The new quantum computing could result in an enormous jump in processing power
GRANT BUCKLER - Special to The Globe and Mail. With files from AP
Thursday, October 14, 2004
It could revolutionize drug research and make the cryptographic codes used to protect today's most sensitive information seem as easy to solve as a child's crossword puzzle. For now, though, researchers just want to get quantum computing past the laboratory stage and -- at least in the private sector -- start making money from it.
Quantum computing is one aspect of the broader field of quantum information research. It applies quantum theory, a set of ideas about how the smallest particles of matter behave, to traditional computing. Whereas conventional computers deal with bits of information with two possible values, zero or one, a quantum bit (qubit) can actually represent a zero and a one at the same time. This increases the computational ability of the computer exponentially -- using a single cubit offers two values, linking two qubits lets it work with four values simultaneously, and so on.
It involves complex theories, but harnessing the phenomenon holds the promise of a mind-boggling leap in computing power.
Quantum computing research today focuses on two main areas. One is simply determining the types of problems quantum computers are good at and figuring out the best methods to solve them.
David DiVincenzo, a researcher at International Business Machines Corp.'s research labs in Yorktown Heights, N.Y., says the field is so new that there are no clear criteria for what problems typically lend themselves to quantum computing. They could include anything from modelling the behaviour of molecules, to deciphering complex cryptographic codes or identifying potential new drugs.
Conventional computers try thousands of possibilities until they find the best answer, which can be time consuming. Some of those types of jobs should be relatively easy for quantum computers to handle, says Michele Mosca, deputy director of the Institute for Quantum Computing at Ontario's University of Waterloo.
The other area of research involves designing actual quantum computers. A tiny market already exists for quantum computing devices, made up mainly of researchers who need quantum "toy computers" to conduct research in the field, says Geordie Rose, president of D-Wave Systems Inc. The Vancouver company hopes its quantum-computing efforts will start producing revenue in the next few years as it moves beyond the lab environment.
Today, a $100,000 supercomputer will still "beat the pants off the most sophisticated quantum computer that's ever been built," Dr. Rose admits. Numerous technical problems must be solved before quantum computers can be built powerful enough to outrun the fastest chip-based machines.
Researchers point out that they aren't expecting to outmuscle traditional supercomputers for some time. Quantum computers that can model molecular behaviour and break the toughest encryption will need to process thousands of qubits, Dr. DiVincenzo says, and the prototypes in labs today can't manage more than about seven. But, he adds, quantum machines as small as 10 qubits could start to prove useful for specialized computing tasks.
Barry Sanders, director of the Alberta Institute for Quantum Information Science at the University of Calgary, agrees, but adds that quantum systems able to challenge the fastest chip-based supercomputers could be as much as 20 years away. In the meantime, he says, "there are other things out there that are working and even commercial."
An example is quantum encryption, which applies quantum principles to encoding information. There are several companies selling quantum cryptography technology, such as MagIQ Technologies Inc. and ID Quantique SA. A group in Europe recently made the first quantum-encrypted bank transaction, and researchers at Harvard University, Boston University and BBN Technologies (funded by the Pentagon), have built a multipoint quantum encryption system on the Internet.
On a quantum network, a laser separates light particles called photons, which are polarized and sent over a fibre-optic link. On the receiving end, the coded photons form a "key" to unscramble encrypted data sent over the network by conventional means. Eavesdropping on the photons by setting up a photo detector disrupts them, making the codes unusable and alerting the network to the snooper.
Like quantum computing, Dr. Sanders says quantum cryptography is expensive and unlikely to replace mainstream systems any time soon. But he says it will be useful where security is crucial, because attempts to intercept or tamper with a message can be detected.
© The Globe and Mail. Republished with permission. All Rights Reserved. No part of this article may be reproduced or republished or redistributed without the prior written consent of the copyright holder.