Convolutions with Radio-Frequency Spin-Diodes

2025-01-27

AI Summary

To address the high analog-to-digital conversion overhead and scalability limitations of spintronic hardware in radio-frequency (RF) signal classification, this work proposes an analog-domain RF sensing and computing paradigm based on metallic spin diodes (NiFe/Pt). Leveraging ferromagnetic resonance responses, the approach enables in-situ weighted summation and 2×2 convolution operations directly on RF signals—demonstrating, for the first time, a scalable hardware implementation of spin-diode-based convolutional units. A four-device chain co-designed hardware–software system achieves 88% top-1 accuracy on the first 100 Fashion-MNIST images, approaching the software baseline (88.4% with noise, 90% noise-free). This work overcomes the scalability bottleneck of spintronic devices in edge RF intelligence and establishes a new pathway toward low-power, high-throughput analog neuromorphic computing.

Technology Category

• Machine Learning: Hardware-aware ML
• Cognitive Modeling & Cognitive Systems: Neural Spike Coding
• Humans and AI: Brain-Sensing and Analysis

Application Category

• Responsible Web: Machine-in-the-loop, human agency and autonomy
• Security and Privacy: Large-scale security measurements
• Systems and Infrastructure for Web, Mobile and WoT: Applied ML and AI for Web-based mobile applications

Abstract

The classification of radio-frequency (RF) signals is crucial for applications in robotics, traffic control, and medical devices. Spintronic devices, which respond to RF signals via ferromagnetic resonance, offer a promising solution. Recent studies have shown that a neural network of nanoscale magnetic tunnel junctions can classify RF signals without digitization. However, the complexity of these junctions poses challenges for rapid scaling. In this work, we demonstrate that simple spintronic devices, known as metallic spin-diodes, can effectively perform RF classification. These devices consist of NiFe/Pt bilayers and can implement weighted sums of RF inputs. We experimentally show that chains of four spin-diodes can execute 2x2 pixel filters, achieving high-quality convolutions on the Fashion-MNIST dataset. Integrating the hardware spin-diodes in a software network, we achieve a top-1 accuracy of 88 % on the first 100 images, compared to 88.4 % for full software with noise, and 90 % without noise.

Problem

• RF signal classification
• analog signal processing
• spintronic device scalability

Innovation

• Metal Spin Diode
• Signal Classification
• Hardware-Software Integration