Neuromorphic Engineering For Asynchronous Event Based Systems

# Neuromorphic Engineering For Asynchronous Event Based Systems

## Core Concepts

Neuromorphic engineering aims to mimic the biological nervous system's structure and function to create novel computing architectures. Unlike traditional von Neumann architectures, which process information in discrete time steps, neuromorphic systems leverage *asynchronous event-based* computation. This means processing occurs only when a change in input is detected, leading to significant power efficiency.

**Key Principles:**
*   **Sparsity:** Biological neural activity is sparse; only a small percentage of neurons fire at any given time. Neuromorphic systems aim to replicate this sparsity.
*   **Parallelism:** The brain operates in a massively parallel manner. Neuromorphic architectures exploit this parallelism for faster processing.
*   **Locality:** Computation and memory are co-located, reducing data movement and energy consumption.
*   **Event-Driven:** Information is encoded and transmitted as *events* (spikes) rather than continuous values.

## Event-Based Sensors

Traditional image sensors capture frames at a fixed rate, even when the scene is static. Event-based sensors, such as Dynamic Vision Sensors (DVS), only report changes in brightness. This results in: