Neuromorphic hardware as a self-organizing computing system
This work addresses the challenge of creating scalable and fault-tolerant neuromorphic computing systems for applications in AI and hardware design, but it appears incremental as it builds on existing concepts of neural plasticity.
The paper tackles the problem of designing self-organizing neuromorphic hardware by proposing the SOMA architecture, which aims to achieve hardware efficiency and scalability through digital spiking neurons and neuro-cellular structures, though no concrete performance numbers are provided.
This paper presents the self-organized neuromorphic architecture named SOMA. The objective is to study neural-based self-organization in computing systems and to prove the feasibility of a self-organizing hardware structure. Considering that these properties emerge from large scale and fully connected neural maps, we will focus on the definition of a self-organizing hardware architecture based on digital spiking neurons that offer hardware efficiency. From a biological point of view, this corresponds to a combination of the so-called synaptic and structural plasticities. We intend to define computational models able to simultaneously self-organize at both computation and communication levels, and we want these models to be hardware-compliant, fault tolerant and scalable by means of a neuro-cellular structure.