摘要
步入信息高速快车的时代,基于MPI的并行计算是解决物联网、图像处理获取数据,人工智能处理数据的重要途径之一;同时也是MIMD(多指令多机)并行计算实现高效、高速目标的重要策略与方法。笔者从MPI(Message-Passing-Interface消息传递接口)一个跨语言的通信协议机制,通过Linux或Windows操作系统平台,以C++语言平台为基础,论述并行计算实现几种方法。
In the era of information high-speed train, MPI-based parallel computing is one of the important ways to solve the Internet of Things, image processing to obtain data, and artificial intelligence to process data;It is also an important strategy and method for MIMD(multiple instruction multiple machine) parallel computing to achieve high efficiency and high speed. Starting from MPI(Message-Passing-Interface), a cross-language communication protocol mechanism, through Linux or Windows operating system platform, based on the C++ language platform,the author discusses several ways to realize parallel computing.
作者
郭福洲
GUO Fuzhou(Huanggang Polytechnic College,Huanggang,Hubei Province,438002 China)
出处
《科技资讯》
2021年第28期7-9,共3页
Science & Technology Information
关键词
MPI
并行计算
算法
进程
消息
通信
MPI
Parallel computing
Algorithm
Process
Message
Communication