According to a New York Times article, Hewlett-Packard is about to announce a breakthrough in adaptable circuitry that will allow circuits to be reconfigured even after a consumer has purchased a product [“H.P. to Report an Advance in Adaptable Circuitry,” by John Markoff, 16 January 2007]. These flexible circuits are one of the first widespread commercial applications of molecular computing research. Markoff writes:
“The technology grows out of an advance in nanocomputing, which involves creating circuitry on a molecular scale and making it interact with today’s silicon wires and transistors. A cellphone using the technology could be wirelessly upgraded to take advantage of improved wireless network standards. Another potential use would be in making ultracheap memory chips, and one early application could be in the ink-jet cartridges which Hewlett-Packard manufactures by the tens of millions. The results of the research, which the company plans to report on Tuesday and will be the subject of an article in the Jan. 24 issue of the British journal Nanotechnology, are the clearest evidence yet that the once highly speculative technology could be commercialized soon. The H.P. researchers are among dozens of groups in the United States and elsewhere who have been pursuing molecular computing for more than a decade. Even as today’s microelectronics industry continues to shrink the size of the wires and switches that make up silicon chips, most engineers believe that sometime in the next decade the microelectronics industry will run up against fundamental limits.”
Markoff notes that unless a way is found to breakthrough the limits of microelectronics, the era of Moore’s law will come to end. Markoff goes on to note:
“That challenge has led a hunt for a new technology in which wires will be no more than several molecules wide and switches will be composed of single atoms. So far many laboratories have fabricated experimental switches and wires on this scale, but little progress has been made on the crucial technical challenge of how to move signals between the world of molecular computing and today’s microelectronic systems. Now the researchers report that they have capitalized on a simple idea proposed by researchers at Stony Brook University in New York. Last year two Stony Brook scientists, Dmitri B. Strukov and Konstantin K. Likharev, proposed a novel way to overlay a mesh of molecular-scale wires, or nanowires, on top of a conventional chip circuit to move data between the two worlds.”
The Hewlett-Packard design will apparently produce a hybrid circuit board that contains a traditional array of transistors made using conventional photolithography techniques, but it will also have a mesh of nanowire-connected switches as proposed by Strukov and Liharev. They anticipate the result will extend the life of Moore’s law by a decade.
“The Hewlett-Packard researchers, who are based in Palo Alto, Calif., have extended the Stony Brook concept and applied it to a class of computer chips known as field programmable gate arrays, or FPGA. FPGA chips are widely used in the computer industry to design prototype circuits that can later be manufactured less expensively. To gain flexibility, the FPGA chips use large numbers of transistors that can be reconfigured into an infinite array of different circuits. Therefore the flexibility entails much higher cost, and the circuits are not routinely used in final products, but rather in development systems. The Stony Brook and H.P. design, however, would make it possible to build FPGA circuits that are one-eighth to one-tenth the scale of today’s commercial chips. Moreover, they would have the advantage of consuming far less power than conventional microchips because the molecular computing switches are nonvolatile — that is, they consume power only when switching from one state to another. Such a breakthrough would allow the flexible FPGA-style chips to be used routinely in consumer products manufactured by the tens of millions. It is this advance that could lead to the ability to modify or upgrade the circuitry of standard consumer electronics products already in use.”
Such chips could bring the world closer to what Tom Barnett calls the “Evernet,” a world where, if one desires, he or she can be connected 24/7/365. I’m not sure people are ready for such connectivity or that manufacturers are either. Most companies make a living selling consumers the latest and greatest rather than simply providing an upgrade to something consumers already own. The exciting prospect is that we will begin to see more and better products that take advantage of decades of research into nanotechnology.