Quantum annealing (QA) and Quantum Alternating Operator Ansatz (QAOA) are both heuristic quantum algorithms intended for sampling optimal solutions of combinatorial optimization problems. In this article we implement a rigorous direct comparison between QA on D-Wave hardware and QAOA on IBMQ hardware. These two quantum algorithms are also compared against classical simulated annealing. The studied problems are instances of a class of Ising models, with variable assignments of $+1$ or $-1$, that contain cubic $ZZZ$ interactions (higher order terms) and match both the native connectivity of the Pegasus topology D-Wave chips and the heavy hexagonal lattice of the IBMQ chips. The novel QAOA implementation on the heavy hexagonal lattice has a CNOT depth of $6$ per round and allows for usage of an entire heavy hexagonal lattice. Experimentally, QAOA is executed on an ensemble of randomly generated Ising instances with a grid search over $1$ and $2$ round angles using all 127 programmable superconducting transmon qubits of ibm_washington. The error suppression technique digital dynamical decoupling is also tested on all QAOA circuits. QA is executed on the same Ising instances with the programmable superconducting flux qubit devices D-Wave Advantage_system4.1 and Advantage_system6.1 using modified annealing schedules with pauses. We find that QA outperforms QAOA on all problem instances. We also find that dynamical decoupling enables 2-round QAOA to marginally outperform 1-round QAOA, which is not the case without dynamical decoupling.
The central idea of the paper is to provide a rigorous comparison between QAOA and Quantum annealing. In order to do this, the authors constructed very specific Ising models that matched the connectivity graph of the current IBMQ quantum computers, which have a heavy-hex graph, and the Pegasus connectivity graph of the D-Wave quantum annealer. The QAOA circuit construction uses a CNOT depth of 6 per round, which means these very short depth QAOA circuits can actually be run on the entire 127 qubit chip of ibm_washington. The Ising models also contained higher order terms, which further tests the capabilities of both QAOA (which can natively handle higher order terms) and Quantum annealing. For the Quantum annealing implementation, custom order-reduction techniques were used to embed the problems onto the chip, and problem instances were executed in parallel. Dynamical decoupling was also tested on the QAOA circuits.
QAOA and Quantum annealing are both types of computation that address combinatorial optimization problems. It is of interest to investigate how these two algorithms perform, especially when there is noise in the computation, and how they compare to each other.
The primary result is that quantum annealing performs better than QAOA, specifically when these QAOA circuits were executed on ibm_washington. Importantly, the QAOA circuit and the QA order reduction algorithm are both scalable, meaning that the problem instances and algorithms can be applied to future hardware as new devices increase in scale and fidelity.