Rod Serling would have loved quantum computers. They compute in the Twilight Zone.
Really. Mathematicians call it Hilbert space, named for David Hilbert, a German mathematician. But it is a dimension - actually, countless dimensions - beyond sight and sound. It is a realm of multiple possible realities, described by the equations of quantum mechanics.
Those equations describe the behavior of subatomic particles, which usually aren't in any definite place at any one time. The large objects of everyday life, though, do seem to have definite positions. A kitchen table is always found in the kitchen, not spread out over different rooms of the house.
That's because for large objects, all but one of those possible realities rapidly disappear as the object interacts with its environment.
But someday, if scientists can overcome some serious obstacles, it might be possible to build a device that preserves those multiple realities long enough to use them for computation.
For some tasks, a quantum computer could outperform a supercomputer the way a supercomputer beats counting on your fingers. Until recently, though, nobody knew of any such task that was useful because even if they could compute in numerous realities, quantum computers could still produce only one answer.
But now, a mathematician at AT&T Bell Labs has shown that a quantum computer could perform a task with enormous ramifications - cracking secret codes.
Spies and the military, financial transactions and other commercial communications require secrecy. Today, most codes for such purposes rely on huge numbers, 100 digits long or more, that a sender needs to know to encode a message. The receiver of the message can decode it only with the help of the prime numbers that, when multiplied together, produce the original long code number. Thus breaking up a big number into its primes (a process called factoring) is the key to breaking a secret code.
"If you were given the prime factors you could easily multiply them out to find what the number was," says Umesh Vazirani, of the computer science department at the University of California, Berkeley. "But if you were given the number, you can't easily take it apart into its prime factors."
Given a 2,000-digit code number, for example, all the computers in the world couldn't factor it before the sun burns out, billions of years from now. But one quantum computer could factor the number quickly, using Hilbert space to conduct more computations, simultaneously, than the number of atoms in the universe.
There remains that problem about producing only one answer. That's the problem that Peter Shor of Bell Labs recently solved. Drawing on earlier work by Dr. Vazirani and his student Ethan Bernstein and by Daniel Simon of the University of Montreal, Dr. Shor showed how the one answer a quantum computer produces would be exactly the answer needed to solve the factoring problem.
(The details are technical. But if you're interested, it involves finding the period of a periodic function. The multiple computations in Hilbert space interfere with each other in such a way that the one answer the computer spits out is precisely the period in question. With that number, finding the prime factors is easy.)
So far, this development has not been classified Top Secret, and quantum computing specialists have not been rounded up and quarantined. But that's probably only because the surprising paper proving this new quantum danger to secrecy has not yet been published. It has been circulated among experts via e-mail, though, who have been expressing astonishment and incredulity about the paper's result in their electronic conversations.
Still, financial secrets and embarrassing e-mail are not in any immediate danger of exposure from quantum computers. Technical obstacles are severe. It might take years to build even a small working model.
Nevertheless, it's clear that quantum computation could be important. No doubt other important problems will surface that only quantum computers could solve.
And there is yet a more significant aspect of Dr. Shor's discovery. It provides evidence once again that the most arcane fields of research can suddenly take on surprising practical importance. For years, physicists have discussed the notion of quantum computation, beginning with the late Richard Feynman and Paul Benioff of Argonne National Laboratory. British physicists David Deutsch and Richard Jozsa, among others, have expended vast intellectual effort envisioning the computational potential of Hilbert Space.
All along, these efforts seemed to be meaningless exercises in mathematical mythology. But now the possibility exists that these games of the mind have revealed deep and consequential truths about the real world - including the realization that it is not all as real as everybody has always thought, except, perhaps, fans of The Twilight Zone.
Back to @Man's Homepage