Micromagnetic simulation for random magnetization switching process of a spin–orbit true random number generator

Abstract

True random number generator (TRNG) is an important component for modern information security technologies. Among the candidates, TRNG with spin–orbit torque (SOT)-induced probabilistic magnetization switching is expected to be competitive for its advantages in anti-radiation, unlimited endurance, robust stability, and broad temperature range. However, specific mechanism of the randomness in SOT-induced magnetization switching are still not clear, which limits the demonstration of applicable SOT-TRNG. Here, we performed micromagnetic simulation of the SOT-induced probabilistic magnetization switching by using MuMax3. When various thermal noise seeds were introduced, not only stochastic precession trails within the current pulses, but also stochastic precession tendency after removing the current pulse were obtained, together resulting in random final magnetization states. Particularly, the consistency between mz orientations at the current withdraw moment and at the final stabilized magnetization state shows a positive correlation with the fluctuation range of mz during the current pulse duration, the latter of which narrows with the increasing of channel current density. Our work suggests the importance of thermal noise-related magnetization precession variations on the SOT-induced probabilistic magnetization switching, and suggesting the MuMax3 to be a practical tool for simulating the SOT-TRNG.