.svg){kind=spacer}
Details
- The University of Tokyo and IBM jointly developed the Krylov Quantum Diagonalization (KQD) algorithm, published in Nature Communications in June 2025, to accelerate the discovery of ground states in complex quantum systems.
- This effort was spearheaded by associate professor Nobuyuki Yoshioka from UTokyo and Antonio Mezzacapo, IBM's Principal Research Scientist, leveraging a partnership established in 2020 with a dedicated IBM Quantum System in Kawasaki City since 2021.
- KQD significantly improves upon the variational quantum eigensolver (VQE) by generating Krylov subspaces directly on quantum hardware, boosting precision and scalability for complex calculations, and can be combined with sample-based quantum diagonalization (SQD) to create the SKQD method.
- The algorithm was validated through successful simulation of a 56-site many-body system using IBM's Heron quantum processor, highlighting its readiness for practical hardware applications.
- As of late 2025, the collaboration has produced 64 joint research papers, strengthening both institutions' leadership in quantum algorithm innovation, while IBM eyes delivery of its advanced Quantum Starling system by 2029.
Impact
The introduction of KQD and its SKQD variant marks a major advance in quantum computing, shifting the focus from qubit count to innovative algorithm design as the driver of quantum advantage. This paves the way for real-world applications in chemistry and materials science as soon as the next two years. The IBM-University of Tokyo alliance now sets a high bar for global quantum research partnerships striving for early, practical quantum breakthroughs.
