D-Wave ‘the mother of all ships’

Revolution AI

Not even the greatest geniuses in the world could explain quantum computing.

In fact, in the early 1930s, Einstein called quantum mechanics — the basis for quantum computing — “spooky action at a distance.”

Then there’s a famous phrase from the late Nobel Laureate in physics, Richard Feynman: “If you think you understand quantum mechanics, then you don’t understand quantum mechanics.”

That may be so, but the mystery behind quantum has not stopped D-Wave Systems Inc. from making its mark. “In the 1980s it was thought maybe quantum mechanics could be used to build a computer. So people starting coming up with ideas,” says Bo Ewald, president of DWave in Burnaby, B.C.

Two of those people were UBC PhD physics grads Eric Ladizinsky and Geordie Rose, who had happened to take an entrepreneur course before founding D-Wave in 1999. “Since there weren’t a lot of businesses in the field, they created and collected patents around quantum,” Ewald says.

While most of those exploring the concept were looking in the direction of what is called the universal gate model, D-Wave decided to work on a different architecture, called annealing. The two do not necessarily compete, but perform different functions.

In quantum annealing, algorithms quickly search over a space to find a minimum (or “solution”). The technology is best suited for speeding research, modelling or traffic optimization.

Universal gate quantum computing can put basic quantum circuit operations together to create any sequence to run increasingly complex algorithms. (A third model, topological quantum computing, could require decades before it can be commercialized.)

When D-Wave sold its first commercial product to Lockheed Martin about six years ago, Ewald says it marked the first commercial sale of a quantum computer. Google was the second to partner with D-Wave for a system being run by NASA Ames Research Center. “Each gets half of the machine,” Ewald says. “They believed quantum computing had an important future in machine learning.”

Most recently D-Wave has been working with Volkswagen to study traffic congestion in Beijing. “They wanted to see if quantum computing would have applicability to their business, where there are lots of optimization problems.” Another coup is a deal with the Los Alamos National Laboratory in the U.S.

Any quantum computing investment is a long-term prospect, but that has not hindered funding efforts. To date, the company has had more than 10 rounds of funding from the likes of PSP, Goldman Sachs, Bezos Expeditions, DFJ, In-Q-Tel, BDC Capital, GrowthWorks, Harris & Harris Group, International Investment and Underwriting, and Kensington Partners Ltd.

“What we have with DWave is the mother of all ships: that is the hardware capability to unlock the future of AI,” says Jérôme Nycz, executive vice-president, BDC Capital. “We believe D-Wave’s quantum capabilities have put Canada on the map.”

Now, Ewing says, the key to moving forward is getting “more smart people working on apps and on software tools in the areas of AI, machine learning and deep learning.”

To that end, D-Wave has recently launched an initiative with Creative Destruction Lab at University of Toronto’s Rotman School of Management to create a new track focused on quantum machine learning. The intensive one-year program will have an introductory boot camp led by Dr. Peter Wittek, author of Quantum Machine Learning: What Quantum Computing means to Data Mining, with instruction and technical support from DWave experts, and access to a D-Wave technology.

While it is still early days, Ewald believes D-Wave’s head start gives it a leg up if and when quantum hits the mainstream. “So far customers tend to be government and/or research related. Google is the notable exception. But once apps come along that are applicable for other industries, it will all make sense.”

The early start has given D-Wave the experience to be able to adopt other architectures as they evolve. “It may be a decade before a universal gate model machine becomes a marketable product. If that turns out to be true, we will have a 10-year lead in getting actual machines into the field and having customers working on and developing apps.”

Ewald is the first to admit that as an early entrant, DWave faces criticism around its architecture. “There are a lot of spears we tend to get in the chest. But we see them coming and can deal with it. If we can survive all that, we will have a better view of the market, real customers, and relationships with accelerators like Creative Destruction Lab. At the end of day we will have the ability to adapt when we need to.” Bo Ewald of Burnaby, B.C.’s D-Wave which aims to use quantum computing to unlock the future of AI. D-Wave will be sending more of its quantum computers to Google, NASA and the Universities Space Research Association in a multi-year agreement.