Flexible Optimization of Noise-Robust Rydberg Gates

Abstract

High fidelity two-qubit gates are a crucial requirement for quantum computers that can run deep circuits and error correction at scale. Yet two-qubit gate fidelities consistently lag behind single-qubit gate fidelities due to entangling operations being considerably harder to control. Nevertheless, In the last few years, two-qubit gates in neutral atom quantum computers have seen dramatic improvements, largely due to better gate protocols produced by quantum optimal control (QOC) algorithms. However, the most popular algorithm for optimizing Rydberg gates, GRAPE, is inefficient in its description of controls and difficult to apply towards optimizing other objectives in addition to gate fidelity that are experimentally relevant. In this report, I explore the Gradient Optimization of Analytic Controls (GOAT) algorithm as a more flexible alternative to GRAPE. I demonstrate that GOAT can achieve gate fidelities and runtimes that are similar to GRAPE and explore some of the ways in which the algorithm can be used to produce gates that are more resilient to experimental imperfections.