Computational Exploration of Quantum Many-body Phenomena

We are exploring novel methods in computational physics based on stochastic method such as the Monte Carlo simulation, path-integral representation of quantum fluctuations, information compression by using the singular value decomposition and the tensor network, statistical machine learning, etc. By making full use of these powerful numerical methods, we aim to elucidate various exotic phases, phase transitions, and dynamics specific to quantum many-body systems, from strongly correlated systems such as the spin systems and the Bose-Hubbard model to real materials. We are also researching parallelization methods for leading-edge supercomputers, and developing and releasing open-source software for next-generation physics simulations.

• Development of simulation algorithms for strongly-correlated systems
• Application of machine learning technique to materials science
• Fundamental theory of quantum computer
• Novel state and critical phenomena in strongly correlated systems
• Cooperative phenomena in non-equilibrium and non-steady states
• Development of open-source software for next-generation parallel simulations