Supercomputing power already fits into birthday cards. The popular musical greetings use chips with roughly the processing capability of the late 1940s' ENIAC, the first operational electronic digital (super) computer that used 18,000 vacuum tubes and required 1,800 sq ft of space. Today's R&D is focused on processors no bigger than a grain of sand. The enabler is nanotechnology: systems for transforming matter, energy, and information based on nanometer-scale components. To reach grain-of-sand status, researchers at Georgia Institute of Technology, Atlanta, have demonstrated binary optical storage using nanoclusters of silver oxide, two to eight atoms each. "These nanomaterials have a remarkable new property: When you shine blue light with a wavelength of less than 520 nanometers onto them, you switch on their ability to fluoresce," Robert M. Dickson, assistant professor of chemistry and biochemistry, wrote in the Jan. 5 issue of Science. "You can then read the fluorescence nondestructively by illuminating the clusters with longer-wavelength light. "We have already demonstrated binary optical storage because we can write fluorescent patterns in which an individual particle is either on or off. But we can imagine being able to write and read more than binary storage. These silver clusters could potentially be very useful optical storage materials because of the potential for writing and reading in parallel and/or storing more than one bit of information per data point." Recent nanotechnology research by IBM Corp. adds more credibility to the grain-of-sand size idea. Its scientists developed a breakthrough transistor technology that could preview how computer chips can be made smaller and faster than is currently possible with silicon, says Tom Theis, director of physical sciences at IBM's Thomas J. Watson Research Center, Yorktown Heights, N.Y. By significantly refining the process of building transistors from carbon nanotubes, the researchers have given important new life to Moore's law (named after Intel Corp. cofounder Gordon Moore, who postulated that the number of transistors on a chip will double every 18 months). The cylinders of carbon atoms (about 10 across) are 1/500th the size of silicon-based transistors. The process bypasses the slow method of manipulating individual nanotubes. IBM succeeded with a process it calls constructive destruction, which allows the scientists to produce only semi-conducting carbon nanotubes with the electrical properties required. The step involves using an electric shockwave to destroy metallic nanotubes, leaving only the semiconducting nanotubes. The problems scientists had faced in using carbon nanotubes as transistors was that all synthetic methods of production yield a mixture of metallic and semiconducting nanotubes that stick together to form ropes or bundles. Only semiconducting nanotubes can be used as transistors, and when they are stuck together the metallic nanotubes overpower the semiconducting nanotubes. Theis notes that the challenge of extending Moore's law in the long run may be less significant than evolving the commercial applications for these new "grains of sand" it enables. "In this century we'll be able to put millions of times more processing power into people's hands for the same amount of money," says Theis, noting, however, that "I don't know what we'll do with that computational power."