IBM Releases Quantum-Centered Supercomputing Reference Architecture, Ushering in a New Era of Hybrid Computing

On March 12, IBM officially released the industry's first quantum-centric supercomputing reference architecture, providing a technical blueprint for integrating quantum processors (QPUs) into modern supercomputing environments. This architecture aims to solve scientific challenges that are difficult to address using single computing methods, such as chemical simulations, materials science, and optimization problems, by combining quantum computing with classical high-performance computing.

The architecture employs an open and composable design, integrating quantum hardware with classical infrastructure such as CPU/GPU clusters, high-speed networks, and shared storage into a unified computing environment. Shared storage, as the core infrastructure layer, plays a crucial role in coordinating workflows between quantum and classical computing, supporting the efficient exchange of massive amounts of data generated during preprocessing, post-processing, and error mitigation. Through the open-source framework Qiskit and the Quantum Resource Management Interface (QRMI), researchers can seamlessly utilize quantum computing power within existing HPC workflows, enabling automated orchestration and scheduling of hybrid tasks.

This architecture has already been validated in several cutting-edge studies: the Cleveland Clinic used it to simulate a 303-atom tryptophan cage mini-protein; IBM and the RIKEN team, using the Fugaku supercomputer and quantum processor in tandem, completed a large-scale quantum simulation of iron-sulfur clusters; and multiple research institutions collaborated to synthesize the world's first "semi-Möbius molecule" and verify its electronic structure.

IBM stated that it will continue to collaborate with research institutions worldwide to expand this architecture and promote the application of quantum supercomputing in more industries.