The title says it all. Chipmakers have tried for many years to build 3-D chips.
Turns out chipmakers and big-city developers have a similar problem: They are running out of space. In less than a decade, engineers figure they will have packed all the computing power they can into a microchip's tiny real estate, ending the yearly leaps in processing speed we take for granted. One long-proposed solution: Pull a Donald Trump and build up instead of out. You can pack more tenants in a skyscraper than in a ranch house; why not turn the microchip into a cube to get more computing power in the same space?
Yet no one has succeeded in 30 years of effort to make a commercially viable, three-dimensional chip. Most chipmakers, including Intel, Hewlett-Packard and IBM, have jumped past 3-D in favor of nanotechnology.
Still, two 3-D chipmakers hope to introduce new products in a near future. And they are using very different approaches.
Matrix Semiconductor, which has raised $80 million from Thomson Multimedia, Sony, Kodak, Benchmark Capital and Microsoft, says it has created a 3-D chip it will start selling by year-end.
For three decades engineers at Stanford, Livermore Labs and elsewhere couldn't overcome the obstacles to a 3-D semiconductor. Executives at Matrix say they have solved the problem by emulating the flat-panel display industry, which has figured out how to place the circuits on glass substrates by developing "thin-film" transistors.
Most important, the Matrix design is meant to be manufactured at existing chip plants, saving money and time.
The other technology is the brainchild of Fujio Masuoka, a former Toshiba engineer who was behind the design and the emergence of "flash memory." His 3-D chip design is much more ambitious and expensive than the Matrix one.
Rather than layer the chip circuitry atop one another, he has designed the circuitry itself to stand vertically; he says the design will have a crucial speed advantage.
Here is an illustration of his design.
I'm not sure that any of these technologies can win, with all the big semiconductor companies taking other routes. But it's sure refreshing to see innovation at work.
Source: Benjamin Fulford, Forbes.com, July 22, 2002
5:41:45 PM Permalink