IBM outlines new architecture for quantum supercomputing in high-performance environments

IBM has introduced a reference architecture that connects quantum supercomputing with classical high-performance systems to accelerate scientific discovery and complex simulations.

IBM presents first quantum-centric supercomputing blueprint

IBM has published what it calls the industry’s first reference architecture for quantum-centric supercomputing, detailing how quantum processors can be tightly integrated into modern supercomputing environments. The company argues that this unified approach will be essential as quantum hardware moves toward practical applications.

Today, quantum computers are progressing toward useful simulations of complex quantum systems. Moreover, emerging hybrid quantum classical algorithms are already producing meaningful results in areas such as chemistry and materials science, where quantum mechanics plays a central role.

However, their ability to address grand-challenge scientific problems remains constrained. The main obstacle is their separation from existing classical HPC infrastructure, which still depends on manual data transfer and ad hoc coordination between quantum and classical systems.

Integrating quantum, GPU, and CPU resources

To close this gap, IBM proposes an architecture that brings together quantum processors, or QPUs, with GPUs and CPUs across on-premises clusters, national research centers, and cloud platforms. This model is designed so that different computing technologies can collaborate on problems beyond the reach of any single system.

The blueprint creates a unified computing environment that fuses quantum hardware with classical resources, including CPU and GPU clusters, high-speed networking, and shared storage. Moreover, this combination is intended to support intensive workloads along with algorithm development, while making it easier to use quantum processors with GPUs in production-scale workflows.

In practice, the design aims to streamline quantum classical workflow orchestration, so that scientists do not need to manually manage the data movement between processors. That said, the architecture still depends on robust middleware and software abstractions to hide underlying complexity from end users.

Three-phase roadmap for integrated systems

IBM scientists describe a three-phase roadmap toward fully integrated quantum classical systems that can support end-to-end scientific workflows. The first phase focuses on deploying QPU accelerators in HPC environments, where quantum processors operate as specialized accelerators attached to existing supercomputers.

In the second phase, IBM envisions middleware-enabled heterogeneous platforms that abstract system complexity. Moreover, these platforms would allow developers to treat quantum, CPU, and GPU resources as components of a single logical system, rather than isolated machines that must be managed separately.

Ultimately, the third phase aims at fully co-optimized quantum-classical systems designed from the ground up for complete workflows. In this stage, quantum computing and supercomputing will be tightly coupled so that workloads can be dynamically partitioned between quantum and classical resources according to performance and accuracy requirements.

Software stack and developer access

With this foundation, IBM plans to support coordinated workflows that span quantum and classical computing within the same application. The company highlights integrated orchestration and open software frameworks as key components of the architecture.

In particular, IBM points to the Qiskit open software framework as a way for developers and scientists to access quantum capabilities using familiar tools. Moreover, by exposing quantum resources through standard interfaces, IBM expects to broaden adoption in fields like chemistry, materials science, and complex optimization.

The firm argues that, over time, this ecosystem could enable scalable chemistry simulations for quantum supercomputing and other demanding workloads. That said, realizing this vision will depend on continued progress in both quantum hardware and classical infrastructure.

Scientific impact and long-term vision

IBM executives frame this effort as a step toward a new era in supercomputing and quantum computing. According to the company, the goal is not to replace classical machines, but to combine their strengths with those of quantum hardware in a coherent architecture.

“Today’s quantum processors are beginning to tackle the hardest parts of scientific problems—those governed by quantum mechanics in chemistry,” said Jay Gambetta, Director of IBM Research and IBM Fellow. He emphasized that this progress is already visible in early research projects.

“The future lies in quantum supercomputing, where quantum processors work together with classical high-performance computing to solve problems that were previously out of reach. IBM is building the technology and systems that bring this future of computing into reality today,” he stated.

Overall, IBM’s reference architecture aims to provide a clear technical pathway for combining quantum and classical resources, positioning the company at the center of the emerging quantum-centric supercomputing landscape.