Modeling Synapse Formation and Growth as a Non-Biological Analog to the Brain

Abstract

Efficiency in computing systems is a pressing concern as global reliance on machines and automation grows. Leveraging an understanding of the brain's exceptional computational capabilities, this study presents a preliminary nondimensionalized model of synapses, an essential component for developing brain-inspired computing systems. The model simulates a physical analog of synapse formation wherein a single two-nanowire junction in an electrolytic medium undergoes an electric potential, causing electric field-driven ion transport and subsequent filament growth. Simulations allow for the extraction of meaningful parameter relationships as well as governing equations relating both filament length and time, and current and time. By investigating electric potential-driven cation diffusion, the model provides insights for designing more advanced computing technologies. Future directions involve refining assumptions, adapting system geometry for dendritic growth, and modeling an entire nanowire network. This research bridges the gap between brain-inspired and physical computing, paving the way for highly efficient computing systems beyond traditional approaches.