For reference, that's small enough to fit 30,000 transistors on the head of a pin, or 400 on the surface of a human red blood cell.

Why bother with such mini marvels? The logic is simple. The smaller the transistor, the more you can fit on a chip. The more you can fit on a chip, the more data the chip can process.

And the more data it can process, the faster it can power everything from a streaming movie to a copy of Microsoft Word, not to mention the Web browser you're using to read this story.

Moore's Law

Intel and IBM -- whose methods differ subtly -- believe their research will let the chip industry extend Moore's Law another decade. Named after Intel cofounder Gordon Moore, Moore's Law says the number of transistors that will fit on a chip will double roughly every two years, letting chip speeds double in concert.

What did Gordon Moore think of Intel's announcement? He called it the "biggest change in transistor technology" since the 1960s.

IBM was optimistic as well. "Until now, the chip industry was facing a major roadblock in terms of how far we could push current technology," said T.C. Chen, vice president of Science and Technology at IBM Research. "After more than ten years of effort, we now have a way forward."

Nuts and Bolts

How does it work? A transistor is a crucial component on a computer's processor, a tiny switch that turns on and off to represent the ones and zeroes that form the basis of all of modern computing.

Both IBM and Intel are using exotic metals to improve the design of a processor's gates, which turn the transistors on and off, and dielectrics, which insulate the transistors. As a result, the transistors work more efficiently, losing smaller amounts of power as they process data (a problem called "leakage") and throwing off less heat.

And because the new transistors are smaller -- 0.0000018 of an inch -- they can switch on and off more quickly, roughly 300 billion times per second in Intel's case. For a sense of scale, consider that light travels only one-tenth of an inch in the time it takes one of Intel's new transistors to switch on and off, according to the Santa Clara, California-based firm.

Production Plans

Only 10 years ago, transistors that spanned some 250 nanometers were standard. But neither Intel nor IBM plans to spend that much time putting their new technologies to work.

Intel, which has code-named its new chips Penryn, plans to produce next-gen Core 2 Duo, Core 2 Quad, and Xeon chips in the second half of this year, all with Penryn designs. Prototypes have already been tested with Windows Vista and XP, as well as with Mac OS X and Linux.

IBM claims that its new designs can be produced now, with minimal changes to current manufacturing equipment. Intel's archrival AMD will use IBM's technique to produce 45-nanometer chips in 2008.