D-Wave Systems Inc. became an early leader in the race to develop the world’s most powerful processor by taking a different approach than rival developers of technology called quantum computers. That allowed D-Wave to get to market faster – but limited what problems its machines could solve.
Now, after a challenging stretch, the pioneering Burnaby, B.C., company says it will follow the lead of IBM, Google and others and develop an all-purpose quantum computer using a “gate-model” design similar to those its rivals are pursuing. However, the company said it will still continue advancing its own more specialized approach, dubbed “quantum annealing.” It was this technology which allowed the company to claim to have put the first quantum computer on the market years ago.
“We’ve taken what we’ve learned and built over the past 20 years and developed a quantum platform road map that will further the benefits of annealing quantum computing … while accelerating our ability to expand into other problem classes,” chief executive officer Alan Baratz said in a release.
He told Fortune D-Wave is still “absolutely committed” to developing annealing systems. “But there are important questions that annealing cannot address” in areas such as physics and chemistry equations. “So if we could also bring a gate-model system, we would be the only company in the world” with both.
The move came as a surprise to the industry, Daniel Lidar, director of the Center for Quantum Information Science & Technology at University of Southern California, said in an interview. “I don’t think this is admitting defeat,” Dr. Lidar said of the company’s decision to move to develop gate-model computers. “What I think they’re doing is clever. Their strategy was to scale up as quickly as possible using the focus on annealing … Now I expect with the appropriate modifications they’ll take that know-how and put it to work in gate-model devices. It’s a natural next step.”
D-Wave was cofounded in 1999 by Geordie Rose shortly after he earned his PhD in theoretical physics from the University of British Columbia in Vancouver. Scientists had theorized for years that quantum processors could solve vastly more complex problems than the world’s most powerful computers. But they figured they would take decades to develop. IBM was already trying.
In 2004, Mr. Rose broke from conventional thinking, concluding the approach most were pursuing would be impossible to build effectively. Instead, he embraced emerging theories and set out to make an annealing device.
Classic computer systems are made of bits, or tiny circuits on a microchip that are either open or closed – representing ones and zeros in conventional software.
In a gate-model quantum computer, bits are replaced with entities called qubits. Each qubit holds a value that can simultaneously behave as though it is a one and a zero owing to the weird properties of matter at small scales. If many such qubits are linked together – a huge technical challenge – they can zip through calculations of a sort that would tie up a conventional computer for trillions of years. A gate-model machine that can be made to work at a practical scale would have many potential applications.
In contrast, D-Wave’s annealing system uses qubits that start off with a similar mixed identity, then quickly settle into one pattern or another. D-Wave’s proposition was that it could use this transition to solve a more limited class of “optimization” problems – problems which call for identifying the best solution from a vast number of possibilities.
To build such a machine, D-Wave had to generate temperatures colder than those in deep space inside the machine, to slow the atoms inside D-Wave’s processors and harness their quantum effects. Mr. Rose called it “the hardest engineering project that has ever been attempted in the history of man.”
Mr. Rose, who left the company in 2014, was greeted with skepticism by many in the scientific community. But D-Wave built progressively better and faster versions of its machine, raised US$300-million from the likes of the U.S. Central Intelligence Agency’s venture-capital arm and Amazon founder Jeff Bezos. It sold models to Google, Lockheed Martin, NASA and the Los Alamos National Laboratory in the United States.
D-Wave CEO Vern Brownell hired Mr. Baratz in 2017 to build software systems that worked with its hardware. Mr. Baratz had led Sun Microsystems Inc.’s effort in the 1990s to transform Java from a nascent programming language into the internet’s main software-writing program.
Mr. Baratz led a shift in D-Wave’s strategy, which accelerated after he replaced Mr. Brownell as CEO in early 2020. D-Wave focused on providing online access to its technology and stopped selling its expensive shed-sized machines. Online customers included Volkswagen and biotechnology startups that used D-Wave to solve complex optimization problems such as improving traffic flows and identifying proteins that could become drugs.
D-Wave also completed a costly US$40-million refinancing last year that wiped out most of the value of some long-time investors and cut its valuation to less than US$170-million from US$450-million. This past March, the Government of Canada committed $40-million to support D-Wave’s research and development costs.
The company must still spend years and tens of millions more to develop not only its annealing technology but now the gate-model machines. D-Wave also faces a new wave of rivals, including Toronto’s Xanadu Quantum Technologies Inc., which use different, less costly approaches to develop quantum computers.
With a file from Ivan Semeniuk