- 1、本文档共63页,可阅读全部内容。
- 2、有哪些信誉好的足球投注网站(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。
- 3、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。
- 4、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)。
查看更多
Matrix Multiplication参考
Parallel Programmingin C with MPI and OpenMP Michael J. Quinn Chapter 11 Matrix Multiplication Outline Sequential algorithms Iterative, row-oriented Recursive, block-oriented Parallel algorithms Rowwise block striped decomposition Cannon’s algorithm Iterative, Row-oriented Algorithm Performance as n Increases Reason:Matrix B Gets Too Big for Cache Block Matrix Multiplication Recurse Until B Small Enough Comparing Sequential Performance First Parallel Algorithm Partitioning Divide matrices into rows Each primitive task has corresponding rows of three matrices Communication Each task must eventually see every row of B Organize tasks into a ring First Parallel Algorithm (cont.) Agglomeration and mapping Fixed number of tasks, each requiring same amount of computation Regular communication among tasks Strategy: Assign each process a contiguous group of rows Communication of B Communication of B Communication of B Communication of B Complexity Analysis Algorithm has p iterations During each iteration a process multiplies(n / p) ? (n / p) block of A by (n / p) ? n block of B: ?(n3 / p2) Total computation time: ?(n3 / p) Each process ends up passing(p-1)n2/p = ?(n2) elements of B Isoefficiency Analysis Sequential algorithm: ?(n3) Parallel overhead: ?(pn2)Isoefficiency relation: n3 ? Cpn2 ? n ? Cp This system does not have good scalability Weakness of Algorithm 1 Blocks of B being manipulated have p times more columns than rows Each process must access every element of matrix B Ratio of computations per communication is poor: only 2n / p Parallel Algorithm 2(Cannon’s Algorithm) Associate a primitive task with each matrix element Agglomerate tasks responsible for a square (or nearly square) block of C Computation-to-communication ratio rises to n / ?p Elements of A and B Needed to Compute a Process’s Portion of C Blocks Must Be Aligned Blocks Need to Be Aligned Rearrange Blocks Consider Process P1,2 Consider Process P1,2 Consider Process P1,2 Consider Process
您可能关注的文档
- IQC来料管理和SQE供应商管理培训学习资料(全面而且实用)参考.ppt
- Learning About Return and Risk from the Historical Record参考.ppt
- LED灯具公司物料系统编码规则参考.doc
- Leveraging the Business参考.ppt
- Libnids在商用多核系统上的并行化详细设计说明书参考.doc
- linux下文件压缩与解压的分析研究参考.doc
- Introduction to CloudStack 3.0 Technical Presentation参考.pptx
- Loan Amortization;Mortgage loan amortization schedule参考.ppt
- Linux进程间通信的研究与实现参考.doc
- loan balances and details of individual payments参考.ppt
- 2024年证券分析与咨询服务项目投资申请报告代可行性研究报告.docx
- 2024年铬酸酐项目资金申请报告代可行性研究报告.docx
- 2024年清洁胶项目资金申请报告代可行性研究报告.docx
- 2024年肉松饼项目投资申请报告代可行性研究报告.docx
- 2024年陆上泵项目资金需求报告代可行性研究报告.docx
- 2024年未硫化复合橡胶及其制品项目资金需求报告代可行性研究报告.docx
- 2024年精密温控节能设备项目资金筹措计划书代可行性研究报告.docx
- 2024年汽车覆盖件模具项目资金筹措计划书代可行性研究报告.docx
- 宋词行书钢笔字帖.pdf
- 我的暑假生活作文三年级300字10篇.pdf
文档评论(0)