Reinforcement-Learning-Designed Field-Free Sub-Nanosecond Spin-Orbit-Torque Switching

TL;DR

This paper demonstrates ultra-fast, deterministic magnetization switching in nanomagnets within 300 ps using reinforcement learning to optimize current waveforms, revealing new insights into spin-orbit torque dynamics.

Contribution

It introduces a reinforcement learning approach to discover optimal current waveforms for field-free, sub-nanosecond spin-orbit-torque switching, providing a new physical understanding and analytical model.

Findings

Achieved magnetization reversal within 300 ps at high current density

RL-discovered waveforms exploit precessional shortcuts via SOTs

Control strategy is robust against damping variations and thermal noise

Abstract

We demonstrate deterministic, field-free magnetization reversal of a single-domain nanomagnet within 300 ps under a current density of $3 \times 10^{10}~\mathrm{A/m^2}$ by coupling reinforcement learning (RL) to the Landau-Lifshitz-Gilbert equation with the spin-orbit torques (SOTs). The RL agent autonomously discovers a current waveform that minimizes the magnetization trajectory path and exploits a precessional shortcut enabled by the field-like SOT and hard-axis anisotropy. From the learned pulse, we extract a clear physical picture of the dynamics and develop a model-based analytical framework that establishes a lower bound on the switching time. The control strategy remains robust across a wide range of damping constants and is stabilized against thermal fluctuations at higher current densities. We also discuss feasible experimental implementations for the precessional switching.