The Feature,Programming Model and Performance Optimization Strategy of Heterogeneous Many-Core System:A Review

JU Tao; ZHU Zheng-dong; DONG Xiao-she

doi:10.3969/j.issn.0372-2112.2015.01.018

您当前的位置：

首页 >

文章列表页 >

The Feature,Programming Model and Performance Optimization Strategy of Heterogeneous Many-Core System:A Review

更新时间：2025-07-16

- The Feature,Programming Model and Performance Optimization Strategy of Heterogeneous Many-Core System:A Review
- Acta Electronica Sinica Vol. 43, Issue 1, Pages: 111-119(2015)
- 作者机构：
  
  西安交通大学电子与信息工程学院,陕西,西安,710049
- 作者简介：
- 基金信息：
- DOI：10.3969/j.issn.0372-2112.2015.01.018
  CLC： TP301
- Published：2015
- 稿件说明：
移动端阅览
JU Tao, ZHU Zheng-dong, DONG Xiao-she. The Feature,Programming Model and Performance Optimization Strategy of Heterogeneous Many-Core System:A Review[J]. Acta Electronica Sinica, 2015, 43(1): 111-119.
DOI：

JU Tao, ZHU Zheng-dong, DONG Xiao-she. The Feature,Programming Model and Performance Optimization Strategy of Heterogeneous Many-Core System:A Review[J]. Acta Electronica Sinica, 2015, 43(1): 111-119. DOI： 10.3969/j.issn.0372-2112.2015.01.018.

摘要

异构众核系统已成为当前高性能计算领域重要的发展趋势.针对异构众核系统

从架构、编程、所支持的应用三方面分析对比当前不同异构系统的特点

揭示了异构系统的发展趋势及异构系统相对于传统多核并行系统的优势;然后从编程模型和性能优化方面分析了异构系统存在的问题和面临的挑战

以及国内外研究现状

结合当前研究存在的问题和难点

探讨了该领域进一步深入的研究方向;同时对两种典型的异构众核系统CPU+GPU和CPU+MIC进行不同应用类型的Benchmark测试

验证了两种异构系统不同的应用特点

为用户选择具体异构系统提供参考

在此基础上提出将两种众核处理器(GPU和MIC)结合在一个计算节点内构成新型混合异构系统;该新型混合异构系统可以利用两种众核处理器不同的处理优势

协同处理具有不同应用特点的复杂应用

同时分析了在该混合异构系统下必须要研究和解决的关键问题;最后对异构众核系统面临的挑战和进一步的研究方向进行了总结和展望.

Abstract

The heterogeneous many-core system has emerged as a promising developing trend in the high performance computing area.In this paper

we first revealed the developing trend and dominant position of the heterogeneous systems via analyzing their architectures

programming and application characteristics.Secondly

we discussed the programming model and performance optimization of current heterogeneous systems

and summarized the relative research trends.Thirdly

we verified the different application behaviors of the GPU and MIC heterogeneous system by running different types of Benchmark

which provides the reference for user to select the specific heterogeneous computing platform and

the basis of composing the hybrid heterogeneous system which combines the two types of many-core processor(GPU and MIC)into an individual computing node.This hybrid heterogeneous system can fully exploit the computing potential of the two types of many-core coprocessors to coordinate handling the complex application with different application behaviors.Finally

some open issues and future directions in the heterogeneous system were viewed.

关键词

Keywords

references

Views

下载量

CSCD

Alert me when the article has been cited

提交

Tools

Publicity Resources

Reading and Writing Performance Optimization of Cross-Language FUSE Framework

Research of Dynamic Proxy Based Context-Aware Programming Model EIPM

Transactional Memory Architecture Supporting I/O Operations within Transactions

A Data-Driven Uniform Programming Model for Reconfigurable Computing

Related Author

HUANG Yi-hua

DAI Hai-peng

WANG Zhao-kang

QIU Ling-wei

LUO Yi-li

GU Rong

FONT

FONT face

Related Institution

College of Computer Science and Technology， Nanjing University of Aeronautics and Astronautics

Department of Computer Science and Technology， Nanjing University

State Key Laboratory for Novel Software Technology， Nanjing University

（西安交通大学电信学院软件研究所，）

徐州师范大学数学科学学院江苏徐州

⁰