| name | distributed-qaoa-simulator |
| description | Distributed Quantum Approximate Optimization Algorithm (DQAOA) simulator for QUBO problems across multiple QPUs. Supports monolithic and distributed QAOA execution modes with configurable QPU capacities, cross-QPU coupling handling, and runtime optimizations. Activation: distributed QAOA, DQAOA simulator, QUBO optimization, multi-QPU quantum, quantum unit commitment |
| metadata | {"arxiv_id":"2606.26297","published":"2026-06-24","authors":"Ali Rajabi, Milad Hasanzadeh, Amin Kargarian","tags":["quantum","optimization","distributed-computing","QAOA","QUBO","economics"]} |
Distributed QAOA (DQAOA) Simulator
Description
Open-source Qiskit-compatible DQAOA simulator for QUBO problems arising in engineering design and economic decision applications. Supports monolithic QAOA on single QPU and distributed QAOA across multiple QPUs with configurable capacities.
Activation Keywords
- distributed QAOA
- DQAOA simulator
- QUBO multi-QPU
- quantum unit commitment
- distributed quantum optimization
- 分布式量子近似优化
- QUBO分布式求解
Core Architecture
Workflow Pipeline
- QUBO Canonicalization: Standardize QUBO model formulation
- Cost Hamiltonian Mapping: Map QUBO to quantum cost Hamiltonian
- Variable Allocation: Distribute variables across QPUs by capacity
- Coupling Identification: Separate local vs cross-QPU couplings
- Circuit Construction: Build corresponding circuits per QPU
- Execution & Aggregation: Run modes and aggregate results
Runtime Optimizations
- Parameterized circuit reuse (avoid recompilation)
- Objective reuse at fixed depth
- Batched evaluations
- Parallel multi-start execution
Execution Modes
- Monolithic: Single QPU, standard QAOA
- Distributed: Multiple QPUs with cross-QPU coupling via remote operations
- Hybrid: Classical preprocessing + quantum optimization
Methodology
Step 1: QUBO Formulation
Express problem as QUBO: minimize x^T Q x for binary x.
Step 2: Distributed Allocation
Partition variables across N QPUs based on capacity constraints. Identify:
- Local terms (within single QPU)
- Cross-QPU terms (require remote operations)
Step 3: Circuit Construction
For each QPU:
- Local cost Hamiltonian → RZZ gates
- Mixer Hamiltonian → RX gates
- Cross-QPU couplings → remote entangling operations
Step 4: Parameter Optimization
Optimize QAOA angles (γ, β) using classical optimizer.
Step 5: Solution Recovery
Measure bitstrings, recover optimal solution, compare across modes.
Usage Patterns
Pattern 1: Engineering Design Optimization
Apply DQAOA to QUBO-formulated engineering design problems (power generation unit commitment, resource allocation).
Pattern 2: Economic Decision Optimization
Use for portfolio optimization, scheduling, and other economic problems formulatable as QUBO.
Pattern 3: QPU Capacity Planning
Simulate different QPU configurations to determine minimum hardware requirements for target problem sizes.
Pitfalls
Cross-QPU Communication Overhead
Distributed QAOA is more demanding than monolithic because cross-QPU couplings require remote operations. Evaluate whether distribution actually reduces wall-clock time vs. waiting for larger single QPU.
Enforcement Cap Limitation
Similar to SCUC paper (2606.26345), distributed QAOA shows coverage bottleneck when enforcement cap no longer spans complete commitment period.
Qiskit Compatibility
Simulator is Qiskit-compatible — requires Qiskit installation. Not all quantum backends support the required gate set.
Problem Size Scaling
Distributed approach shines for problems too large for single QPU but requires careful variable partitioning. Poor partitioning → excessive cross-QPU communication → slower than monolithic.
References
- arXiv: 2606.26297 - "A Distributed Quantum Approximate Optimization Algorithm Simulator for Engineering Design Optimization"
- Related:
qaoa-manifold-optimization (QAOA parameter optimization)
- Related:
quantum-rl-scuc-qsample (quantum RL for unit commitment)
