D-Wave CEO Alan Baratz discusses how the quantum computing company competes against AI, potential acquisitions, and more

Dr. Alan Baratz just wants to hear about your problems.

The CEO of D-Wave Quantum is aiming to position his firm as the best solution for business optimization challenges.

The quantum computing company reported second-quarter results last week, which included better-than-expected quarterly sales of $3.1 million and a near doubling in bookings compared to the same period last year.

One issue, as outlined in a recent report commissioned by D-Wave, is that all the attention — and hundreds of billions of dollars — paid toward AI capex is effectively sucking the oxygen out of the room when it comes to other technologies like quantum computing.

“We have a heavy focus on trying to penetrate more of the market by bringing on new logos,” he told us. “We simply need to help them understand the value that Quantum can bring to their business, and the way we do that, in part, is by engaging with them and demonstrating to them what we can deliver by spending some time talking, trying to understand the problems that they’re trying to solve, and then maybe coming in and saying, ‘Look, here’s a quick and dirty example of what we can do to deliver value for you and to improve your business operations.’”

We spoke with Baratz following the company’s earnings last week to discuss quantum computing competing against (and working with) AI, his potential M&A plans, a recent technological breakthrough from D-Wave that competitors are itching to get their hands on, and more.

This transcript has been lightly edited for clarity. Emphasis added.

Sherwood News: One line from you that continues to live rent-free in my head is that “most business problems are optimization problems.” To that end, when I look at what Alphabet considers some of its successes in terms of AI applications, a lot of these are on the optimization front, in industrial planning for automotives, fulfillment for retail, and logistics and supply chain for ports. How does quantum computing solve optimization problems differently, better, or more cheaply than AI?

Alan Baratz: First of all, most of the hard optimization problems that businesses need to solve, they are solving today. Otherwise, they wouldn’t be running a business. But they’re so hard that they’re not solving them optimally. They’re using heuristics. They’re simplifying the problem using heuristics to try to come up with what they hope are good enough solutions, and then they’re using those solutions to run their business.

AI is exactly the same. It’s another heuristic that will give them a solution that is hopefully good enough, but typically not the optimal solution. Quantum computing is able to give them much better, if not optimal, solutions, and that’s where the business benefit comes from.

So AI is another heuristic. The question is, will it do better than what they’re doing without AI? Sometimes yes, sometimes no. But it’s typically not going to get them to optimal.

Why is it not going to get them to optimal? Because we know that these problems are so hard that the amount of compute required to get to optimal is beyond what classical can do. So it’s always going to be heuristics. It’s always going to be to try to get what you hope is good enough until you get to quantum. That’s when you’re able to make the jump toward optimality.

Sherwood: You referenced on the earnings call that you’re working on more to come in the AI-quantum integration space. Any hints, any teases as to what you’re working toward there? 

Baratz: Well, it really is about two things. One is the two working together to solve problems better. AI is a really good predictive tool. You might use AI to predict demand for a product set in the future, and then you would use quantum to optimize the supply chain to meet that demand. This is the two working together to solve a problem, each focused on what it’s good at solving.

The second is quantum used to basically improve AI and make it more efficient. Introducing quantum into the AI model training and inference workflow, we think will enable a generation of better models built faster and with much lower energy consumption. The tool kit that we introduced earlier this week is a first step toward actually demonstrating that; that tool kit is all about the first steps in using quantum as a part of training large language models.

Sherwood: On that second part, we’re talking about, potentially, not a large group of companies with incredibly deep pockets and acquisitive histories. Is there a need to lean into partnership with these larger players, many of which have their own gate quantum capabilities in development, for this use case?

Baratz: I think that once we have a complete offering in this space, you are right. This is going to be very interesting to the hyperscalers as a part of their AI compute infrastructure and to the AI companies, companies like OpenAI and so on. We’re not there yet. We’ve got the first steps toward that, but we still have more work to do and we will be delivering more capability through our product set over time.

But there are companies like Japan Tobacco that want to use AI to build their own smaller models for solving very specific problems. In the case of Japan Tobacco, it’s developing new molecules for drugs. And we are working with them today to do that, and they’re seeing improved molecular discovery as a result of introducing our quantum systems into their AI workflow.

Sherwood: Your most recent capital raise announced in June, and spelled out clearly in Q2 earnings, included that you were raising money to pursue strategic acquisitions. What will new acquisitions unlock for D-Wave? What are the capabilities that you are looking to get?

Baratz: I have not announced our acquisition strategy. I simply said that we are developing that strategy.

We now have $800 million in the bank, so we have sufficient resources to pursue acquisitions in a variety of different areas. And we are now developing the strategy by, you know, kind of spending time thinking through what’s most important to us to accelerate growth of the company, to create shareholder value. What are companies out there that might be able to help us with that? And then evaluating those companies to try to put together a prioritization, with respect to what our plan should be in this area.

But to take a stab at answering your question more specifically, the way I think about it is accelerating growth in areas that are important to our business. For example, we are the only company in annealing today. We are very far advanced in that area. We pretty much own all the technology for annealing. There aren’t any other companies out there doing it. So acquisition isn’t really an option there. But for gate model, there are still a lot of hard problems that need to be solved.

We have, like everybody else, a longer rather than shorter road map. There’s good work going on in solving or addressing a lot of these hard technical problems out there in the industry. So there may be an opportunity to accelerate our work in that area by some nonorganic growth.

AI is another area. We don’t have AI domain expertise. We’ve started working on how to leverage our quantum systems in the AI space. We brought an initial set of capabilities to market through the developer tool kit that we announced earlier this week, but I think we could move faster if we had more expertise in that area. Those are just some examples.

Sherwood: Another area where D-Wave has demonstrated leadership is cryogenic control. That’s something that’s not just a choke point in annealing quantum systems, but it’s also in gate model development as well. This seems like a technology that would be incredibly useful for competitors to have. Is there any inbound interest you’ve received from them in terms of paying for access to this technology? Is licensing an opportunity for D-Wave here, and is this something that is indicative of any opportunities for horizontal integration across the industry?

Baratz: The answer to that question is yes, from several companies. There’s been interest in licensing this technology and getting some help from us on how to actually implement it because there’s a lot of trade secrets as well — it’s not just published patents.

But this is something that we are thinking through right now relative to transitioning from bringing quantum systems to market and helping other quantum companies bring quantum systems to market. Basically, being a systems supplier as well as an arms supplier, that’s a different business model and we need to think it through very, very carefully. It can create some challenges with respect to supplying competitors, and not all companies are comfortable with that, either as the supplier or as the purchaser, the recipient of it.

Then there’s also a distraction for us in the sense that we are very focused on bringing the best commercial quantum computers to market as quickly as possible, and we don’t want to lose that focus. So there is interest. We have had conversations. We may choose to do that, but so far we have not chosen to do it.

Go Deeper: D-Wave CEO explains where the US is falling behind the rest of the world on quantum computing