Low-power Spike-based Wearable Analytics on RRAM Crossbars

2025-02-10

AI Summary

To address the noise sensitivity, resource constraints, and energy-efficiency bottlenecks of RRAM-based in-memory computing hardware, this work proposes a low-power spiking neural network (SNN) online learning system tailored for wearable applications. We introduce direct feedback alignment (DFA) for the first time to train SNNs on RRAM crossbar arrays, enabling layer-wise parallel gradient updates and circumventing the high latency and energy overhead of backpropagation in in-memory architectures. Leveraging our custom hardware simulator DFA_Sim, we perform end-to-end software-hardware co-optimization. Experimental results demonstrate that, compared to conventional backpropagation, the proposed system reduces energy consumption by 64.1%, decreases chip area by 10.1%, cuts inference latency by 2.1×, and improves human activity recognition accuracy by 7.55%. These advances significantly enhance online adaptability and edge-deployment feasibility.

Technology Category

• Machine Learning: Hardware-aware ML
• Cognitive Modeling & Cognitive Systems: Neural Spike Coding
• Search and Optimization: Learning to Search

Application Category

• Search and Retrieval-Augmented AI: Web learning to rank, online learning, and counterfactual learning for ranking
• Systems and Infrastructure for Web, Mobile and WoT: Applied ML and AI for Web-based mobile applications
• Graph Algorithms and Modeling for the Web: Graph neural networks and deep learning approaches for Web-related graphs

Abstract

This work introduces a spike-based wearable analytics system utilizing Spiking Neural Networks (SNNs) deployed on an In-memory Computing engine based on RRAM crossbars, which are known for their compactness and energy-efficiency. Given the hardware constraints and noise characteristics of the underlying RRAM crossbars, we propose online adaptation of pre-trained SNNs in real-time using Direct Feedback Alignment (DFA) against traditional backpropagation (BP). Direct Feedback Alignment (DFA) learning, that allows layer-parallel gradient computations, acts as a fast, energy&area-efficient method for online adaptation of SNNs on RRAM crossbars, unleashing better algorithmic performance against those adapted using BP. Through extensive simulations using our in-house hardware evaluation engine called DFA_Sim, we find that DFA achieves upto 64.1% lower energy consumption, 10.1% lower area overhead, and a 2.1x reduction in latency compared to BP, while delivering upto 7.55% higher inference accuracy on human activity recognition (HAR) tasks.

Problem

• Develop low-power wearable analytics system
• Utilize Spiking Neural Networks on RRAM
• Implement online adaptation with DFA

Innovation

• Spiking Neural Networks on RRAM
• Direct Feedback Alignment learning
• Energy-efficient online SNN adaptation