LEESBURG, Va., October 5, 2021 — Quantum Computing Inc. (Nasdaq: QUBT), a company focused on bridging classical and quantum computing, today announced research that provides evidence that “domain-wall” encoding — a method for representing information in quantum computers —delivers better performance for discrete optimization problems than other methods. This breakthrough has significant implications for a wide range of real-world challenges, such as the traveling salesperson problem, which requires choosing an optimal solution from an extremely large number of possibilities. QCI Technical Advisor Dr. Nick Chancellor, who developed the method and was part of the research team that demonstrated its efficacy, will present the paper at the D-Wave Qubits 21 conference on October 6 at 3:45 PM ET.
Domain-wall encoding leverages topological defects (when adjoining structures are out of phase) and Ising spin chains (discrete variables that represent magnetic dipole spin moments) to efficiently represent information in quantum computing systems. Dr. Chancellor’s UK team, which also included other QCI experts, proved that the domain-wall method is better than others on annealers, like D-Wave’s, and will also likely excel for gate model computers. For example, in a problem where discrete variables can take three values, domain wall-encoding uses two-thirds as many qubits to solve the problem vs. other methods. It also found results that other encoding techniques missed. Dr. Chancellor will present their findings and discuss the relevance for optimization problems, such as reconciling the distance between cities for a traveling salesperson.
Interestingly, they demonstrated how encoding interacts with the physics of the quantum processor to get better answers, an important and often overlooked consideration. This post on the QCI blog provides a plain English but detailed explanation.
“It is really important to get as much as we can out of early quantum computers, given how common discrete vs. binary problems are in the real world,” said Dr. Nick Chancellor, who is also a research and teaching fellow at Durham University. “This innovation is an important step to advance our capabilities, especially given the value we’ve found using this encoding. It did better in every way we could think of for critical problems.”
“Today’s initial domain-wall encoding innovation increases the size of a problem that a quantum computer can solve by a factor of 30%,” said Rebel Brown, VP, Strategy & Marketing for QCI. “As quantum computers scale the number of qubits they support, we expect this innovation, and others in development at QCI, to significantly accelerate the time-to-viable-solution for production problems using Qatalyst and quantum systems.”
Most real optimization problems involve discrete variables vs. binary decisions. Consider transportation routing, in which a truck can take any of three roads; microchip design, where a component can be placed any of four places; scheduling an event that can happen at any of seven times; or choosing the best of ten locations to build a plant. While the choices are often not binary, classical computers usually are. Discrete-to-binary encodings like the domain-wall are highly useful for solving real problems that demand discrete answers with quantum computers.
A paper describing the study is available on the ArXiv.org preprint server and is also undergoing peer review.
About Quantum Computing Inc.
Quantum Computing Inc. (QCI) (Nasdaq: QUBT) is focused on accelerating the value of quantum computing for real-world business solutions. The company’s flagship product, Qatalyst, is the first software to bridge the power of classical and quantum computing, hiding complexity and empowering SMEs to solve complex computational problems today. QCI’s expert team in finance, computing, security, mathematics and physics has over a century of experience with complex technologies; from leading edge supercomputing innovations to massively parallel programming, to the security that protects nations.
