| |
|
|
|
|
|
|
QUANTUM COMPUTING
devices like the ones used by researchers from IBM and Stanford University
operate on completely different principles from the classical computers
you’re using to read this story. Instead of dealing with binary 1’s and
0’s — black-and-white bits of information — they work in the fuzzy, gray
world of subatomic particles, where a quantum bit can be thought of as
representing 1 and 0 at the same time.
The field is of far more than academic interest: Quantum computers hold the
promise of solving types of problems far beyond the capability of current
machines, such as searching through huge databases — and even unlocking
the crypto codes used to protect secret messages and secure online
transactions. That’s why the Pentagon, other government agencies and
corporations like IBM are spending tens of millions of dollars on quantum
research. Way back in 1994, Bell Labs
mathematician Peter Shor showed that quantum computers could break a large
number down into its smaller prime factors much more quickly than
classical computers. That just happens to be the trick used by today’s
most advanced encryption schemes to hide information — and as a result,
“Shor’s algorithm” became the driving force behind the development of
quantum computer techniques. |
|
|
Advertisement
|
|
Instead of conventional electronics, quantum computers use
properties such as the magnetic spin of atomic nuclei to represent quantum bits,
or “qubits.” But it’s devilishly difficult to link together enough qubits
to do meaningful work. The record — first set by Los Alamos researchers
and now equaled by the IBM-Stanford team — is seven qubits.
Seven qubits is just barely enough to calculate the
simplest form of Shor’s algorithm, and the IBM-Stanford team describes how
they did it in Thursday’s issue of the journal Nature.
FIVE TIMES
THREE EQUALS... The first
calculation isn’t exactly earth-shattering: It basically determined that
15 is divisible by the prime numbers 5 and 3. But it’s still a big step
forward, said Nabil Amer, who manages research into quantum computing at
IBM’s Almaden Research Center in San Jose, Calif.
"This is just like the first flight of the
Wright Brothers,” Amer told MSNBC.com. “It didn’t go too far, it didn’t go
too high, but nonetheless it proved that you could fly.”
To make the seven-qubit calculation, the researchers created
a new kind of molecule with seven nuclear spins. Those spins can interact
with each other and be programmed by radio frequency pulses. An estimated
1 quintillion of these molecules were synthesized — that’s a 1 followed by
18 zeroes — and poured into a test tube. |
|
|
A diagram charts a specially
designed molecule that currently ranks as the world's most advanced
quantum computer. The seven "qubits" -- quantum bits that are actually
fluoride and carbon nuclei -- are marked by arrows. The chemical name of
the custom-synthesized molecule is dicarbonylcyclopentadienyl
(perfluorobutadien-2-yl) iron
(C11H5F5O2Fe)
|
|
The
results of the calculation can be read by nuclear magnetic resonance
instruments similar to those found in hospitals, a technology known as NMR
for short. Reading the results is particularly tricky because of a problem
called decoherence: Any outside interaction can destroy the quantum
information. “We were able to model this
decoherence process, predict exactly where it would occur, and use that to
optimize our quantum circuits so we could minimize the errors,” explained
Isaac Chuang, the research team’s leader, who is now an associate
professor at the Massachusetts Institute of Technology.
“You can think of this tool we developed as the beginnings of
a quantum CAD (computer-aided design) tool,” Chuang told
MSNBC.com. THE ROAD AHEAD
|
|
|
|
|
From now on, the goal is to build up the number of linked qubits so
that a future quantum computer can really do the types of jobs that
conventional computers can’t. Unfortunately, the NMR technique is being
pushed close to its theoretical limit: Researchers believe that the
technology can’t be scaled up beyond 10 to 20 qubits, and Amer says it
would take hundreds of qubits to put a quantum-computing system to a
challenging test. “We’ve done all that we
wanted to do with NMR, and now it’s time to move on,” he said. That means
adapting the lessons learned so far to solid-state technology — perhaps
involving exotic technologies such as condensed-matter
systems and high-temperature superconductors, or even good old
semiconductors like silicon. “We have
several possible routes to take,” Amer said. “They all at this point in
time, in my view, look pretty much equally promising. ... The one that
will win will be the one that combines the practicality and
economics.” Because of all the question
marks, Amer and others involved in the field shy away from predicting when
we’ll have quantum computers sitting on our desktops. Thus, it will be a
long time before quantum cyber-crooks have the capability to crack
strongly encrypted credit-card data. In the
meantime, researchers are trying to harness unconventional physics to open
the way for a quantum leap in data security. Amer, for example, is working
on a quantum-crypto card that could be inserted into a computer server to
create a “secure island” for sensitive information. |
|
|
|
|
“It’s relying on physics and a physical
process to ensure that data is secure,” he explained. “And if someone
tries to hack into it — because quantum mechanics tells us that you cannot
copy or clone (quantum) information — the data is rendered
useless.” Amer said such quantum-crypto
add-ons are likely to become the first applications to hit the market,
appearing in the next five to 10 years. “It
may be even sooner,” he said.
|
|
| |