By Tom Rickey
Could the future of computing lie in a glowing red dot? Or perhaps in an ordinary-looking loop of silvery wire?
Either could be mistaken for a discarded fragment of ENIAC, the world's first computer. But, in fact, those dots and loops embody atoms doing a dance between two very different states of matter, pointing the way to an intriguing new way of computing.
In the five decades since ENIAC--a 150-foot behemoth full of wires and lights that used to crank out all of 14 calculations per second--computers have consistently become smaller and faster. Now University engineers and scientists are at the forefront of an effort to take a truly quantum leap forward in computing technology: shrinking a computer processor to perhaps just a few atoms, while boosting its speed to billions of times that of today's fastest computer chip.
This could have important consequences because the chips we use today are produced by a technology whose days are numbered.
In their endless quest to maximize performance, engineers have shown their genius at packing ever more millions of miniaturized chips into ever tinier spaces. Just a few months ago, Intel announced the first home-computer chip that performs one billion calculations per second (an effort, incidentally, toward which Rochester's Eby Friedman contributed crucial research).
So there's not exactly a crisis yet. But how much smaller can we go with our chips? Most engineers expect current integrated-circuit technology to top out in 10, maybe 20 years. Then what?
Of possible solutions, one of the sexiest--if seemingly the most far-fetched--is "quantum computing," a field so riddled with paradoxes that just a few years ago it was regarded as no more than a laboratory curiosity. But now serious-minded agencies like the Department of Defense are sitting up and taking notice. Last year the DOD awarded the University's new Center for Quantum Information Systems a $5 million grant for its investigations, making Rochester home to one of the leading efforts in the nation exploring this admittedly bizarre discipline.
Drawing on two of the University's longstanding engineering strengths-- quantum optics and low-temperature superconducting electronics--the Rochester center includes about a dozen researchers at The Institute of Optics, the Department of Electrical and Computer Engineering, and the Department of Physics and Astronomy. Also collaborating are scientists at Cornell, Stanford, Harvard, and Rutgers universities, and Lucent Technologies.
The whole effort revolves around the eccentric rules of quantum mechanics, a batch of physics principles that describe the behavior of matter at the subatomic level. This quirky field lays out theories and predictions that scientists like to call "counter-intuitive" (that's science-speak for "weird" or "odd"). Among them: that an object can exist in two different places --or in two very opposite states--at one and the same time. The rules open the door to "impossibilities" like time travel or (shades of Star Trek) a transporter device that could beam information or matter across the galaxy faster than the speed of light.
The possibilities are bizarre even to the most intellectually elastic of science fiction fans. But the most shocking aspect? It all appears to be true.
Perhaps the world leader in putting quantum theory through its paces is Rochester physicist Leonard Mandel, who over the last decade has conducted a series of experiments that many describe as "the most elegant ever" in demonstrating the validity of quantum mechanics.
Using a simple laser system, Mandel has verified what Einstein once called "spooky action at a distance"--i.e., that any given action has an effect on some other action in some other place. In this case, Mandel showed how making changes to one laser beam in his laboratory altered the characteristics of another, unrelated beam located elsewhere in the lab.
"Mandel's experiments are some of the cleverest and clearest ever done to prove that quantum mechanics can have this 'action at a distance,' " says Ian Walmsley, professor of optics and the head of the new center. Walmsley affirms that whenever scientists have managed to conjure up ways to test the rules of quantum mechanics, the theory has held up as predicted.
Many of today's ordinary devices, such as CD players and fiber optic cables, already rely on quantum technology, which is based on the fact that everything acts as both a wave and a particle. In everyday life, we're used to things mainly behaving as particles: We see it, we feel it, we know where it is. But on the microscopic scale, wave-particle duality is a way of life, and it opens the door to all kinds of possibilities.
In fact the technology is already being used that way, although in a small way so far. At Los Alamos National Laboratory in New Mexico, engineers are using quantum science to send signals via a laser beam between points two kilometers apart. The key to the set-up is knowing exactly how many photons (individual particles of light) are in the message to start with, and making sure that exactly those same photons arrive at the receiving end--something they're able to do thanks to techniques that Mandel has developed. "It's guaranteed by physics that this information is secure," Walmsley says. "Quantum mechanics provides one of the first security systems that does not depend on trust between two parties." If anyone does try to steal a peek, the photons that arrive are irrevocably altered. For instance, an interference pattern normally present might disappear if the integrity of a communiqué has been breached.
The manipulation of units as small as a single photon has come a long way in the last 10 years, aided by the development of ultrafast lasers that give scientists like Walmsley and Carlos Stroud, another University researcher, the ability to nudge atoms, molecules, and electrons ever so precisely.
Like Mandel, Stroud has used lasers in his laboratory to demonstrate a central tenet of quantum mechanics--that an object can be in two places at once. This rule, called super-position, is the one that's the hardest to swallow and is at the same time so promising for the future of computing. Stroud has, however, accomplished the seemingly impossible by placing an electron at one and the same time on opposite sides of a very large (one micron wide) atom. It's a bit like two volleyballs rolling around the rim of a basketball hoop--only there's really just one "volleyball."
An object like the volleyball that is synchronously in two different modes is said to be in a "Schrodinger's Cat" state, an occurrence that takes its name from a physicist who thought up a quantum-mechanics experiment in which a cat could be both dead and alive at the same time.
"Trying to imagine systems that are in this super-position, like the cat that's both dead and alive, is more than your mind can handle," admits Mark Bocko, an electrical engineer who nonetheless is out to create his own version of super-position.
Bocko and colleagues Marc Feldman and Alan Kadin work not on the small scale--with single atoms or photons as do Stroud and Mandel--but instead with hundreds of trillions of atoms contained in a superconducting material, typically a loop of niobium.
When cooled to near Absolute Zero, the atoms act in sync, allowing current to flow without resistance in one direction or the other. Bocko's team is trying to achieve a super-position with current simultaneously flowing both clockwise and counter-clockwise.
Back in 1996--the early days of quantum computing--the Rochester team laid out a scheme for performing quantum computation using conventional circuit technology. That paper has become the blueprint for dozens of groups around the world that have since entered the field, and it has helped the Rochester team earn support from the National Security Agency.
The state of super-position that these scientists are working on is at the heart of the promise and power of quantum computing. To understand why, picture how today's computers work.
Conventional devices rely on transistors that are always in one of two states --on or off, represented by a 0 or a 1. The semiconductor materials on which they're based either transmit an electrical signal, or they don't. Every calculation made, every e-mail sent, every photograph manipulated is the culmination of millions or billions of combinations of these two types of signals.
In a quantum computer, however, super-position would allow the device to work out a nearly infinite array of possibilities all at the same time. While a transistor can be in only one of two states at any given moment, an entity like an atom can be in many states simultaneously.
"This is way beyond binary logic," says Stroud, who has shown how the infinite energy states of a single atom could be used to store information. And it's not only atoms that show potential; Walmsley has discovered that even a single photon can store far more information than was once thought possible.
In a quantum computer, such molecular-sized units, known as qubits, would comprise an information currency that can be in two or more different states simultaneously. A qubit might represent both 0 and 1 at the same time, or it might even represent every number from 1 to 100. In practice, of course, depending on the qubit, engineers would need to link at least a few, and perhaps a few thousand, of them to carry out meaningful calculations.
Ultimately, though, to be of any use, the outcome of the calculations must be accessible, and that points up the deal-breaker that currently makes such schemes literally collapse: measurement.
The rules hold that once someone measures a quantum system--as you would in checking for the answer to a given calculation--the system collapses. Not only that: Mandel has proven in the laboratory that if it's even possible to measure the system, no matter that no one actually does so, the system breaks down. In other words, just the possibility of interaction with anything in the environment, even a single atom or photon, denotes "quantum leakage" and brings calculations to a stop.
"It's as if you have a bunch of stuff going on inside this box, and the minute you open the lid and take a look, everything sort of collapses," Bocko says. "But you can do lots of interesting computations before you take that look."
The challenge, then, is to isolate the system completely from environmental interference. Engineers are now working on various schemes to keep the "box" closed, as it were. Bocko's team is trying to sidestep the problem by developing extraordinarily fast circuits that could steal a quick peek and store the information for later analysis, keeping perturbation to a minimum. Another way around the problem is to make a measurement only at the very end of the computation. That's not as simple as it sounds. As Mandel has shown, engineers must maintain the impossibility of ever being able to measure the system, even while secreting away a method to make a last-moment measurement.
"It's difficult to conceive of a way where at some point you wouldn't have to introduce an extra machine to check the system," Walmsley concedes. "You might have to cleverly conceal any checking in such a way that you can't accidentally look at the result of the computation at the wrong time."
But, perhaps--in seeking a way to overcome their biggest barrier to breakthrough technology in a field full of paradoxes--it's only fitting that scientists have to learn how not to find out information, how not to measure, and, further, how not to make it even possible for measurements to be taken.
Talk about counter-intuitive, or just plain weird. Such are the issues that the pioneers in quantum information technology are trying to answer.
"One way to describe what we're doing," says Stroud, "is that we're developing a quantum toolbox. There's a series of problems in the whole field of information theory, and we want to see how quantum mechanics can provide the tools for real-world applications.
"The excitement is no longer purely theoretical," he adds. "Experiments we were just dreaming of doing a few years ago can now be used for something practical."
Scientists today are torn on the role quantum computers might play in our future. Will each of us have a terahertz computer on the desktop powered by a little loop of superconducting wire, or maybe a tiny vacuum chamber with laser beams shooting within?
Bocko, for one, thinks that this is unlikely. Instead, he envisions developments more like specialized machines for tasks like data encryption, so no one can swipe your credit card number during an Internet transaction, or a special library computer that could provide a way to search the entire Internet for information, instantly.
"The odds are slim that any of this will work, but if it does, then things are never going to be the same," says Bocko. "The payoff would be tremendous."
Can it be done? Perhaps the best answer, given in the true spirit of a quantum-mechanical super-position, is both Yes and No.
Tom Rickey is senior science editor for the Office of University Public Relations.
Maintained by University Public Relations
Maintained by University Public Relations
|©Copyright 1999 — 2002 University of Rochester|