Benchmarking QAOA on Amazon Braket devices

In this section, we will use what we have learned with QAOA to compare the performance of Amazon Braket devices. This includes the SV1 and TN1 quantum simulators, IonQ’s 11-qubit Ion Trap quantum computer, Rigetti’s new Aspen-11 38-qubit superconducting quantum processor, D-Wave’s quantum annealer, and the classical simulated annealer, which is also available through D-Wave. We will start with an 11x11 matrix that represents our problem and work our way up to a 100x100 matrix. Along the way, we will discover the strategies that are needed to solve these types of matrices on different quantum devices and the limits of each device.

Optimizing an 11x11 matrix

We will continue to use the IonQ_matrix.csv file in this section. The IonQ device that’s available on Amazon Braket has 11 qubits, so we cannot solve a matrix larger than this with the available device. However, in the IonQ device, every qubit can be entangled...