AI-assisted hybrid quantum-classical benchmarks for optimization, simulation, and graph algorithms.
The QHPC Shared Tasks invite participants to use AI-assisted hybrid quantum-classical workflows to solve benchmark problems across optimization, quantum simulation, and graph algorithms. Submissions will be evaluated on solution quality, quantum resource efficiency, robustness, reproducibility, and transparency in the use of AI. Submit your solution as a workshop paper.
Tracks:
1. Quantum Optimization
2. Quantum Simulation
3. Graph Theory
Quantum Optimization
T1. AI-Assisted QAOA for Weighted Max-Cut
Solve weighted Max-Cut instances using QAOA, variational quantum methods, quantum-inspired approaches, or AI-designed hybrid workflows.
- Approximation ratio
- Success probability
- Circuit depth and two-qubit gate count
- Shots and optimizer calls
Quantum Optimization
T2. Constrained QUBO for Scientific Resource Allocation
Solve constrained binary optimization problems inspired by HPC scheduling, quantum resource allocation, and experiment planning.
- Feasible objective quality
- Constraint violation rate
- Penalty-adjusted QUBO energy
- Runtime and evaluation cost
Quantum Simulation
T3. Molecular and Materials Ground-State Energy
Estimate ground-state energies for molecular or materials Hamiltonians using VQE, QPE-inspired, or AI-assisted quantum simulation workflows.
- Energy error against reference
- Chemical accuracy rate
- Measurement cost
- Qubits, depth, and gate count
Quantum Simulation
T4. Detection of Quantum Phase Transitions
Simulate spin models such as the transverse-field Ising model and use quantum measurements with AI-assisted analysis to identify phase transitions.
- Critical-point error
- Observable accuracy
- Phase-classification accuracy
- Noise-aware performance
Graph Algorithms
T5. Maximum Independent Set on Hardware-Relevant Graphs
Solve maximum independent set instances, including unit-disk and neutral-atom-inspired graphs, using quantum or hybrid methods.
- Independent-set ratio
- Validity rate
- Embedding overhead
- Resource cost
Graph Algorithms
T6. Graph Partitioning and Community Detection
Partition scientific or synthetic graphs using QAOA-style, variational, quantum-inspired, or AI-assisted graph workflows.
- Modularity or normalized-cut score
- Balance constraint score
- Approximation ratio
- Runtime and reproducibility
Evaluation Framework
Each submission will include code, result files, quantum circuits or workflow
artifacts, and a short method report. Participants may use AI tools for code generation, ansatz design, parameter search, denoising, analysis, and documentation, provided that the AI-assisted workflow is described clearly.
- Solution quality 50%
- Quantum resource efficiency 20%
- Robustness / noise-aware performance 15%
- Reproducibility 10%
- AI workflow transparency 5%