Simulated annealing with surface acoustic wave in a dipole-coupled array of magnetostrictive nanomagnets for collective ground state computing

Abstract

In a two-dimensional arrangement of closely spaced elliptical nanomagnets with in-plane magnetic anisotropy, whose major axes are aligned along columns and minor axes along rows, dipole coupling will make the magnetic ordering "ferromagnetic" along the columns and "anti-ferromagnetic" along the rows. Noise and other perturbations can drive the system out of this ground state configuration and pin it in a metastable state where the magnetization orientations will not follow this pattern. Internal energy barriers, sufficiently larger than the thermal energy kT, will prevent the system from leaving the metastable state and decaying spontaneously to the ground state. These barriers can be temporarily eroded by globally straining the nanomagnets with time-varying strain if the nanomagnets are magnetostrictive, which will allow the system to return to ground state after strain is removed. This is a hardware emulation of simulated annealing in an interacting many body system. Here, we demonstrate this function experimentally.

Applications of nanomagnets as dynamical systems: I

“…Consequently, they are much more forgiving of single device errors than Boolean logic. They include application-specific-integrated-circuits for image processing [6,59], simulated annealing in energy minimization computing for solving combinatorial optimization problems [60], probabilistic computing with p-bits [12][13][14][15], computer vision [16,17], Bayesian inference engines for belief networks and machine learning [18][19][20][21], restricted Boltzmann machines [22], ternary content addressable memory [23] and non-volatile and reconfigurable equality bit comparators [32] for electronic locks and hardware countermeasures against cyberattacks.…”

Applications of nanomagnets as dynamical systems: II

“…Fortunately, there are a host of non-Boolean computing applications for which magnetic devices may be eminently suitable. They include application-specific-integrated-circuits (ASICs) for image processing [16,17], simulated annealing in energy minimization computing for solving combinatorial optimization problems [18], probabilistic computing with p-bits [19][20][21][22], computer vision [23,24], Bayesian inference engines for belief networks and machine learning [25][26][27][28], restricted Boltzmann machines [29], ternary content addressable memory [30] and non-volatile and reconfigurable equality bit comparators [31] for electronic locks and hardware countermeasures against cyberattacks.…”