Memristive synapse enables learning and inference in spiking neural networks
A Memristive Synapse for Online STDP Learning and Inference in SNNs
Neural and Evolutionary ComputingHardware ArchitectureEmerging Technologies
Summary
Spiking neural networks learn by adjusting connections between neurons based on the timing of their signals. The authors propose a tiny circuit that mimics this learning directly in hardware without needing extra digital control. This circuit uses a memristor to update connection strengths as signals arrive, enabling learning during normal operation. They tested the circuit design through simulations and showed that a small network can specialize and learn on its own.
What this means in practice
- •For neuromorphic hardware developers: Design hardware neural networks that learn on-chip through spike timing without external digital control.
- •For embedded system engineers: Implement low-power learning systems enabling real-time adaptation in edge devices using analog memristive synapses.
Authors
Elia Mateu-Barriendos, Álvaro Gómez-Pau, Daniel Arumí, Rosa Rodríguez-Montañés, Salvador Manich
Abstract
This work presents a fully analog memristive synaptic circuit for online spike-timing-dependent plasticity (STDP) learning in spiking neural networks (SNNs). The proposed synapse integrates a local STDP circuit generating gradual timing-dependent conductance updates directly from pre- and post-synaptic spikes. Learning occurs during normal network operation without requiring external digital control or explicit STDP waveform synthesis. Post-layout simulations of the memristive synapse implemented in a 130 nm CMOS technology show spike-timing-dependent conductance adaptation during SNN operation. A 2x2 SNN simulation further illustrates online neuron specialization through unsupervised learning.