The demonstration of the “world's first commercial quantum computer” is as ambitious as it is exciting. But where are the data?
Tycho Brahe didn't like to talk about his data. Throughout his life, he kept his meticulously collected, unique astronomical observations under lock and key, hidden from everyone including (or, perhaps, in particular) Johannes Kepler. Only after Tycho's death in 1601 did Kepler gain full access to the data, on which he then based the derivation of his laws of planetary motion, a crucial step towards understanding celestial mechanics.
D-Wave Systems doesn't like to talk about its data either. But the implication of their claims is clear: practical quantum computing — that is, using quantum effects to perform certain computations more efficiently than a classical computer can do. In February 2007, the private company from Burnaby, British Columbia, invited the world to see their 16-qubit quantum computer 'Orion' in action (a press release is available at www.dwavesys.com). According to D-Wave Systems, more than 600 curious minds have attended two demonstration runs, in Mountain View, California, and in Vancouver, to watch Orion matching molecular patterns, devising a seating plan for a banquet (at which the 'guests' were 15 historical characters including Isaac Newton, if not Tycho or Kepler), and solving Sudoku puzzles. In fact, the audiences followed Orion's output only via a web link, while the spooky action happened at safe distance, at D-Wave's headquarters.
Specific data to prove that D-Wave's 'star' can function faster than a classical computer are yet to come. But, in the aftermath of the Orion rollout, there is no indication that the physics community will see scientific or technical reports on the data behind it anytime soon — data that many are eager to see, analyse and discuss. A number of fundamental issues have already been raised — see, for example, Wim van Dam's News & Views article on page 220 of this issue — but discussion without data can be little more than speculation.
D-Wave's approach is unique and refreshing. The company has focused on building a machine, a quantum computer that actually feels like a computer, rather than resembling a laser table or an NMR spectrometer. This ambitious attempt to 'just do it' and build a dedicated quantum computer (even if there is still room for improvement) should be commended. But, perfect as its demonstrations are for inspiring venture capitalists (something Tycho didn't have to worry about), D-Wave must also make an effort to back them up with hard data. Potential investors might not care exactly how or why D-Wave's quantum computer works, and there are now issues of commercial sensitivity to face, but approval from those who are able to judge its scientific merit is certainly preferable to their scepticism. After all, it was the successful theories, proposals and proof-of-principle experiments in the community at large that made 'quantum computing' marketable buzz words.
More insight might emerge through D-Wave's newly launched programme to provide free access to one of their systems to those with promising proposals for its use. It remains to be seen whether this includes academic users, and to what degree results will be published in the formal literature. But information on even the basic characteristics of Orion is needed to promote the discussion of the practical and fundamental aspects of D-Wave's seemingly stellar progress in a fair, objective manner, and to end the speculation about what they really did back in February. Giving the physics community concrete data to work with will be a step from which both sides could profit.