A new study assessing the D-Wave Two, an updated version of the first commercial quantum computer, suggests that the device's apparent failure to outperform traditional computers could be attributable in part to reliance on inadequate benchmarks.
SFI Omidyar Fellow Ruben Andrist is part of a team led by Helmut Katzgraber at Texas A&M University that is trying to understand how to best benchmark D-Wave's quantum chip.
"It seems that current benchmarks are like pitting world class athletes in a discipline not challenging enough to demonstrate their true potential," says Andrist, who compares the traditional benchmarks to racing against world class skier Bode Miller on a beginner's slope where he probably won't excel.
In a paper published in the arXiv earlier this month, Katzgraber's team shows that because of the processor's particular design, the careful selection of hard optimization problems is crucial to allow the D-Wave Machine to demonstrate its potential. The team further proposed alternative benchmarks more likely, they say, to result in a matched comparison.
Katzgraber advocates for additional testing and better benchmark design prior to proclaiming either defeat or victory for the D-Wave Two machine.
Read the paper in the arXiv (January 12, 2014)
Read the news release from Texas A&M university (January 30, 2014)