Recurrent Neural Networks Made of Magnetic Tunnel Junctions

TL;DR

This paper demonstrates that recurrent neural networks built from magnetic tunnel junctions can efficiently generate and recognize periodic time series, offering a promising hardware platform for neuromorphic computing with high speed and durability.

Contribution

It introduces a theoretical and numerical framework for implementing recurrent neural networks using magnetic tunnel junctions, highlighting their potential for brain-inspired hardware.

Findings

Networks with as few as 40 magnetic tunnel junctions can perform time series tasks

Magnetic tunnel junction-based RNNs operate at ultrahigh speeds

Devices exhibit nonvolatile memory, high endurance, and reproducibility

Abstract

Artificial intelligence based on artificial neural networks, which are originally inspired by the biological architectures of human brain, has mostly been realized using software but executed on conventional von Neumann computers, where the so-called von Neumann bottleneck essentially limits the executive efficiency due to the separate computing and storage units. Therefore, a suitable hardware platform that can exploit all the advantages of brain-inspired computing is highly desirable. Based upon micromagnetic simulation of the magnetization dynamics, we demonstrate theoretically and numerically that recurrent neural networks consisting of as few as 40 magnetic tunnel junctions can generate and recognize periodic time series after they are trained with an efficient machine-learning algorithm. With ultrahigh operating speed, nonvolatile memory and high endurance and reproducibility, spintronic devices are promising hardware candidates for neuromorphic computing.