Spintronics for image recognition: performance benchmarking via ultrafast data-driven simulations

TL;DR

This paper demonstrates image classification using spintronic nanostructures within an echo-state network, employing ultrafast data-driven simulations to evaluate performance on standard datasets, highlighting potential for advanced neuromorphic computing.

Contribution

It introduces the use of a data-driven Thiele equation approach to simulate spintronic devices for neural network applications, enabling efficient benchmarking without experimental constraints.

Findings

Comparable performance to traditional activation functions like ReLU and sigmoid.

Achieved state-of-the-art accuracy on MNIST dataset.

Potential for improved accuracy with deeper architectures.

Abstract

We present a demonstration of image classification using an echo-state network (ESN) relying on a single simulated spintronic nanostructure known as the vortex-based spin-torque oscillator (STVO) delayed in time. We employ an ultrafast data-driven simulation framework called the data-driven Thiele equation approach (DD-TEA) to simulate the STVO dynamics. This allows us to avoid the challenges associated with repeated experimental manipulation of such a nanostructured system. We showcase the versatility of our solution by successfully applying it to solve classification challenges with the MNIST, EMNIST-letters and Fashion MNIST datasets. Through our simulations, we determine that within an ESN with numerous learnable parameters the results obtained using the STVO dynamics as an activation function are comparable to the ones obtained with other conventional nonlinear activation functions like the reLU and the sigmoid. While achieving state-of-the-art accuracy levels on the MNIST dataset, our model's performance on EMNIST-letters and Fashion MNIST is lower due to the relative simplicity of the system architecture and the increased complexity of the tasks. We expect that the DD-TEA framework will enable the exploration of deeper architectures, ultimately leading to improved classification accuracy.