基于linux的小型集群的分析与实现-analysis and implementation of small cluster based on linux.docxVIP

下载本文档

0
0
约5.09万字
约 73页
2018-05-18 发布于上海
举报
版权申诉

基于linux的小型集群的分析与实现-analysis and implementation of small cluster based on linux.docx

1、原创力文档（book118）网站文档一经付费（服务费），不意味着购买了该文档的版权，仅供个人/单位学习、研究之用，不得用于商业用途，未经授权，严禁复制、发行、汇编、翻译或者网络传播等，侵权必究。。
2、本站所有内容均由合作方或网友上传，本站不对文档的完整性、权威性及其观点立场正确性做任何保证或承诺！文档内容仅供研究参考，付费前请自行鉴别。如您付费，意味着您自己接受本站规则且自行承担风险，本站不退款、不进行额外附加服务；查看《如何避免下载的几个坑》。如果您已付费下载过本站文档，您可以点击这里二次下载。
3、如文档侵犯商业秘密、侵犯著作权、侵犯人身权等，请点击“版权申诉”（推荐），也可以打举报电话：400-050-0827(电话支持时间：9:00-18:30)。
4、该文档为VIP文档，如果想要下载，成为VIP会员后，下载免费。
5、成为VIP后，下载本文档将扣除1次下载权益。下载后，不支持退款、换文档。如有疑问请联系我们。
6、成为VIP后，您将拥有八大权益，权益包括：VIP文档下载权益、阅读免打扰、文档格式转换、高级专利检索、专属身份标志、高级客服、多端互通、版权登记。
7、VIP文档为合作方或网友上传，每下载1次，网站将根据用户上传文档的质量评分、类型等，对文档贡献者给予高额补贴、流量扶持。如果你也想贡献VIP文档。上传文档

基于linux的小型集群的分析与实现-analysis and implementation of small cluster based on linux

AbstractTraditional single-processor computing can’t meet computing requirements, and at the same time, the high-performance computer based on vector processor and symmetric multi-processors has many problems, such as expensive, low scalability, poor expansibility. In order to accelerate the speed of solving the problem, the paper builds a cluster system under Liunx environment.The system has good scalability, availability and cost-effective.A Linux-based high-performance cluster includes many software modules, such as NFS file sharing system, MPI parallel programming library and so on. And then the paper compares the serial and parallel algorithms, which calculate the value of PI by using the C + MPI programming mode.The numerical results demonstrate the feasibility of parallel computing and the advantages of parallel algorithm.In order to solve the degree of parallelism caused by different matrix division method, the paper uses two methods (the striped matrix multiplication parallel algorithm and the checkerboard divided matrix multiplication parallel algorithm) to analyze the computation time and other performance indicators. The experimental results show that the checkerboard division method can develop a higher degree of parallelism than the striped method.The thesis analyzes the parameters of HPL test which includes the dimension of the coefficient matrix in solving dense linear equations, the LU decomposition block size, processor mesh size, the number of processors participating in the calculation and other parameters of HPL package. The paper evaluates the floating point performance of cluster system by adjusting the parameters in HPL.dat. And the Optimal floating point performance of the cluster system is 13.61Gflops.Key words: parallel computing, computer clusters, MPI, Benchmark, HPL, matrix multiplication目录第一章绪论 ..............................................................................................................................................