Quantum computing use cases and apps

Omdia organized a seminar exploring quantum computing use cases. Alexander Harrowell, Senior Analyst, Advanced Computing and AI, Omdia, was the moderator. 59 percent of the respondents from a survey were still in a thinking mode. Only 4 percent were actually using the technology in some way. By 2023, respondents expected 5x as many live deployments and first scale-up projects.

Best algorithms, best cases
Marco Magagnini, Global Quantum Practice Leader, Reply, said they were struggling to find some use cases over the last three years. Now, they have some quantum projects in development. According to him quantum computing has key concepts that need radically new hardware and new software. We are looking at the best algorithms to fit the best business cases. There are quantum annealers (D-wave) and universal quantum computers (IBM, Rigetti, IonQ, Google, etc.).

In the first case, there is adiabatic quantum computing, and the system evolves toward the configuration of minimum energy. The second case is the gate model, where qubits are manipulated using quantum logic gates. This is relevant for the market as you can do portfolio, risk, collateral, workforce management, routing, aircraft landing, communication quality, and customer strategy optimization.

Quantum computing is focusing on quantum use case analysis and optimization, quantum ML, quantum security and quantum financial services. We select the best use cases, and introduce the right quantum algorithms. You can achieve long-term quantum supremacy. He showed the accelerator for GPUs or QPUs. There can be microservices on premise or any cloud. It is already deployed in finance, energy, telco, logistics, etc. The company has a quantum roadmap for implementation.

VW tests
David Von Dollen, Lead Data Scientist, Volkswagen, talked about quantum computing apps at VW Group. VW has tested quantum navigation app in real traffic. They also did traffic flow optimization using a quantum annealer.

VW also did a quantum-assisted feature selection for vehicle price prediction modelling. They will be further experimenting the extending of algorithms to other paradigms.

Quantum computing may show some benefit in finding solutions. The problems may include hardness, combinatorial aspects, timing, and problem size. We may discover classical algorithms in future. You need to have collaboration and communication between business entities, R&D, and engineering teams.