Towards quantum advantage in many-body physics via quantum-HPC hybrid computation
ORAL · Invited
Abstract
As quantum processors rapidly advance in both scale and quality, tightly integrated workflows that combine quantum computing with high-performance classical computing are emerging as a promising strategy to achieve practical quantum advantage in many-body physics, well before fault-tolerant quantum computing becomes feasible. These hybrid architectures overcome the limitations of each platform by offloading classical intractable subroutines to quantum hardware, while leveraging HPC systems for large-scale preprocessing and postprocessing.
In this talk, I will present recent advances in quantum-classical workflows for simulating quantum many-body systems. Highlights include ground-state calculations of strongly correlated systems via sample-based subspace diagonalization [1,2], and imaginary-time evolution combined with phase estimation as an example of quantum-enhanced matrix product operator (MPO) computation [3]. These demonstrations are enabled through close integration between on-premise quantum processors with flagship supercomputers such as Fugaku. Our results show that these hybrid systems can already tackle problem sizes beyond the reach of conventional exact diagonalization, with accuracies approaching those of state-of-the-art classical methods. These developments point to a new regime of quantum computational science, where quantum-HPC integration not only expands the boundary of classically tractable problems, but also enables practical applications in physics and chemistry through quantum-enhanced capabilities.
[1] J. Robledo-Moreno et al., “Chemistry beyond the scale of exact diagonalization on a quantum-centric supercomputer”, Science Advances 11, eadu9991 (2025).
[2] T. Shirakawa et al., “Closed-loop calculations of electronic structure on a quantum processor and a classical supercomputer at full scale”, arXiv:2511.00224.
[3] Qedma and RIKEN collaboration, in preparation.
In this talk, I will present recent advances in quantum-classical workflows for simulating quantum many-body systems. Highlights include ground-state calculations of strongly correlated systems via sample-based subspace diagonalization [1,2], and imaginary-time evolution combined with phase estimation as an example of quantum-enhanced matrix product operator (MPO) computation [3]. These demonstrations are enabled through close integration between on-premise quantum processors with flagship supercomputers such as Fugaku. Our results show that these hybrid systems can already tackle problem sizes beyond the reach of conventional exact diagonalization, with accuracies approaching those of state-of-the-art classical methods. These developments point to a new regime of quantum computational science, where quantum-HPC integration not only expands the boundary of classically tractable problems, but also enables practical applications in physics and chemistry through quantum-enhanced capabilities.
[1] J. Robledo-Moreno et al., “Chemistry beyond the scale of exact diagonalization on a quantum-centric supercomputer”, Science Advances 11, eadu9991 (2025).
[2] T. Shirakawa et al., “Closed-loop calculations of electronic structure on a quantum processor and a classical supercomputer at full scale”, arXiv:2511.00224.
[3] Qedma and RIKEN collaboration, in preparation.
–
Publication: [1] J. Robledo-Moreno et al., "Chemistry beyond the scale of exact diagonalization on a quantum-centric supercomputer", Science Advances 11, eadu9991 (2025).
[2] T. Shirakawa et al., "Closed-loop calculations of electronic structure on a quantum processor and a classical supercomputer at full scale", arXiv:2511.00224.
[3] Qedma and RIKEN collaboration, in preparation.
Presenters
-
Seiji Yunoki
- RIKEN