Evolution of Convolutional Neural Network (CNN): Compute vs Memory bandwidth for Edge AI
It addresses the challenge of optimizing CNN performance on resource-constrained edge devices, which is incremental as it builds on existing knowledge of CNN design.
The paper analyzes the historical evolution of CNN architectures to explore the trade-off between compute requirements and memory bandwidth for Edge AI, providing insights for designing efficient architectures and hardware accelerators.
Convolutional Neural Networks (CNNs) have greatly influenced the field of Embedded Vision and Edge Artificial Intelligence (AI), enabling powerful machine learning capabilities on resource-constrained devices. This article explores the relationship between CNN compute requirements and memory bandwidth in the context of Edge AI. We delve into the historical progression of CNN architectures, from the early pioneering models to the current state-of-the-art designs, highlighting the advancements in compute-intensive operations. We examine the impact of increasing model complexity on both computational requirements and memory access patterns. The paper presents a comparison analysis of the evolving trade-off between compute demands and memory bandwidth requirements in CNNs. This analysis provides insights into designing efficient architectures and potential hardware accelerators in enhancing CNN performance on edge devices.