Watch stocks you care about
The single, easiest way to keep track of all the stocks that matter...
Your own personalized stock watchlist!
It's a 100% FREE Motley Fool service...
It's a small world, after all -- and it's only getting smaller.
International Business Machines (NYSE: IBM ) just produced the first SRAM chip -- a simple sort of high-speed computer memory -- on a new 22-nanometer manufacturing process. To put that achievement into perspective, you could fit more than three thousand of these minuscule chip components across the width of your average strand of human hair.
It's just a proof of concept at this point, but will translate to full-scale processor production by 2011, according to Big Blue and its research partners at Advanced Micro Devices (NYSE: AMD ) , Toshiba, STMicroelectronics (NYSE: STM ) , and Freescale. Before making that leap, chipmakers will make a stop at 32nm in 2009 or 2010. IBM confidently boasts that "no other company or consortium can match" its high-tech 32nm technology; that's a not-too-subtle jab at semiconductor titan Intel (Nasdaq: INTC ) , which has similar process improvements of its own under development.
These advances are important, because they drive the seemingly neverending progress of faster and more capable processors, lower power draws, and cheaper manufacturing. For now, we're stuck with a mixture of 45nm and 65nm microchips powering everything from Dell's (Nasdaq: DELL ) servers and laptops to Apple's (Nasdaq: AAPL ) computers and handheld gadgets.
We've come a long way from the 800nm chip traces of the 486 era, and we're getting close to the theoretical limits of smallness. My physicist friends tell me they'd be between 5nm and 10nm, thanks to quantum effects, X-ray laser wavelength issues, the size of a silicon atom, and other really cool limitations. That's "cool" in a pocket-protector kind of way, of course, and the semi dudes are already doing an end-run around some of the problems by focusing on multicore processors, rather than just cramming more functionality and performance into each core.
It's way too early to call the race to 22nm today, even with working silicon in the IBM consortium's labs. But keep an eye on this space over the next couple of years, and we'll figure out the next-generation winners together.