industry, most agree, will reach a physical limit. Engineers will no longer be able to shrink down circuits on tiny slabs of silicon. To continue doubling computing capacity every 18 months, while cutting costs -- the formula that has powered tech growth for 40 years -- the industry will look to new technology, a new architecture, and perhaps a move beyond silicon.
This will be a crucial transition in the history of computing (see BW, 2/14/05, "The Business Of Nanotech"). And the key question is this: Will the current titans of the Silicon Age, led by the semiconductor industry, maintain their leadership as they attempt to shepherd the industry through this rough passage? Or will the new components and architecture of computing come from outside the silicon economy, from the new forces of nanotechnology?
Put it another way: Will nanotechnology save the chip industry or mount a challenge to it?
ALREADY NANO? For semiconductor execs, the question itself is absurd, even offensive. After all, they have customers, they operate perhaps the world's most highly refined industrial process, and they took in $214 billion in revenue last year. How could a few thousand researchers who are just now piecing together early prototypes of nano transistors ever challenge them?
What's more, no one knows better than the chipmakers how to manufacture at the nano level: The nodes in many of today's chips are only 90 nanometers apart, which technically qualifies these semiconductors as nanotechnology. (Nanotech covers all engineering at a scale between 1 and 100 nanometers. A nanometer is 1 billionth of a meter.)
Here's the counterargument: Nanotechnologists are experimenting with entirely new paradigms of computing. They could conceivably marry electronics, for example, to biology, coming up with self-replicating computational devices. Researchers in Israel have already harnessed transistors to strands of DNA.
"I COULD SPIT." Further, by working at the atomic level, they're operating at a scale far tinier than the chipmakers do. If they can coax these infinitesimal bits to transmit information, the result could transform computing. Steve Jurvetson, a managing director of the Silicon Valley venture firm, Draper Fisher Jurvetson, has invested some $95 million in nano startups. Few people, he says, understand the vast potential of information technology at the atomic scale. And molecules are not only tiny, but plentiful. "There are more molecules in a drop of water than in all the transistors ever built," he says.
Such statements infuriate chip execs. "I could spit on this table, and you'd have even more molecules," says Randy Goodall, an associate director of Sematech International, an Austin (Tex.)-based semiconductor research consortium. "But what's it worth?"
His point: Nanotechnologists and their fans dwell too much on nano's potential, and pooh-pooh the potential of the industry that has innovated continually, managing to turn Gordon Moore's 40-year-old prediction about computing power into a law.
Adds Phaedron Avouris, manager of nanoscale science at IBM's Nanoscale Science & Technology Group (IBM): "Unfortunately, there's a lot of hype in this technology. There's no point in making things smaller if they don't work well."
UNIMAGINABLE POWER. Despite the arguments, a chip industry facing its physical limits needs nano breakthroughs. "I hope that in 10 years we have a breakthrough. I don't know for sure that I expect it. But I know it will be needed," says Avouris.
Over the next decade, in addition to pouring billions into nanotechnology R&D, chipmakers will be scooping up nano startups and using their innovations on the chip. Still, this means that chipmakers will be incorporating nanotechnology into their own industrial process.
But what happens if a more powerful, cost-effective architecture for computing emerges from the nano labs in the next decade? This would create almost unimaginable computing power, systems that could revolutionize energy, medicine, transportation. Computing could narrow its gap with human consciousness.
This advance, exciting as it would be, also would give birth to a melee in the business world. The prize: Control over the next stage of the Information Age. No wonder the arguments get a bit heated. Baker is a senior writer for BusinessWeek in New York