By Anne Stuart, Oct. 2005
Light is the solution. It’s also the problem.
That’s the paradox HP Labs' Quantum Information
Processing Group is beginning to unravel with its research
into optical quantum computing.
The group has been investigating ways to use photons,
or light particles, for information processing, rather
than the electrons used in digital electronic computers
today. Their work holds promise for someday developing
faster, more powerful and more secure computer networks.
“Quantum processing can attack problems we can’t
attack with conventional computers,” says Tim Spiller,
the HP Distinguished Scientist who is leading the research. “Even
a small quantum computer has the potential to enhance communications and
information processing.”
Today's computers work by manipulating bits that exist
as either 0s or 1s. What makes quantum computing so powerful
is that quantum bits (or qubits) have an infinite choice
of values, meaning they can potentially perform multiple
operations simultaneously. For example, a quantum computer
could efficiently factor large numbers that today's, or
even tomorrow’s, conventional machines might never
be able to crack.
Scientists have believed for years that high-speed light
is the best candidate for moving quantum information from
place to place. However, previous designs for quantum computing
using light have been extremely inefficient, and so completely
impractical for actual technology.
“Light is very good for communication because the
bits of lights don’t talk to each other,” Spiller
explains. “You can send light over long distances – for
instance, with optic fibers – and it preserves its
state pretty well. It doesn’t communicate with other
bits of light -- or with much else, either. That’s
why you can have many different conversations going on
at the same time in the same telephone cable and they don’t
interfere with each other.”
Therein lies the problem. “To do any kind of data
processing, the bits of data need to be able to interact,” as
they do in today’s computer systems, Spiller says. “So
on the face of it, light isn’t good for information
processing because the bits of light don’t talk to
each other. We need a process to get pieces of light at
the quantum level to talk to each other.”
That’s exactly what his team -- including Principal
Research Scientist Bill Munro, other HP Labs researchers
and Kae Nemoto, an associate professor of quantum information
sciences at the Tokyo-based National Institute of Informatics
-- has spent two years trying to develop.
Their starting point was solving yet another puzzle: how
to detect photons, or individual chunks of light, without
absorbing or damaging them. Typically, detecting a single
photon requires letting it "smack into" something,
such as a piece of semiconductor material, Spiller says. “That
creates a lot of electrons and holes, and the piece of
light is lost. So you can detect it, but in the process
of doing so, it’s destroyed.”
Now the team has developed a method for both detecting
photons without appearing to harm them and for allowing the bits
of light to communicate with each other by having the photons
'talk' with one another via a probe light signal. The photon
leaves an imprint of itself on the probe light signal without
being damaged in the process, allowing the researchers
to detect it without apparently demolishing it.
“You can measure the probe light signal and look
for the photon’s imprint,” Spiller explains. “If
you see it, it’s there; if you don’t, it isn’t.”
Equally important: the probe light signal lets photons
interact, albeit indirectly. “If you do a certain
type of measurement on the probe after it’s talked
to two photons, you’ll find that although they don’t
talk directly to each other, the photons have interacted
because they both talk to the probe,” Spiller says.
Theoretically, the researchers say, there’s no limit
to the number of photons that can interact this way. For
that reason, their work represents a quantum step toward
creating a scalable method for optical quantum computing.
“The nice thing about this is it’s moving
toward having the best of both worlds,” Spiller says. “The
best communication is done with light. If you can also
compute with light, you can do everything with light." As
a result, there is no need to convert quantum information
from some type of electronic format to the new optical
one and back.
The HP Quantum Information Research effort circles the
globe, with researchers in Bristol, Palo Alto and Tokyo
(Nemoto). Their efforts, which could fundamentally change
the way computers and people work, have caught the attention
of scientists worldwide. Many call the team’s early
findings important and promising, but warn that researchers
are years away from being able to build a real optical
quantum computing system.
Spiller readily concurs. “Our vision is long-term,
and we’re starting small,” he says.“
We hope to have some experimental results in the next couple
of years. Once we’ve got that started—once
we know for sure how one photon can talk to one beam --
we can go forward to build quantum processors.”
Quantum computing has the potential to revolutionize information
technology. Small quantum processors containing just a
few qubits could be used to stretch out the distances over
which secure quantum communications work, analogous to
the way that conventional optical repeaters are used to
amplify ordinary optical communications. Such small processors
may also enable new sensing and measurement technology.
Comparably small (in terms of qubit number), but distributed,
quantum processors could enable new protocols such as secure
quantum auctions between separated parties, or quantum
voting.
Mid-sized quantum processors (50-100 qubits) could be
used as research tools, allowing simulation of quantum
systems that currently cannot be performed on even the
most powerful supercomputers. Large quantum computers (with
tens of thousands of qubits) will likely be able to search more
quickly than conventional technology, and factor very large
numbers efficiently, leading to quantum code-breaking.
Other potential applications are likely to evolve over
time. Munro points out that transistors were initially
used in hearing aids.
“At that time, you only had
big valves made of glass so you couldn’t have a small
integrated circuit. Then somebody put a few components
on a small device used for hearing aids,” he says. “The
people who did that didn’t foresee that down the
road you’d have millions of transistors, or integrated
circuits, on pieces of silicon, running devices of all
types and sizes."
Spiller
says he expects that quantum computing will enhance, rather than replace, the current
standard.
“We believe quantum computing will grow
alongside conventional computing. You might have a quantum
processor sitting next to your conventional machine,” he
says. “It’s not that quantum technology will
sweep everything else away. Instead, it will enable new
things.”
Anne Stuart is a Boston-based freelance journalist who has written
about business, technology, and the Internet for more than a
decade.
|
Printable version
