Fujitsu’s Vancouver office is breaking boundaries with its quantum-inspired tech

By applying quantum ideas to AI, the company is helping organizations curb emissions and discover new drugs

    1 of 3 2 of 3

      There are a number of good reasons why, this week, Japanese IT giant Fujitsu officially opened an AI-focused office in Vancouver. Its welcoming culture, strong developer talent base, and ties to first-rate universities drew the interest of the company’s executives, who hoped to bolster the organization’s staggering number of artificial intelligence patents. But what really tipped the scales for the multinational corporation was the rich pool of companies that populate the Vancouver tech ecosystem.

      North America is one of the biggest AI markets in the world. The region’s investment in the technology is 17 times bigger than that of Japan, meaning that—in Canada alone—a significant percentage of tech companies are currently working on integrating artificial intelligence into their products. But while Toronto might be seen as the Canadian leader for AI—and the more obvious choice for the new Fujitsu HQ—Vancouver had a trump card: it’s home to one of the leading experts in quantum computing, 1QBit.

      In 2017, Fujitsu became the lead investor in 1QBit’s Series B round, which raised $45 million. That cash injection allowed 1QBit to continue pushing towards creating quantum technology: a breakthrough that will completely transform how we solve some of the world’s most complex problems.

      On the surface, quantum technology is relatively simple. Currently, our computers operate by reading bits: the smallest units of measurement on a computer, which have a value of either 1 or 0, like an on/off switch. Quantum computing works by exchanging those bits for qubits. Through a series of complex calculations and conditions, qubits are able to exist as both a 1 or 0, which vastly increases the processing power and speed of computers. Creating the conditions where that can occur is spectacularly difficult, but two local companies—1QBit and D-Wave Systems—are getting close to creating something commercially viable.

      Fujitsu wanted in. Setting up its Vancouver office within spitting distance of the 1QBit headquarters, the two companies worked together to create a quantum-inspired product that Fujitsu believes will transform how companies use AI. They called it the digital annealer.

      “It’s a digital technology that’s actually allowing us to achieve quantum scale and quantum speeds in terms of calculations,” said Dean Prelazzi, the company’s vice president and head of business development and marketing, at a product event. Instead of requiring a genuine quantum state—something that has yet to be replicated satisfactorily outside of a lab—the digital annealer uses conventional bits, he explained, but its technology allows the free exchange of signals across all of those bits. “The digital annealer is a digital technology which is currently achieving similar and even better performance in handling some of the largest kinds of problems.”

      Dean Prelazzi, the company’s vice president and head of business development and marketing, believes quantum-inspired technology can help solve some of the world's most complex issues.
      Twitter

      The unit shines when dealing with a huge amount of variables. Take, for instance, shipping routes. A simple journey from the Port of Vancouver to Yokohama Port in Japan is subject to fluctuations in everything from wind speeds to sea currents and weather changes. Each variable has a significant impact on the trip—how fast the ship should travel, for example, or what line it should take in order to arrive as quickly and efficiently as possible. The digital annealer is able to parse these options, and predict very quickly and accurately which route has the highest probability of producing the optimum result.

      Those results benefit not just the shipping companies, but the environment. At the moment, the world’s largest container ships release about as much pollution as 50 million cars. By following routes that are more effective—generated by a technology that is able to update in seconds—the ships can significantly cut their emissions.

      The digital annealer’s route-finding technologies aren’t limited to ships. Deciding the best directions for cars, too, is a particularly complex problem. To optimize five pairs of start and finish points, a computer would have to consider 10100 routes if it wanted to avoid overlaps between vehicles and traffic jams. The digital annealer is able to solve that problem in seconds: an accolade that has seen its technology being considered by both autonomous car manufacturers and governments as a means to increase efficiency and reduce the carbon footprint of local transit.

      “Why are we so excited about reaching and sort of pushing the limits on computational capabilities? The main reason is because we’re reaching the limits of classical computing,” Prelazzi says. “Moore’s law basically says that every two years, the speed of processors doubles. And we've been experiencing a doubling of processing speed. With that immense computational power, over the past 20 years,…we've gone from big mainframe computers to your cell phone, which today has as much processing power as several rooms for computers 20 to 25 years ago. But we're reaching the limit, so there's a lot of activity happening around the world to pursue quantum calculations—quantum capabilities. We are driving aspects of that through the extraordinary innovation in our digital annealer.”

      The digital annealer runs a program during a test simulation at Fujitsu's Vancouver office
      Kate Wilson

      Another realm that Fujitsu hopes its technology can help is drug discovery. Currently working with Toray, an industrial pharmaceutical manufacturer, the company is using the processing power of the digital annealer to predict how to create new medicines. Most drugs are formed of proteins, and every protein has a different chain of amino acids attached to its backbone, which helps to determine the characteristics of the molecule. To create the most efficient drugs, pharmaceutical companies search for the most stable structures—a task that involves sifting through 10100 possible candidates.

      “10100 is a number that’s larger than the number of particles in the universe,” says Prelazzi. “We’re doing calculations on this kind of scaled problem in minutes and seconds. So we're just scratching the surface with this enormous and innovative technology. This, particularly in the drug development context, really holds tremendous promise for accelerating the pace of pharmaceutical companies around the world to identify molecular structures that have high binding affinity—it’s that binding affinity that you need in order to attack the disease cells. There’s an enormous amount of effort around the world to try and do these kinds of things.”

      By locating its new office in Vancouver, the second-largest AI hub in Canada, the company hopes that it will be in the right spot to continue fixing humanity’s most complex questions.

      “If we’re going to solve the largest problems in the world around climate change, if we're going to solve the diseases of the world, if we're going to push the limits on drug discovery, we need computing power that's generations ahead of where we are now. And we’re there with the digital annealer,” he says.

      Kate Wilson is the Technology Editor at the Georgia Straight. Follow her on Twitter @KateWilsonSays

      Comments