Data-driven Thiele equation approach for solving the full nonlinear spin-torque vortex oscillator dynamics (2206.13596v1)

Abstract: The dynamics of vortex based spin-torque nano-oscillators is investigated theoretically. Starting from a fully analytical model based on the Thiele equation approach, fine-tuned data-driven corrections are carried out to the gyrotropic and damping terms. These adjustments, based on micromagnetic simulation results, allow to quantitatively model the response of such oscillators to any dc current within the range of the vortex stability. Both, the transient and the steady-state regimes are accurately predicted under the proposed data-driven Thiele equation approach. Furthermore, the computation time required to solve the dynamics of such system is reduced by about six orders of magnitude compared to the most powerful micromagnetic simulations. This major breakthrough opens the path for unprecedented high-throughput simulations of spin-torque vortex oscillators submitted to long-duration input signals, for example in neuromorphic computing applications.