D-Wave bets that government will lead the way for quantum computing
- By Mark Hoover
- Jan 05, 2017
In October, the first company to build a commercial quantum computer set up a government subsidiary to expand its work with the federal government.
While the subsidiary is new, a focus on the government market is not. D-Wave has worked with U.S. government customers and other organizations related in some shape or form to the government.
D-Wave International of Burnaby, British Columbia, first launched its commercial quantum computing product five years ago and has since acquired three major customers, with Lockheed Martin being the first, said Bo Ewald, president, D-Wave International.
In addition to Lockheed, Google and the Los Alamos National Laboratory became customers. What was interesting about the Google relationship, Ewald said, is that Google and the NASA Ames Research Center program—whose offices Ewald said are right next door—created a partnership where NASA Ames shares the use of the quantum computer. Both use the product for machine learning research.
“Even as we got the business started, there was a huge focus on government,” Ewald said. “It’s only natural that as we move this business forward, that we would expect the U.S. government to be the largest customer.”
René Copeland, former director of government sales, was tapped to be president of D-Wave Government Inc.
The company opens its new government subsidiary at a time of political tension in the United States, but Ewald does not foresee many hiccups in the company’s business due to the presidential transition, referencing the strong momentum for high performance computing that is coming out of the Obama administration.
“The names of some of the programs might change, the players might change, but the reality of needing this type of advanced computing and technology to be able to solve problems of national interest is not going away, so we’re expecting to move forward quite crisply,” Ewald said.
One of the main differences between a quantum computer and a regular computer, Ewald said, is that a quantum computer is probabilistic—not deterministic. Whereas with a deterministic, conventional computer, you could run the program and get the same answer every time, a probabilistic, quantum computer aims to give a probable approximation.
In one example, Ewald explained a program that tries to determine the lowest valley the Alps. “If you ran it 100 times, 87 of the answers would be in the lowest valley,” he said. Three more of those answers would be in an adjacent valley, a few more in a further—but still nearby—valley, and the remainder would be wild answers.
This has potential applications across government, Ewald said, listing the Defense Department, intelligence community, logistics organizations and the Energy Department as potential areas of opportunity.
“In the longer term, as these machines turn out to be more useful in production-oriented applications, then you would see perhaps help with weather forecasting or global climate modeling,” Ewald said, who expects early customers will include various laboratories belonging to different agencies, both defense and civilian.
Lockheed Martin sees the potential applications that quantum computing could yield in the government space, which was the reason the company purchased D-Wave's chip in 2010, said Dr. Kristen Pudenz, senior quantum applications engineer at Lockheed Martin.
One such application is verification and validation of classical software. "The impact of verification and validation of software on our business as a whole is huge," Pudenz said. "Our devices keep getting more complex, and we want to make them more capable and more intelligent, so we write more and more software for them, and the cost of verifying that software grows exponentially with the number of lines of code that we put onto the system."
Even a small improvement from quantum technology could have a great impact on the bottom line of many of the company's products and processes, she added.
Quantum computing also lends itself to general-purpose optimization—Pudenz offered up the example of logistics and the allocation of resources to different tasks—and machine learning.
Right now, the company's work with quantum computing is largely experimental, but Lockheed Martin is focusing on application development as it experiments with the technology. The company has established a quantum computing lab for research at the University of Southern California that is built around a D-Wave computer.
Mark Hoover is a senior staff writer with Washington Technology. You can contact him at firstname.lastname@example.org, or connect with him on Twitter at @mhooverWT.