JIJ Launches SQOA Solver on the Qiskit Functions Catalog
TOKYO, [November 5, 2026] — JIJ Inc. today announced the availability of its SQOA Solver in IBM’s Qiskit Functions Catalog as jij/sqoa-solver.
The solver implements JIJ’s Sampling-based Quantum Optimization Algorithm (SQOA), bringing an approach previously demonstrated in research and benchmarking into a form that researchers and practitioners can apply to their own optimization problems.
Current quantum hardware remains constrained by factors including the number of available qubits. JIJ has been developing SQOA to explore how quantum computing can be applied to combinatorial optimization under these hardware constraints.
From Research and Benchmarking to an Accessible Solver
The research underlying SQOA was published in “Sampling-based Quantum Optimization Algorithm with Quantum Relaxation” by Hiromichi Matsuyama and Yu Yamashiro and presented at the 2025 IEEE International Conference on Quantum Computing and Engineering (QCE 2025).
In August 2026, JIJ extended the approach introduced in the paper and conducted benchmark evaluations using MaxCut problems on 3-regular graphs and Maximum Independent Set problem instances from the Quantum Optimization Benchmarking Library (QOBLIB).
This research and benchmarking work has continued to inform the development of SQOA and its implementation in SQOA Solver.
Key Capabilities of SQOA Solver
Handles More Decision Variables Than Available Qubits
SQOA uses Quantum Relaxation to represent multiple binary variables on a single qubit. This enables users to search for solutions to optimization problems with more decision variables than there are qubits available on the target quantum device.
No Iterative Variational Parameter Optimization on the QPU
The quantum device is used for sampling, while solution recovery and refinement are performed classically.
State-preparation parameters are determined classically for each problem instance, eliminating the need for iterative QPU execution to optimize variational parameters.
End-to-End Optimization Workflow
Users provide a binary combinatorial optimization problem formulated as a QUBO or Ising model. The workflow combines classical preprocessing, SQOA-based problem encoding and quantum sampling, classical solution extraction, and local-search postprocessing in a single optimization workflow.
Comment from Scott Crowder, Vice President, IBM Quantum Adoption and Business Development
“Now that JIJ’s quantum optimization approach is available to the wider community, via new Qiskit Function SQOA Solver, I’m excited to see their team benchmark these results—and publish them in the Quantum Optimization Benchmarking Library as we track progress toward advantage in quantum optimization.”
Comment from Yu Yamashiro, CEO of JIJ.Inc.
“To translate advances in quantum computing research into real-world value, it is not enough to simply develop and evaluate algorithms. It is equally important to make them available in a form that allows researchers and developers to test them on their own problems.
Making SQOA Solver available through IBM’s Qiskit Functions Catalog is an important step in bringing JIJ’s quantum optimization technology to a broader user community and strengthening our connection with the global quantum development ecosystem. By feeding insights from real-world use and evaluation back into our R&D, we aim to continue advancing quantum optimization toward practical applications.”
Related Information
SQOA Solver on IBM Quantum Platform
Qiskit Function: jij/sqoa-solver
Related Paper
Hiromichi Matsuyama and Yu Yamashiro, “Sampling-based Quantum Optimization Algorithm with Quantum Relaxation,” 2025 IEEE International Conference on Quantum Computing and Engineering (QCE), 2025.
DOI: 10.1109/QCE65121.2025.00253
arXiv:2504.12629
Related Announcement
JIJ Benchmarks Its Quantum Optimization Method Using QOBLIB and Is Referenced in IBM Quantum Blog, August 21, 2026
https://www.j-ij.com/en/news/20260821