.png){kind=icon}
返回
Efficient Precision-Adjustable Architecture for Softmax Function in Deep Learning
.png){kind=icon}
PRE
AI
DOI：10.1109/TCSII.2020.3002564.png)
.png){kind=icon}
原文求助.png){kind=icon}
分享.png){kind=icon}
收藏
摘要
En 中文
The softmax function has been widely used in deep neural networks (DNNs), and studies on efficient hardware accelerators for DNN have also attracted tremendous attention. However, it is very challenging to design efficient hardware architectures for softmax because of the expensive exponentiation and division calculations in it. In this brief, the softmax function is firstly simplified by exploring algorithmic strength reductions. Afterwards, a hardware-friendly and precision-adjustable calculation method for softmax is proposed, which can meet different precision requirements in various deep learning (DL) tasks. Based on the above innovations, an efficient architecture for softmax is presented. By tuning the parameter P, the system accuracy and complexity can be adjusted dynamically to achieve a good tradeoff between them. The proposed design is coded using hardware description language (HDL) and evaluated on two platforms, Xilinx Zynq-7000 ZC706 development board and TSMC 28-nm CMOS technology, respectively. Hardware implementation results show that our architecture significantly outperforms other works in speed and area, and that by adjusting P, the accuracy can be further increased with little hardware overhead.
Keyword:
Hardware
Computer architecture
Complexity theory
Neural networks
Circuits and systems
Task analysis
Mathematical model
Softmax
hardware architecture
deep neural network
transformer
VLSI
AI总结
对已上传原文的论文进行重点信息的提取，主要内容包括：简要概述、研究摘要、背景介绍、关键亮点、图文解析、展望与总结。
期刊
I
IF:
4.9
论文数:
8.8K
被引数:
2.5W
机构
引用论文
Effect of milking vacuum level and overmilking on cows' teat characteristics挤奶真空度和过度挤奶对奶牛奶头特性的影响
OAAI
.png){kind=icon}
OAAI分享收藏
Intracellular Co‐Delivery of Carbon Monoxide and Nitric Oxide Induces Mitochondrial Apoptosis for Cancer Therapy细胞内共递送一氧化碳和一氧化氮诱导线粒体凋亡用于癌症治疗
PREAI
分享收藏分享收藏 没有更多内容
没有更多内容