深度学习硬件方案综述

Review of Deep Learning HW Acceleration Scheme

深度学习技术促进了人工智能的迅速发展,被广泛应用于各个领域。但深度神经网络模型规模庞大、结构复杂,对其进行优化需要耗费巨大的计算资源。随着计算机硬件的快速发展,各种加速器的处理能力显著增强,为深度学习提供了硬件基础。本文首先介绍深度学习背景及其对硬件的需求;然后对当前主要的硬件加速器进行对比分析;最后进行总结展望。

Deep learning has promoted the rapid development of artificial intelligence, and has been widely used in various fields. However, the deep neural networks are usually is huge in scale and complex in structure, whose opti- mization needs huge memory and computational resources. With the development of computer hardware technol- ogy, the accelerator performance is significantly enhanced, which provides the hardware foundation for deep learning. Firstly, this paper presents the background of＂ deep learning and its hardware requirements. Then, the main hardware accelerators are compared and analyzed. Finally, it is summary and prospect.