Improving the Performance of Rydberg Atom Array Quantum Simulators through Low-Latency Feedback Control Systems

Programmable quantum simulators offer a near-term opportunity to explore many-body physics in regimes inaccessible through classical computation, with implications for solving practically relevant problems. Realizing this opportunity requires leveraging real-time feedback control systems (RTFCs) to prepare, manipulate, and read out the state of quantum simulators, ideally in a closed-loop fashion. However, these RTFCs usually suffer from high latency, especially when involving the acquisition and processing of large amounts of data, as well as computing the feedback signals on a timescale faster than relevant error rates. Here, we report on the development of a low-latency RTFCs suitable for improving the performance of programmable quantum simulators based on configurations of atomic particles, such as Rydberg atom arrays. The system is based on a modular and cost-effective computer architecture built around a motherboard and peripheral cards, supplemented with easy-to-deploy software and efficient pre-optimized algorithms. We first quantify its performance at generating large arrays of optical traps with homogeneous intensity profiles using active diffractive optical elements such as acousto-optic deflectors and spatial light modulators. We then quantify its performance at preparing large configurations of neutral atoms without defects in linear chains and grids of optical traps, using both operational and runtime performance as our metric. This low-latency reconfiguration system relies on efficient implementation of atom reconfiguration algorithms and seamless integration of imaging systems, processors, and actuation devices. Upon optimizing all steps of the closed-loop reconfiguration cycle, we show that the runtime performance is ultimately limited by hardware performance. These results demonstrate that the proposed RTFC architecture can achieve low-latency execution of quantum state manipulation for a broad range of applications; the system can readily be used for preparing large arrays of optical traps and large configurations of atomic particles, as well as stabilizing control parameters, implementing adaptive control protocols and quantum error codes.