论文标题
使用操作数锤距离优化提高神经网络加速器的效率
Improving Efficiency in Neural Network Accelerator Using Operands Hamming Distance optimization
论文作者
论文摘要
Neural network accelerator is a key enabler for the on-device AI inference, for which energy efficiency is an important metric.数据路径能量，包括算术单元之间的计算能量和数据运动能量，声称是总加速器能量的重要组成部分。通过重新访问算术逻辑电路的基本物理学，我们表明数据路径能与将输入操作数流到算术单元中时的位截面高度相关，该单元定义为输入操作数矩阵的锤击距离。基于洞察力，我们提出了一种训练后优化算法和一种锤击 - 距离感知的培训算法，以协调并优化加速器和网络协同效应。基于MobilenetV2的Layout后模拟的实验结果平均证明了2.85倍数据路径能量降低，并且对于某些层的数据路径降低高达8.51倍。
Neural network accelerator is a key enabler for the on-device AI inference, for which energy efficiency is an important metric. The data-path energy, including the computation energy and the data movement energy among the arithmetic units, claims a significant part of the total accelerator energy. By revisiting the basic physics of the arithmetic logic circuits, we show that the data-path energy is highly correlated with the bit flips when streaming the input operands into the arithmetic units, defined as the hamming distance of the input operand matrices. Based on the insight, we propose a post-training optimization algorithm and a hamming-distance-aware training algorithm to co-design and co-optimize the accelerator and the network synergistically. The experimental results based on post-layout simulation with MobileNetV2 demonstrate on average 2.85X data-path energy reduction and up to 8.51X data-path energy reduction for certain layers.
