Authors: Aijaz H. Lone, Meng Tang, Daniel N. Rahimi, Xuecui Zou, Dongxing Zheng, Hossein Fariborzi, Xixiang Zhang, Gianluca Setti

Published on: February 06, 2024

Impact Score: 8.22

Arxiv code: Arxiv:2402.03767

Summary

• What is new: Demonstration of magnetic field-gated Leaky integrate and fire (LIF) neuron characteristics in spintronic devices for spiking neural networks (SNNs), which indicates a step forward in neuromorphic computing by integrating spintronics with SNN architectures.
• Why this is important: The quest for energy-efficient data storage and computing architectures beyond traditional CMOS technology.
• What the research proposes: Development of a magnetic multilayer spintronic device that mimics LIF neuron characteristics, controlled by current pulses and external magnetic fields, for use in SNN applications.
• Results: Integration of the developed LIF neuron models in SNN and CSNN frameworks for the MNIST and FMNIST datasets resulted in classification accuracies above 96%.

Technical Details

Technological frameworks used: SNN and CSNN frameworks

Models used: Leaky integrate and fire (LIF) neuron model, modified LIF neuron model

Data used: MNIST and FMNIST datasets

Potential Impact

Data storage technologies, computing hardware manufacturers, and companies invested in neuromorphic computing developments could be significantly impacted or benefit from the advancements presented in this paper.

Want to implement this idea in a business?

We have generated a startup concept here: NeuroSpinTech.