Singapore – IBM has published a reference architecture outlining an approach to combining quantum computing with established high-performance computing systems. The framework outlines how quantum processing units can operate alongside central and graphics processing units across a range of environments, including on-premises infrastructure, research facilities, and cloud platforms, to address computational problems that are difficult to solve with a single method.
The proposed model is designed to support current computational demands while remaining adaptable to future developments. It brings together quantum hardware and classical computing resources, such as processor clusters, networking capabilities, and shared data storage, into a unified system designed to handle complex workloads and support algorithm development.
The architecture also describes coordinated workflows spanning both quantum and classical systems. It incorporates orchestration tools and open software frameworks, including Qiskit, enabling users to access quantum processing capabilities through established programming environments. This approach is intended to facilitate applications in areas such as chemistry, materials science and optimisation.
“Today’s quantum processors are beginning to tackle the hardest parts of scientific problems—those governed by quantum mechanics in chemistry,” Jay Gambetta, director of IBM research and IBM fellow, stated.
“The future lies in quantum-centric supercomputing, where quantum processors work together with classical high-performance computing to solve problems that were previously out of reach.”
Recent collaborative research involving organisations such as the University of Manchester, University of Oxford, ETH Zurich, EPFL and the University of Regensburg has explored the application of this approach.
In one example, researchers reported the creation and validation of a half-Möbius molecule using a quantum-enabled computational framework. Other projects have included molecular simulations conducted with the Cleveland Clinic, investigations into quantum system energy states by teams including the University of Chicago and RIKEN, and large-scale simulations involving integration with the Fugaku supercomputer.
Additional work involving Trinity College Dublin and Algorithmiq has focused on simulation techniques for complex quantum systems using classical resources for error mitigation. These studies collectively indicate ongoing efforts to combine quantum and classical computing methods in scientific research contexts.
Further development of the architecture is expected to involve collaboration across industry and academic institutions. For instance, joint work with Rensselaer Polytechnic Institute is examining methods for coordinating computational tasks across hybrid quantum and classical systems. Continued advances in algorithms and infrastructure are anticipated to support broader applications in scientific and industrial domains.
