Since the days of Edison, the wired world has been carried on the shoulders of the humble electron and the charge it carries.
Every electronic device - from Gameboys to iMacs - is powered by the flow of the negatively charged particle. But now, scientists are turning their attention to another characteristic of this yeoman atomic building block: its spin.
The result is "spintronics" (spin plus electronics). And in the near future, this branch of science could lead to computers without a "save" button, and perhaps even the elimination of disk drives. At some point, spintronics may yield quantum computers millions of times faster than today's fastest supercomputers - and small enough to fit in your pocket.
"It's possible to build a whole new class of electronics based on the spin of the electron," says Stuart Wolf, a program manager at the US Defense Advanced Research Projects Agency (DARPA), the Pentagon's research and development lab. The military hopes to use the technology for weapons and communications systems.
Scientists have long known that spinning electrons create their own magnetic charge, which is oriented either north, "spin up," or south, "spin down." What they didn't know was that this spin could be manipulated - a technique known as "spin polarization."
This newfound property presents an opportunity for computer design that's almost too good to be true. All information stored on computer chips is currently represented either as a "O" or a "1." An electron's "spin up" or "spin down" orientation can serve this same binary function, only in a much smaller package.
Moreover, today's chips only work when current is running through them. Pull the plug and the information on the chips is instantly lost - if it wasn't first saved to the hard drive. Spintronic chips, on the other hand, could have information permanently stored on them, eliminating the intermediary step of transferring information from the hard drive to chips - the process known as booting up. Memory and storage functions could merge, eliminating disk drives entirely and creating instant-on, instant-off computers.
Spin chips are cool, too - literally. As more and more functions are crammed onto current chips, the heat given off by an electron's charge has become a critical design issue. But it takes virtually no energy to flip an electron's spin from up to down. No heat - and the potential for extending laptop battery life is enormous, too.
"Spintronics is bursting out all over the place," says Samuel Bader, a senior physicist at Argonne National Laboratory in Chicago. "Today the focus is on memory, but the possibilities are limitless."
Beyond memory, looming like the Holy Grail, is quantum computing. In the odd world of quantum physics, it's possible for an electron to spin up and down simultaneously. If scientists can learn how to exploit this quality of spin, they say they can create a computer whose power is barely conceivable.
"You could build a computer exponentially faster than today's machines, so what might take a year to solve on today's computers you could solve within half an hour," says David Awschalom, a physicist at the University of California at Santa Barbara, and director of its Center for Spintronics and Quantum Computation, whose pioneering research led to DARPA's involvement in the emerging field.
In a limited sense, spin polarization - the manipulation of an electron's spin - is already being used in today's computers, to store information more efficiently.
When the technique was first discovered in 1988, the computer industry was facing the high-tech equivalent of a brick wall. The tiny information bits on the surface of disk drives had reached a point where they couldn't get any smaller without becoming unreadable and unstable. Spin polarization allowed the march of miniaturization to resume, by making the sensors that read the information many times more sensitive.
In 1998, IBM introduced the first commercial storage product (a large unit with scores of disk drives, typically used by corporations) based on spin polarization. Within a year, the entire industry had adopted the technology.
But storing information is only the beginning, say scientists. A team led by Mr. Wolf - who coined the term "spintronics" - is aiming at incorporating the technology into memory chips, which provide short-term memory for PCs (as opposed to the disk drive, which is a computer's long-term, or permanent memory). Much of that work is already completed, and the research is being transferred to private industry.
Another DARPA program, Spin In Semiconductors (SPINS) is going beyond memory chips and examining the possibility of incorporating spin into logic chips - essentially the brains of a computer, like Intel's Pentium chip - and in a host of optical devices. A third program just now getting under way is called Quantum Information Science and Technology, (QuIST). Its long-range goal is a quantum computer or a quantum communications system.
Although scientists are careful not to exaggerate the promise of spintronics, especially quantum computing, which is thought to be 10 to 20 years away, they aren't ruling anything out. "The good news is that a lot of problems that were thought to be enormous obstacles scientifically don't seem to be there," says Mr. Awshalom. "The future looks very bright."
(c) Copyright 2000. The Christian Science Publishing Society