Probabilistic spin wave computing with quasistatic magnetic inputs

Abstract

Spin wave computing device where an algorithm can be encoded by recording a corresponding magnetization pattern onto a hard magnetic material was previously proposed [K. Rivkin and M. Montemorra, “Spin wave computing using pre-recorded magnetization patterns,” J. Appl. Phys. 132, 153902–153911 (2022)] and a particular implementation of a vector-matrix algorithm was demonstrated. In the present article, we analyze the conditions allowing for implementation of complex algorithms which can combine multiple additive, multiplicative, and conditional operators including logic expressions. Special attention is given to how the input data are provided. Rather than relying on a set of independent sources of the RF field, as is common with the existing spin wave computing methods, we demonstrate usability of more simple solutions using adjustable external quasistatic magnetic fields. We also show how, for the given setup, probabilistic switching of magnetic elements at elevated temperatures can be used to convert deterministic algorithms into a probabilistic form.