A study on supporting spiking neural network models based on multiple neuromorphic processors

Citations Over Time

Abstract

Neuromorphic computing or neuromorphic engineering is an engineering discipline that attempts to simulate human brain function by creating circuits that mimic the shape of neurons. In the field of neuromorphic computing, neuromorphic processors are used. There are many types of neuromorphic processors, and there are neuromorphic processors implemented based on FPGAs. Neuromorphic processors use an artificial intelligence model called a spiking neural network. Each neuromorphic processor has different characteristics. For example, the spiking neural network model supported by each supported neuromorphic processor may be different. In this paper, we propose a service architecture named NAAL (Neuromorphic Architecture Abstract Layer) that enables the use of spiking neural networks by virtualizing various neuromorphic processors with different characteristics.

Related Papers

• → A low-power Si:HfO2 ferroelectric tunnel memristor for spiking neural networks(2022)90 cited
• → Performance and Energy Simulation of Spiking Neuromorphic Architectures for Fast Exploration(2023)8 cited
• → Error estimation and correction in a spiking neural network for map formation in neuromorphic hardware(2020)21 cited
• → Training Robust Spiking Neural Networks on Neuromorphic Data with Spatiotemporal Fragments(2023)2 cited
• → Island model for parallel evolutionary optimization of spiking neuromorphic computing(2019)3 cited