HP to unveil nanotech breakthrough

Researchers are set to announce a milestone in the company's goal to build future generations of smaller, faster and cheaper chips based on "molecular grids."

Michael Kanellos Staff Writer, CNET News.com
Michael Kanellos is editor at large at CNET News.com, where he covers hardware, research and development, start-ups and the tech industry overseas.
Michael Kanellos
3 min read
Hewlett-Packard researchers will unveil a major breakthrough in the field of nanotechnology on Monday in Europe, a milestone in the company's goal to build future generations of smaller, faster and cheaper chips based on "molecular grids."

Molecular grids are the central concept in HP's nanotechnology plans. In HP's vision, layers of molecular strands, laid down in a crisscross fashion like city streets, will form a mesh of tiny, intelligent circuits. This molecular mesh could be sandwiched between layers of ordinary chip wires to act as a communications network or, eventually, used as the foundation for a complete microprocessor.

Corporate research is increasingly focusing on nanotechnology: the science of building computer chips or other devices out of elements measuring 100 nanometers or less. The technology will let manufacturers produce chips and sensors that are far smaller, faster, more energy-efficient and cheaper to make than their present-day counterparts. It will also open up completely new areas of research, as matter behaves differently at this level, researchers have said.

HP, often in conjunction with the University of California at Los Angeles (UCLA), has already accomplished much of the foundational research required to develop chips based on molecular grids. In 1999 and 2000, HP and UCLA announced they had created new molecules that could be switched on and off, laying the groundwork for making molecular circuits. HP has also fabricated strands of these molecules.

In 2001, the two institutions were awarded a patent that addressed one of the key hurdles in using grids: how to connect the molecular wires, which can measure six atoms wide, to the tiny wires found in computer chips, which can be 70 times thicker. Then, in January 2002, HP and UCLA were awarded a patent for forming and managing traffic on these sorts of grids.

To date, HP has demonstrated how a molecular layer can be used to connect standard computer circuits, but it has not shown off more elaborate grids complete with the intersections and signal lights that would be required in practical use.

HP would not comment on the content of the presentation on Monday other than to say that it will involve a "major" breakthrough in molecular electronics. In May, however, R. Stanley Williams, director of Quantum Science Research at HP Labs, said that HP planned on making an announcement in a few months, broadly hinting that the announcement would revolve around forming molecular grids. Williams will be the main speaker at the Monday event.

HP is not alone in its quest. Next week, Intel will partly unveil its plans for nanotechnology at its developer's conference in San Jose.

The nano ball is rolling
The first products of the nanotechnology era will likely emerge without great fanfare. General Motors, for instance, is experimenting with passive nano materials for making new footboards and other auto parts.

Next year, new microprocessors and other computer chips will technically be nanotechnology products because the average size of their internal circuits will measure 90 nanometers. These chips will be faster than current chips, but will largely be made of the same materials and function the same way.

As time goes on though, the funky nature of products with ultrasmall components will emerge. In its labs, IBM has developed a postage stamp-size memory device that can store approximately 25 million textbook pages. If all goes well, such storage devices could be on the market by 2005.

In a decade or so, carbon nanotubes, strings of carbon molecules that act like wires, could begin to be incorporated into microprocessors, gradually replacing copper wires. Not only would this make computer chips faster and more energy-efficient, they would likely be cheaper to manufacture. With current chip manufacturing methods, each transistor has to be precisely laid down through lithography, a time-consuming process that costs billions of dollars. By contrast, carbon circuits will form themselves, with the process being controlled through the laws of physics and chemistry.

"It gives you the ability to conduct electricity as we can today but with smaller conductors," Peter Glaskowsky, editor in chief of the Microprocessor Report , said in an interview earlier this week. "The first guy to figure out how to make these consistently will be a billionaire."

National governments are also funding research. After the United States dedicated $422 million toward nano research in 2001, the Japanese followed with a $410 million project. Europe and the rest of the world have sunk $425 million into nano projects.