Japan Deploys QunaSys Quantum Software on National Supercomputing Platform

Japan's national quantum computing infrastructure has taken a concrete step toward hybrid quantum-classical research workflows. The RIKEN Center for Computational Science (R-CCS) has formally integrated QunaSys QURI SDK Enterprise as the primary quantum software stack for the JHPC-quantum platform, a move that directly connects the country's flagship supercomputer Fugaku and the ROQUO GPU cluster to both on-premises and cloud-based quantum processors.

Hybrid Integration Across Hardware

The JHPC-quantum platform is designed to bridge high-performance classical computing with a range of quantum hardware. The infrastructure includes the Fugaku supercomputer, based on A64FX processors in Kobe, and the ROQUO cluster, which operates 540 NVIDIA GB200 GPUs. On the quantum side, the system links to IBM Quantum System Two (ibm_kobe), Quantinuum System Model H2 (Reimei), and a mix of trapped-ion and superconducting quantum processing units (QPUs). This architecture enables researchers to run hybrid algorithms that require both classical and quantum resources, with job dispatch and simulation managed by unified middleware.

QunaSys QURI SDK Enterprise acts as the central software layer, providing native libraries for hybrid quantum algorithms. Notably, it incorporates proprietary Quantum Selected Configuration Interaction (QSCI) and ADAPT-QSCI routines, which are tailored for electronic structure calculations in materials science, quantum chemistry, and computer-aided engineering. The SDK's distributed simulation capabilities leverage mpiQulacs for parallel circuit simulation across Fugaku's nodes, while NVIDIA cuQuantum accelerates state-vector and tensor-network simulations on ROQUO's GPU hardware.

Technical Evidence and Performance

Operationally, the platform supports multi-node distributed simulation and hybrid variational loops, allowing researchers to offload classical simulation tasks to Fugaku's A64FX nodes and ROQUO's 135 NVL4 nodes (540 Blackwell GPUs). The QURI SDK Enterprise stack enables workflow compilation for both superconducting and trapped-ion QPUs, eliminating the need for backend-specific code rewrites. This cross-platform portability is intended to streamline algorithm development and benchmarking across different quantum hardware architectures.

Access to the JHPC-quantum platform is currently limited to selected academic and industrial research groups through the Test User Program. The integration follows the operational launch of the ROQUO GPU cluster in June 2026, under the framework of NEDO's "Project for Research and Development of Enhanced Infrastructures for Post 5G Information and Communications Systems." The platform is managed jointly by RIKEN R-CCS and NEDO, with funding from the Ministry of Economy, Trade and Industry (METI).

Comparison With Other National Efforts

Japan's approach to hybrid quantum-classical infrastructure reflects a broader trend among national research programs. The integration of QunaSys QURI SDK Enterprise is positioned as a response to the need for scalable, reproducible workflows that can bridge the gap between classical simulation and quantum computation. Similar efforts have been reported in Europe, where institutions such as EPFL have connected trapped-ion quantum processors to supercomputing platforms, as seen in this earlier breakdown. However, the Japanese platform's emphasis on unified middleware and cross-platform portability distinguishes it from more fragmented or hardware-specific deployments.

Despite the technical ambition, the current system remains a testbed rather than a production environment. The platform's effectiveness will depend on the ability to manage noise, error rates, and calibration drift across heterogeneous hardware, as well as the practical utility of hybrid algorithms for real-world scientific and engineering problems. No claims of quantum advantage or fault-tolerant operation have been made, and the platform's performance will require ongoing benchmarking against state-of-the-art classical methods.

Engineering and Scalability Challenges

While the integration of QunaSys QURI SDK Enterprise provides a unified software interface, significant engineering challenges remain. Hybrid workflows must coordinate job scheduling, data transfer, and error mitigation across both classical and quantum resources. The diversity of hardware backends-ranging from IBM's superconducting qubits to Quantinuum's trapped-ion systems-introduces variability in gate fidelity, coherence times, and connectivity. Achieving reproducible results across these platforms will require careful calibration and robust error tracking.

Scalability is constrained not only by the number of available qubits but also by the overhead of classical simulation and the limitations of current quantum hardware. The platform's distributed simulation capabilities are designed to push the boundaries of what can be classically emulated, but the transition to quantum advantage for practical tasks remains unproven. The Test User Program will provide early feedback on workflow bottlenecks and algorithmic performance, but the path to routine, large-scale hybrid computation is still defined by engineering realities rather than marketing timelines.

Japan's decision to standardize on a single quantum software stack for its national platform is a tactical move to reduce fragmentation and accelerate workflow development. However, the real test will be whether this integration can deliver reproducible, scientifically valuable results that justify the investment in hybrid quantum-classical infrastructure. Until error rates, calibration stability, and cross-platform compatibility are demonstrated at scale, the promise of hybrid quantum computing will remain a work in progress.

Hybrid quantum-classical computing refers to workflows that combine conventional high-performance computing with quantum processors. In these systems, classical computers handle tasks such as data preprocessing, optimization, and simulation, while quantum processors are used for specific subroutines that may benefit from quantum parallelism or entanglement. The effectiveness of hybrid algorithms depends on the quality of the quantum hardware, the efficiency of the software stack, and the ability to manage errors and noise across both domains. Achieving practical utility requires not only high-fidelity quantum operations but also seamless integration with classical infrastructure and reproducible benchmarking against the best available classical methods.