Expanding Molecular Dynamics Calculations for Material Development and Drug Discovery to Quantum Computing Platforms

Quemix, a subsidiary of TerraSky specializing in quantum algorithm and software R&D, has announced the development of a foundational technology for executing Molecular Dynamics (MD) simulations on quantum computers, in collaboration with Denso Corporation. MD is an essential computational technique in various industrial fields, such as material development and drug discovery, where atomic-scale behavior determines product performance. Conventional MD is known to have limitations in accuracy regarding time evolution and coordinate space scales close to real-world environments. The two companies have devised a 'Quantum-Classical Hybrid MD Framework' that combines quantum and classical computers, successfully achieving low-load and high-precision chemical state prediction. This method paves the way for advanced quantum computing applications in fields requiring precise atomic-level analysis and functional prediction, such as next-generation battery materials, polymer materials, catalysts, and drug discovery. Moving forward, they will continue to refine this technology and explore its industrial applications, looking ahead to the era of Fault-Tolerant Quantum Computing (FTQC).

Through this joint research, the companies successfully reformulated MD into a new format that maintains and evolves the distribution function, which represents the spread of atomic positions and momenta, as a quantum state. In conventional MD, it was necessary to track the motion of numerous atoms over a long period to construct the distribution function and calculate physical quantities using the resulting set of trajectories. In the newly developed method, the distribution function itself can be directly represented and manipulated as a quantum state, allowing for more direct and efficient calculation of physical properties such as diffusion coefficients. Furthermore, they proposed a method to realize constant temperature (NVT) conditions on quantum circuits and constructed protocols to efficiently read out physical properties like diffusion coefficients and vibrational density of states from quantum states. They applied this framework to the chemical state calculation of hydrogen molecules (H2), demonstrating its feasibility on a quantum computer.