IBM has released what it calls the industry’s first published quantum-centric supercomputing (QCSC) reference architecture, a blueprint that outlines how quantum processors can operate alongside classical computing resources in a unified environment.
The architecture integrates quantum processing units (QPUs) with modern supercomputing environments comprising GPU and CPU clusters, high-speed networking, and shared storage across on-premises systems, the cloud, and research centres.
Three-Layer Architecture
The reference architecture defines three distinct layers. The base layer is the quantum system itself, which includes a classical runtime and one or more interconnected QPUs. The runtime comprises specialised classical accelerators, including FPGAs and ASICs, alongside CPUs whose job is to enable QPU operations from error correction coding to qubit calibration.
A second layer consists of colocated programmable CPUs and GPUs that serve as a testbed for quantum error correction. A separate third layer of cloud or colocated CPUs and GPUs handles the classical workloads that support QPU execution, such as pre- and post-processing.
Unified Workflow
The infrastructure integrates orchestration and open-source frameworks such as Qiskit, the Python-based software development kit for programming quantum computers. This means developers and scientists can access quantum capabilities through tools and workflows they already know.
Paul Smith-Goodson, VP and principal analyst at Moor Insights and Strategy, noted that combining quantum and classical hardware remains a significant challenge. “There is a big difference between quantum hardware and classical hardware, and you have to somehow match those two up,” he said.
IBM contends that the blueprint enables coordinated workflows capable of supporting computationally intensive workloads and algorithm research that no single computing approach can solve on its own, moving quantum computing past the “science curiosity” stage towards practical HPC integration.
