LIF CMOS Neuron for Neuromorphic Computing

Abstract

As Artificial Intelligence is paving way for advancement of computing age, the need for efficient computation is coming to the forefront as most pivotal factor. Analog neurons and neuromorphic computing have emerged as promising candidates to address the need of emerging computational demand. This paper presents a LIF (Leaky Integrate and Fire) CMOS spiking neuron design, which is implemented and simulated in GPDK 180nm technology. The functionality, power efficiency and configuration viability of the neuron is tested through meticulously planned test cases. The neuron layout fits in 0.02mm² and has approximate energy efficiency of 10pJ/Spike.