大数据 ›› 2020, Vol. 6 ›› Issue (4): 69-80.doi: 10.11959/j.issn.2096-0271.2020034

• 专题:大数据异构并行系统 • 上一篇    下一篇

新一代深度学习框架研究

于璠   

  1. 华为技术有限公司,广东 深圳 518000
  • 出版日期:2020-07-15 发布日期:2020-07-18
  • 作者简介:于璠(1980- ),男,博士,华为技术有限公司AI计算框架的创新和生态架构师,曾主导华为云计算资源调度、SDN大规模路由等架构和算法的设计,发表专利30余篇

Research on the next-generation deep learning framework

Fan YU   

  1. Huawei Technologies Co.,Ltd.,Shenzhen 518000,China
  • Online:2020-07-15 Published:2020-07-18

摘要:

从人工智能的历史出发,简述深度学习发展历程以及目前的挑战,通过介绍新一代深度学习框架的特点,分析总体框架,阐述自动并行、自动微分、自动调优等技术优势以及协同昇腾处理器的性能优势,希望可以为深度学习技术研究人员提供参考。

关键词: 人工智能;机器学习, 深度学习, 计算框架, MindSpore

Abstract:

Started from the history of AI,the development and challenges of deep learning were described,the features of the nextgeneration deep learning framework was introduced,the overall framework was analyzed,and the technical advantages of auto parallel,auto differentiation,automatic tuning,as well as the performance advantages of collaborating with Ascend processors were expanded.This article can be used as a reference for deep learning technology researchers.

Key words: artificial intelligence, machine learning, deep learning, computing framework, MindSpore

中图分类号: 

No Suggested Reading articles found!