USEFUSE: Uniform Stride for Enhanced Performance in Fused Layer Architecture of Deep Neural Networks

TL;DR

This paper introduces USEFUSE, a novel approach that employs uniform stride and fusion techniques to optimize CNN performance on edge devices by reducing memory access, power consumption, and redundant computations.

Contribution

It proposes a new fusion methodology with uniform stride and SOP units, enhancing CNN efficiency and performance on resource-limited edge hardware.

Findings

Reduced off-chip memory communication

Lower power consumption without accuracy loss

Improved response time for edge applications

Abstract

Convolutional Neural Networks (CNNs) are crucial in various applications, but their deployment on resource-constrained edge devices poses challenges. This study presents the Sum-of-Products (SOP) units for convolution, which utilize low-latency left-to-right bit-serial arithmetic to minimize response time and enhance overall performance. The study proposes a methodology for fusing multiple convolution layers to reduce off-chip memory communication and increase overall performance. An effective mechanism detects and skips inefficient convolutions after ReLU layers, minimizing power consumption without compromising accuracy. Furthermore, efficient tile movement guarantees uniform access to the fusion pyramid. An analysis demonstrates the utile stride strategy improves operational intensity. Two designs cater to varied demands: one focuses on minimal response time for mission-critical applications, and another focuses on resource-constrained devices with comparable latency. This approach notably reduced redundant computations, improving the efficiency of CNN deployment on edge devices.