On the Design of Abstract Binding Connectors for - ResearchGate在对抽象结合连接器设计研究之门.pptVIP

下载本文档

3
0
约1.09万字
约 31页
2018-06-24 发布于四川
举报
版权申诉

On the Design of Abstract Binding Connectors for - ResearchGate在对抽象结合连接器设计研究之门.ppt

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

On the Design of Abstract Binding Connectors for - ResearchGate在对抽象结合连接器设计研究之门

High-Level Service Connectors for Components-Based High Performance Computing Francisco Heron de Carvalho Junior, Ricardo Cordeiro Correa, Gisele Azevedo Araújo, Jefferson de Carvalho Silva Mestrado e Doutorado em Ciência da Computa??o (MDCC) Universidade Federal do Ceará Fortaleza, Brazil Rafael Dueire Lins Departamento de Eletr?nica e Sistemas Universidade Federal de Pernambuco Recife, Brazil Topics Introduction The # Component Model The # Programming Systems Client/Server bindings Peer-to-Peer bindings High-Order bindings Generic bindings Abstract bindings Conclusions and Further Works Topics Introduction The # Component Model The # Programming Systems Client/Server bindings Peer-to-Peer bindings High-Order bindings Generic bindings Abstract bindings Conclusions and Further Works Introduction Better abstractions are claimed for efficient parallel programming targeting contemporary High Performance Computing (HPC) architectures; Make parallel programmers more productive, but keep paralelism control at their hands; Better paralelism support is claimed for HPC component models and frameworks Fractal, GCM (Grid Component Model), CCA (Common Component Architecture); Be as expressive as MPI, but not adding intrinsical performance overheads; Motivations (more specifically ...) HPC component models and frameworks (Fractal, GCM, CCA) have approached parallelism support by some orthogonal extensions. Parallelism as a cooperation between components; SCMD (Single Component Multiple Data) – cohorts of components; More recently, MCMD (cooperative paralelism between components); Parallel concerns are scattered accross components; Adding new connectors: Collective interfaces, parallel remote method invocation, M?N interactions ... More complexity on the component model; Focus on efficient implementation and experimentation ... very pragmatic issues ... We propose a more generic and flexible approach without performance penalties Move parallel programming from a process-base