D-Wave Quantum Inc. was the first commercial company to enter the quantum computing market. The company was founded in 1999 and developed their first commercial product, a 128 qubit quantum annealing processor called the D-Wave One in 2011. The Quantum Computing Report published our first report about them in August 2015 with the announcement of the D-Wave 2X. Their forthcoming quantum processor, called Advantage2, will be D-Wave’s sixth generation processor. They are also currently developing a gate-based fluxonium processor to provide a solution path for some applications that cannot use a quantum annealer. D-Wave is now a public company after completing a SPAC merger in August 2022. In this article we will take a look at their overall strategy, roadmap, and progress, both technically and commercially, to provide an update on the company.
Commercial Progress
We talk about Quantum Advantage often in these pages, but there is actual something even more important which I will call Quantum Production Revenue. After all, governments and venture capitalists are investing billions into quantum companies and they would certainly like to see a return on their investment in the form of commercial revenue. And what they want to see is something more than the occasional runs to demonstrate a Proof-of-Concept (POC) or to test out a new algorithm. They want to see end users using quantum in regular production usage for recurring revenue.
Some theorists will opine that a quantum annealing technology will not be able to surpass the best classical approaches. As expected, D-Wave does not quite agree and points to a recent paper published in Nature titled Quantum critical dynamics in a 5,000-qubit programmable spin glass as evidence that their quantum annealing processor can be significant faster that classical approaches for certain 3D spin glass optimization problems. But, more importantly, enterprise IT managers and corporate data analysts may not always choose a solution on what is theoretically the best, but what solution is the easiest to implement. D-Wave has several pieces of software and services will do this which will give them an advantage over other quantum hardware providers who have yet to do this. Perhaps highly quantum savvy PhDs and Postdocs who can do it themselves won’t need these tools. But for someone working within a commercial organization looking to achieve quick results, these tools can be very helpful.
In a recent presentation, D-Wave has disclosed that they have several dozen customer currently working with to explore using their annealers, 26 are developing proof-of-concepts, and two customers, the Pattison Food Group and Satispay, that have completed the proof-of-concepts and are now starting production. In addition, the company indicated that will have additional customers moving from POC to production in the next two quarters. D-Wave asserts that they are the only quantum company that has customers who have progressed to this level.
Software and Services
One important piece of technology that D-Wave has developed and enhanced over the years include their Solvers. This is software that allows a problem to be inputted at a high level as a quadratic model and the software will compile it to a hardware configuration that can run on the quantum annealer. Their original solver only allowed variables in the Binary Quadratic Model (BQM) to be specified as a 0 or 1. But over the years, the company has extended their solver to handle Discrete Quantum Models (DQM) and later a Constrained Quadratic Model (CQM) which can handle continuous variables as well as constraints. D-Wave has a next generation solver in development that will be released later this year that will provide even more improved performance for a growing set of use cases (see chart below). The company has also continued to improve its Ocean software which provides tools for software development in a hybrid quantum/classical environment. And their Leap Quantum Cloud Service provides real-time and secure cloud access to quantum computers and hybrid solvers
D-Wave positions themselves as a company that can provide practical quantum computing today and has a focus on landing new customers and moving them to production as quickly as possible. To that end, they previously established a Professional Services group that can provide assistance at a different levels to help customers from problem discovery, proof-of-concept, pilot production, and/or achieving full production. As part of D-Wave’s thrust to accelerate customer’s time to market, they are developing a series of POCs and standard offerings that customers may be able to start with and modify for their own use case. This would result in the customers getting into production much faster than having to formulate a problem from scratch. D-Wave’s roadmap for creating vertical specific POCs and standardized offerings is shown below:
One additional thing that is often overlooked when assessing quantum companies is that production users need real-time access to the quantum systems in order to get timely solutions to meet their schedules. It may be OK for a student to have a job wait in a queue for a few hours before their job can run, but that would be unacceptable for most production applications. D-Wave has put on focus on providing this as well as high levels of reliability and availability so that their production users can get their work done without delay.
Sales and Marketing
D-Wave’s management has made some adjustments in their sales and marketing approach in order to increase the company’s business development efforts. The first is to focus on the Manufacturing and Logistics verticals. Many of their current customers working on POCs come from these domains and use cases in these domains are a good fit with the inherent strength of their annealing technology to solve scheduling and optimization related problems. The company is also working to bring in more sales resources who are familiar with these industries and also recruit partners who can develop business for them and support additional enterprise customers.
Another area that will get increased focus from D-Wave is government sales. D-Wave currently has minimal sales to the government sector, but they were recently encouraged to see that the recent National Defense Authorization Act (NDAA) signed in December 2023 specifically calls out quantum annealing as a technology that should be evaluated for defense purposes. They will be pursuing more government sales and grants.
Hardware Technology
Last, but not least, D-Wave is continuing to advance both their annealing and gate-based hardware efforts. The company has been working on a successor to their current Advantage product that they are calling Advantage2. It will have over 7,000 qubits and have significant improvements in connectivity (from 15 to 20), coherence, and performance. They have been building prototype technology demonstrators of increasing size and have recently announced a 1200+ qubit version of Advantage2 that they will be making publicly available. They will also be using this demonstrator to test out some error mitigation technology they recently announced. The next version is a 4800+ qubit Advantage2 demonstrator that will be built in a new and improved production stack for further improvements. Finally, the full 7000 qubit version is expected to be released by 2025. Moving beyond the Advantage2, the company’s roadmap shows an Advantage2 Performance Update annealer in 2026 and an Advantage3 annealer in 2028 that will incorporate additional features.
As we previously reported on these pages, D-Wave also has a development effort to create a gate-based processor. Once that is released, D-Wave would be the only company to provide both annealing and gate-based solutions which they feel will put them at an advantage. Their goal is to develop a superconducting based machine using fluxonium qubits that includes error correction. It is still several years away but the company is continuing to make progress. D-Wave released some encouraging results from a test chip last September. They recognize that a quantum annealer cannot cover all the potential quantum computing applications but adding a gate based machine will allow them to cover all of them. A key point the company has made is that there is a lot of common technology between quantum annealers and gate-based quantum processor. This includes superconducting qubits, fabrication, control electronics, cryogenics, and other things which will give them a head start in their development program.
D-Wave also has a very extensive patent portfolio that covers over 500 granted and pending patents worldwide covering hardware, software, applications, and other related areas. The company estimates that about 60% of the portfolio is applicable to both annealing and gate-based processors.
Conclusion
D-Wave is advancing on several fronts including hardware, software, and customer engagement. The chart below provided by the company shows a summary of some of their efforts. The currently stand alone as the only commercial provider of annealing systems, but there are other groups researching this area. There are also other groups looking at other approaches for efficiently solving optimization problems. But D-Wave has been working on this for a long time and possess a great deal of technology and has engaged with a significant number of customers. Once many of the customers working on POCs finish them and move into production, D-Wave should be able enjoy increases in revenue that will start to make their many years of investments start to pay off.
February 8, 2024