Beyond Silicon:
HP Researchers Achieve Breakthroughs in Molecular Electronics

Moore's Law has been working for silicon technology for the past 50 years. About every two years, semiconductor performance doubles.

But that could come to an end in about a decade. The laws of physics will make further advances simply impossible. Even if it didn't, Moore's lesser-known Second Law - that the cost of a fabrication facility increases at an even greater rate — could make it financially infeasible.

Enter the Quantum Science Research (QSR) lab in HP Labs, the company's central research facility.

"We're looking for ways to extend Moore's Law another 50 years beyond the limits of silicon," says Stan Williams, HP Fellow and QSR director.

Williams was in Stockholm recently as one of four speakers invited to address a symposium celebrating the 175th anniversary of the Swedish Royal Institute of Technology. He used the occasion to discuss his group's latest work — including three dramatic breakthroughs.

HP announced that it had:

• created the highest density electronically addressable memory reported to date. The laboratory demonstration circuit — a 64-bit memory using molecules as switches — occupies a square micron of space, an area so tiny that more than 1,000 could fit on the end of a single strand of a human hair. The bit density of the device is more than 10 times greater than today's silicon memory chips;
• combined, for the first time, both memory and logic using rewritable, non-volatile molecular-switch devices; and
• fabricated the circuits using an advanced system of manufacturing called nano-imprint lithography — essentially a printing method that allows an entire wafer of circuits to be stamped out quickly and inexpensively from a master.

Chen's method describes how to make a "master" or mold of a chip, using electron beam lithography. The master can then be used to stamp out copies, just like a printing press. (Sort of gives new meaning to HP being in the "printing" business.) The potential is virtually limitless.

"We expect that this method will be much less expensive than the current photolithographic techniques used for creating chips," Williams said.

Chen's patent is the fourth for the group. "We take 'Invent' seriously," Williams said, "and, even though we're engaged in basic research, we also take business seriously. All of these patents are aimed at making this technology affordably manufacturable."

The first products could be a replacement for flash memory or for use in other high-density, portable devices. They could be ready in five to 10 years, says Williams.

Silicon isn't going away, of course. The molecular-scale chips will also be used to supplement traditional silicon structures. Today's chips could become, in effect, the motherboards for the tiny devices.

The type of computing power that could be delivered by molecular-scale electronics could provide devices so tiny that they could be part of the fabric of clothing. And they could be powerful enough to understand ordinary speech. (You might want to be careful about walking down the street, talking to your shirt.)

As part of his invitation to Stockholm, Williams also had dinner with the King of Sweden, Carl XVI Gustaf. He's scheduled to give his presentation in Spain on Tuesday and Germany on Friday.

"Computing efficiency has increased by a factor of about 100 million in the past 40 years," says Williams, "but there appear to be no physical reasons why it can't be improved by a factor of a billion.

"In some sense, you could say that the age of computing hasn't even begun yet."

	Zooming in on the world's highest density electronic memory HP researchers have created a 64-bit laboratory prototype memory in a space one micron square - an area so tiny that more than 1,000 could fit across the end of a single strand of human hair. In this series of pictures, taken with optical and scanning electron microscopes, each image is magnified approximately 10 times more than the previous one. The wafer on which 625 memories and their test structures were simultaneously imprinted. An array of memories with their test connections. A single test structure with the memory, which is still invisible at this magnification, in the center.   Nanowires leading from the test pins to the memory at the intersection of the lines. The crossed-wire structure of the memory. A close-up of a single 64-bit memory. A bit can be stored at each of the intersections of the eight vertical and eight horizontal wires.