Japan’s Quantum Leap: RIKEN and SoftBank Select 21 Teams for Hybrid Quantum–Supercomputer Link Project
21 organizations and research teams have been chosen to take part in Japan’s “JHPC-quantum” test-user program, according to a statement released by RIKEN and SoftBank. This project is a significant step in turning hybrid quantum computing from an intriguing lab experiment into a useful tool for business and education. The main goal of the project is to smoothly integrate quantum computers with the Fugaku supercomputer, Japan’s premier high-performance computing (HPC) equipment.
Through the utilization of actual workloads against the combined quantum-classical system, the “JHPC-quantum” test-user program is designed to enable external teams to investigate and improve hybrid quantum–HPC applications. The goal of this large project is to tackle the crucial “last mile” issues of hybrid systems, such as data transportation, error propagation, resource scheduling, workflow orchestration, and debugging.
You can also read WiMi Quantum Computing Advances AI with QDCNN Research
The Near-Term Value of Hybrid Architectures
Japan’s national quantum policy has continuously underlined that hybrid designs that closely combine quantum accelerators with traditional HPC resources will provide value in the near future rather than isolated quantum systems. This strategy supports the commonly held belief in the quantum community that a basic division of labor is necessary for the majority of industrial applications. According to this concept, classical systems manage crucial functions including large-scale data processing, modelling, optimization, and control, whereas quantum routines are in charge of resolving particular, computationally demanding subproblems.
Stakeholders emphasize that hybrid quantum–HPC is the best feasible course of action in the medium term, not a short-term compromise. Currently, pure quantum machines are not well adapted to handle complex, large-scale problem cases on their own due to severe restrictions in qubit count, fidelity, and coherence. Therefore, it will be necessary for industry-relevant workflows to integrate quantum modules, such as simulation or optimization kernels, into reliable conventional pipelines that handle control logic and bulk processing.
You can also read IonQ Algorithmic Qubits 64 Record Marks in Quantum Advantage
Accessing Fugaku and Two Quantum Processors
Participants in the test program have access to an extensive collection of tools created to support these interconnected applications. Among these resources are:
- The backbone of classical computing is the Fugaku supercomputer.
- The ability to experiment with quantum kernels being handled as callable subroutines is made possible by two quantum processors.
- RIKEN, SoftBank, and other partners developed a collaborative software platform that manages work orchestration across the two different computing layers.
- APIs, resource managers, and data interfaces are examples of essential operational tools that facilitate software prototyping that combines classical and quantum operations.
Instead of needing custom, one-time integrations, RIKEN and SoftBank hope to eventually turn quantum resources into “callable accelerators” within current HPC procedures. Achieving this objective could significantly quicken the transition from scholarly algorithm research to broad corporate implementation.
Rigorous Selection and Key Participants
The 21 candidates were chosen using a phased admission procedure by RIKEN and SoftBank. Formal application windows were held in April 2025 and again in June–July 2025, after an initial trial solicitation that took place from December 2024 to January 2025. A diverse range of industries, including materials, manufacturing, chemistry, healthcare, life sciences, and information and communications, were represented in the submitted concepts.
When assessing applications, reviewers concentrated on three primary criteria:
- Concrete use of the platform: Candidates were required to provide specifics about how they intended to combine the HPC and quantum components.
- Role of HPC in the workflow: Priority was given to applications where HPC was a key, necessary element rather than a pointless shell.
- Feasible timeline: Within the operating duration of the test program, applicants had to present practical methods for generating early results.
Nine of the chosen organizations have completed their memorandum of understanding (MoU) and were given the go-ahead to begin their development work by the end of September 2025.
You can also read Drafting Quantum-as-a-Service QaaS Technical Agreements
Participants who have been made public include:
- Toyota Motor Corporation
- Mitsubishi Chemical Corporation
- JSR Corporation
- SoftBank Corp.
- Kyoto University
- Oita University
- Toyota Central Research Institute
- Ochanomizu University
- University of Electro-Communications
As their Memorandums of Understanding are finalized, more grantees will be made publicly available on the JHPC-quantum website.
Target Domains and Technical Hurdles
The sectors that were selected show areas where hybrid architectures are expected to gain traction early on.
When included into larger conventional simulation pipelines, quantum subroutines are anticipated to improve modelling of processes including molecular interaction, catalysis, and adsorption in the fields of materials and chemistry. Routing, layout, and scheduling are examples of optimisation tasks that are frequently the focus of design and manufacture. Quantum heuristics may provide speedups on certain combinatorial subproblems, while classical resources manage scaling and evaluation. Lastly, with validation and large-scale simulations handled by classical computation systems, quantum modules could speed up subproblems like property prediction or docking processes in drug development and the life sciences.
However, there are actual technical dangers when combining these two computational approaches. These difficulties include preventing quantum gains from being negated by data transfer overheads, addressing the problem of error amplification, which occurs when quantum defects reverberate throughout the traditional process, and reworking algorithms that might not transfer well to the hybrid architecture. Importantly, success will be determined by proving quantifiable improvements in execution time, resource utilization, reproducibility, and robustness on integrated, representative workflows, rather than just by job completion.
You can also read IonQ QC-AFQMC Algorithm For Climate and Industrial Research
