An Application Benchmark for Quantum Optimization
Quantum computers hold promise as accelerators onto which classically, intractable problems may be offloaded. Consequently, deeply understanding how quantum and classical compute resources can be used in tandem is necessary to identify areas for which near-term quantum computers could be useful. In this work, we study such ”hybrid quantum-classical workflows” in the context of quantum optimization, using variational quantum factoring (VQF) as a prototypical example algorithm for which such workflows are naturally suited. We implement VQF on a superconducting quantum system made available by IBM Quantum and examine the tradeoff between the amount of quantum compute resources used (i.e., number of qubits and circuit depth) and performance of VQF. Our results on 3,4, and 5-qubit VQF instances reveal coherent noise coming from static ZZ couplings between the qubits substantially impacts performance. These results suggest that VQF could form an “application-based” benchmark for the quantum computers of today, and tomorrow.
Zapata Speaker bio: