Low-power plasmonic SiC nanowire network-based artificial photo-synaptic device for musical classification neural network systems
Artificial synaptic devices for neuromorphic computing must reduce energy consumption to approach biological femtojoule levels. This work reports a SiC/SiO2@Ag nanowire network (NWN) device that emulates both ultraviolet visual and electrical synaptic functions under biased electric field and zero-bias photoexcitation. The NWN architecture and Ag nanoparticle-induced localized surface plasmon resonance (LSPR) enable substantial synaptic responses at ultra-low currents. The device achieves energy consumption of 0.471–0.218 pJ per synaptic event, significantly lower than conventional artificial synapses. In a musical classification task using a spiking neural network with hardware-implemented spike-timing-dependent plasticity (STDP), the system reaches >95% accuracy within 20 training epochs, surpassing software-based STDP and backpropagation after 10 epochs. The SiC NWN structure ensures robust synaptic performance and high precision. These results demonstrate a scalable, energy-efficient hardware foundation for neuromorphic music information processing, with potential for spiking neural networks that mimic biological operational principles.