Skip to comments.IBM Gets Small - Big Blue researchers are trying to move quantum computing from theory to reality.
Posted on 12/11/2001 8:06:30 AM PST by Straight Vermonter
Most of us take it for granted that computers will get faster, cheaper, and more powerful every year. The bedrock of the Digital Age is Moore's Law, which observes that microprocessors double in power every 18 months as the transistors that make up microchips become smaller and smaller -- which means that more can be packed onto chips of the same size.
One of the most basic elements of a transistor is a feature called a gate. Today, gates are made up of about 1,000 atoms. If Moore's Law were to continue unabated, in 20 to 40 years those gates would reach the realm where single atoms are used to store and process information -- the physical limits of miniaturization. That would mark the end of both Moore's Law (if something else does not kill it first, as is more likely) and the increases in processing power that have come with it. So researchers are now scrambling to figure out how to make quantum computers, whose fundamental parts work at the atomic level.
There's nothing new about the ideas behind quantum computing.
In 1959, at a conference of the American Physical Society, physicist Richard Feynman described a set of famous challenges that would have to be overcome for nanoscale computing to become a reality. Feynman dared the assembled scientists to produce several technologies that seemed far-fetched at the time. Among them: a microscope that could look at individual atoms, a device that could manipulate those individual atoms, a computer with wires no more than 100 atoms wide, and circuits that could take advantage of the quantized energy levels of individual atoms. The first three were accomplished in the 1980s and 1990s, but Feynman's final challenge has not yet been met.
Quantum computers would manipulate the different quantum states of electrons, such as their energy levels or their nuclear spin, to perform calculations and store data. They could also be designed to operate on atomic nuclei or photons. Quantum computers exist only in theory today, but at IBM's Almaden research lab, set on a stunning hilltop just south of San Jose, scientists are trying to meet Feynman's last challenge. During a recent visit, I met with two scientists, Don Eigler and Bill Risk, who are taking different approaches to building computational structures out of individual atoms and molecules. Their work hints at how quantum computers may one day operate.
Eigler was the first scientist to overcome one of Feynman's other challenges -- to manipulate atoms one by one -- by laying out xenon atoms to spell "I-B-M" in 1989. The ability to precisely place atoms one at a time is the key to understanding the physics of nanoscale structures. Eigler is still moving atoms around. Ponytailed and serene, he walks through the Almaden research halls with his furry, lumbering dog, a Leonburger called Argon. "The name of the game is to make things smaller," he says.
Right now Eigler is trying to figure out how the wave nature of electrons comes into play when you build structures at microscopic levels. State-of-the-art semiconductors have transistors and circuits with elements that are 130 nanometers (0.13 microns) wide. Somewhere below the 100-nanometer range, when these elements become as small as individual molecules and atoms, quantum physics will start to wield its influence.
One of Eigler's experiments involves precisely placing iron atoms one by one in an elliptical ring called a quantum corral. When he places a magnetic cobalt atom on one of the foci of the elliptical ring, an interesting effect occurs: The cobalt atom creates a magnetic disturbance not just in the location where it is placed, but also at the other focus of the ellipse, which is an empty spot across the ring.
This "quantum mirage,"as Eigler has dubbed it, is similar to an acoustic phenomenon familiar to visitors of the U.S. Capitol. If you stand near a particular column inside the dome, you can hear every word whispered at a corresponding spot across the hall. John Quincy Adams used his knowledge of this effect to spy on political opponents; Eigler, on the other hand, thinks the quantum mirage may be a new way to transport information within microchips. Instead of moving electrons around in a linear fashion -- one at a time down millions of crisscrossing wires -- you could change the resonance of an electron in one place and trigger an effect somewhere else.
"We're using the electrons as waves instead of as particles," Eigler explains. "You can always send waves through each other." In other words, using Eigler's approach, multiple channels of information could be transmitted through the same physical space. Chips that take advantage of these quantum properties could eliminate or relieve the contemporary need to pack so many wires onto each chip.
Another Almaden scientist is Bill Risk, the manager of IBM's Quantum Information group. In classical computing, bits are represented by ones or zeroes. There is a one-to-one relationship between the number of bits and the information they can represent. But in quantum computing, bits are represented by the quantum states of electrons, atoms, molecules, or photons. These states are not binary in nature. A quantum bit, or qubit (pronounced "cue-bit"), can simultaneously embody 80 percent of one state and 20 percent of another, for example, or any other combination.
And unlike in classical computing, there is an exponential relationship between the number of qubits and the information they can represent. Taken together, all of this means that a quantum computer could be exponentially more powerful than the most powerful binary machines. For instance, figuring out the prime factors of large numbers -- the basis of modern public-key cryptography -- is a particularly computation-intensive task. The fastest computer in existence today would take 10 billion years to factor a 400-digit number, whereas a quantum computer could theoretically do it in 3 years.
Drafting the blueprints for such a computer is the philosophers' stone of our day. "One approach," Risk suggests, "is to use molecules as individual quantum computers." The benefits of this approach are obvious. "Nature makes every one identical," he notes, "and the atoms that represent quantum bits are precisely positioned and have precisely controlled interactions through chemical bonds."
Risk's group is creating molecular cocktails in the lab to perform quantum computations. One of his former scientists, Isaac Chuang (now at MIT), was the first to demonstrate this type of quantum computer. At IBM, Chuang designed molecules with three, five, and seven atoms exhibiting nuclear spin, each of which represented a qubit. He put a billion billion of these molecules into a solution, aligned their nuclear spins with a magnetic field, exposed them to radio waves, and then changed their spins en masse, which he detected with a nuclear magnetic resonance spectrometer. In this way, all the molecules performed the same calculation simultaneously, and he was able to poll them for the result.
Chuang proved that such quantum calculations can actually be performed. But as Risk acknowledges, "Reality is setting in among experimentalists. The quantum states are very delicate and easily degraded or destroyed by whatever is surrounding the quantum bits."
Figuring out ways to manipulate those states will be the key to unlocking the promise of quantum computing. But for scientists like Eigler and Risk to do that they must continue to think small.
Sorry. No can do. My head exploded after looking at that numeral with the 64 zeroes.
I guess no tinfoil hat could protect you then. Though, the power of medicine would be awesome.
I don't believe man will ever create a machine to compare to the sophistication and complexity of the human brain.
The best novel I've read this year is Michael Crichton's Timeline. It deals with quantum computing, but you don't need to understand it to completely enjoy the book.
But if you see the big picture, then you know those terrorists exist everywhere, not just over there...
Increasing conscious knowledge of existence, through free enterprise business, is the key to everlasting conscious human happiness on planet earth...