Quantum vs regular computer race ends in draw (Wired UK)


The D-Wave Computer is housed at the USC-Lockheed Martin Quantum Computing Center in Marina del Rey, California

Mae Ryan/WIRED


Advertisement

Advertisement

A speed test between quantum and classical computers has ended
in a draw. New research suggests the commercial quantum computer
sold by Canadian company D-Wave
Systems
isn’t faster than the PC on your desk.

In theory a quantum computer, which uses the quirks of quantum
mechanics to perform calculations, should leave today’s most
powerful machines in the algorithmic dust. A classical computer
encodes data as familiar zeros and ones, known as bits. A quantum
computer, on the other hand, uses subatomic particles known as
qubits that can be zero, one, or a simultaneous superposition of
the two. This should technically allow a quantum computer to test a
huge number of possibilities at the same time. Just how much faster
a quantum computer is able to run through a problem than a
classical computer is known as “quantum speedup.”

Since 2011, D-Wave has been selling a commercial product that
uses a particular architecture known as adiabatic quantum computing
to perform calculations. So far, many
researchers are skeptical
of just how much of an advantage the
product gives over other computers. But earlier this year, a team
at Google and Nasa raced one of D-Wave’s machines against several
off-the-shelf algorithms and found that the quantum product was
about 35,500 times quicker than the classical solvers
.

Not so fast. Google’s test wasn’t exactly a one-to-one
comparison because the off-the-shelf products were not optimised to
solve these particular problems. Another team has also been testing
D-Wave’s machine against regular old computers, running
optimisation algorithms on both that fairly matched their
abilities.


The heart of Nasa’s D-Wave computer

Nasa Ames / John Hardman


“We found no evidence of quantum speedup,” said physicist Matthias
Troyer
of the Swiss Federal Institute of Technology in Zurich,
Switzerland, co-author of a paper that appeared today in Science. “When we looked at all
problems, the machine worked the same as a classical computer.”

Though the team found some instances where the D-Wave machine
could solve five times faster than an ordinary PC, they also
discovered certain problems where the quantum computer was about
100 times slower. Troyer said that it was impressive that D-Wave’s
brand new technology was able to keep up with modern computer
chips, which have been honed over many decades of research. But its
quantum speedup abilities remain unproven.

“It’s a solid piece of research that puts a big question mark on
whether or not the D-Wave approach will give us a tremendous
speedup,” said computer scientist Wim van Dam from the
University of California, Santa Barbara, who was not involved with
the work.

So is that the end of D-Wave’s speedy powers? Not yet. Quantum
speedup effects could still show up in other types of algorithms
run on D-Wave’s machine. The company thinks that the classes of
algorithms that Troyer and his team used went too easy on the
classical computers.

“This particular result in Science isn’t that significant,” said
computer scientist Colin Williams, D-Wave’s director of business development.
“They need to pick problems that are much harder.”

Williams pointed to another researcher, computer scientist Itay Hen of the University
of Southern California, who has done more recent work that seems to
show particular specialised problems where the D-Wave machine
computed faster than a classical computer. Hen presented some of
his results last week at the Third Workshop in
Adiabatic Quantum Computing
in Los Angeles. But he says his
work is still fairly preliminary.


Thomas Porostocky


“I would say the D-Wave guys are, at least for now, too
optimistic,” Hen said.

D-Wave is still banking on other research that might vindicate
their machine. The Google/Nasa team has yet to publish their full
results, which Williams said could appear in the not-too-distant
future.

For now, researchers will continue to search for classes of
problems that show a quantum speedup on D-Wave’s machine. Troyer
said that no experiment could ever rule out the existence of such
problems, but it remains to be seen if they will only be specially
tailored instances or if they could have real-world applications.
D-Wave’s architecture is particularly suited to optimisation
problems, such as finding the quickest route between a set of
points. If it showed an advantage over classical techniques, it
might be a boon to businesses that need such information, such as
airlines, delivery companies, or those making mapping software.

Advertisement

Advertisement

This article originally appeared on Wired.com

If the article suppose to have a video or a photo gallery and it does not appear on your screen, please Click Here

20 June 2014 | 10:34 am – Source: wired.co.uk
———————————————————————————————————————

Leave a Reply

Your email address will not be published.