RAISE LAB's DQAOA Simulator scales quantum optimization across multiple QPUs

A team led by Ali Rajabi at RAISE LAB has unveiled a distributed quantum approximate optimization algorithm (DQAOA) simulator tailored for quadratic unconstrained binary optimization (QUBO) problems in engineering design. The open-source package, built on IBM's Qiskit framework and available on GitHub, supports two execution modes: monolithic QAOA on a single QPU, and distributed QAOA across a user-specified number of QPUs with configurable capacities. Key runtime enhancements—parameterized circuit reuse, objective reuse at fixed depth, batched evaluations, and parallel multi-start execution—slash repeated overhead compared to naive implementations.

In validation tests, including a power generation unit commitment application, DQAOA recovered the exact optimal bitstring and operating cost identified by brute-force enumeration. The simulator also includes a Streamlit GUI for uploading QUBO instances, configuring solver settings, and visualizing solution quality without coding. Staged runtime analysis shows substantial reductions across implementation stages, though distributed mode remains more demanding due to cross-QPU coupling requiring remote operations. The paper (arXiv:2606.26297) provides 37 pages of benchmarks, confirming consistency with classical monolithic QAOA references.