检索结果-内蒙古大学图书馆

International Computer Symposium (ICS)

作者： Chou, Chau-Yi Chen, Kuen-Tsann Natl Chung Hsing Univ Dept Appl Math Taichung Taiwan Natl Ctr High Performance Comp Hsinchu Taiwan

ISBN: (纸本)9781509034383

The chip heat dissipations defeat the clock speed increment. Multi-core clusters and the heterogeneous platforms including accelerators become a main trend recently. parallel programming paradigms surfs on these diverse platforms: CUDA C, CUDA Fortran, OpenCL, OpenACC, OpenMP, MPI, pthread, MapReduce, and so on. The quantitative performance indexes help get a good picture of parallel programming paradigms for the applications. This study employ two examples: Pennes bioheat equations to simulating local hyperthermia destroying tumor cells and Navier-Stokes equations to simulating driven cavity flow at high Reynolds numbers via parallel programming paradigms: CUDA C, CUDA Fortran, OpenMP and MPI. parallel programming in MPI for Pennes bioheat equations shows super-linear speedup on NCHC (National Center for High-performance Computing) ALPS and significantly faster than the original author, whereas parallel programming in CUDA C framework for Navier-Stokes equations achieves around 24 times speedup on a NVIDIA C1060 GPU. We hope these results to support useful suggestions.

关键词： parallel programming paradigms bioheat transfer model Pennes equation Navier-Stokes equations performance evaluation

来源：评论

学校读者我要写书评

暂无评论

programming MASSIVELY-parallel ARCHITECTURES WITH SEQUENTIAL OBJECT-ORIENTED LANGUAGES

引用

FUTURE GENERATION COMPUTER SYSTEMS 1994年第1期10卷 59-70页

作者： JEZEQUEL, JM BERGHEUL, F ANDRE, F I.R.I.S.A. Campus de Beaulieu F-35042 Rennes Cedex France

Most parallel object oriented languages (OOL) are currently using a general parallelism model based on communicating sequential processes. This approach makes it difficult to program massively parallel systems in an easy and efficient way. So we propose to use another form of parallelism, known as data parallelism. We describe how it can be integrated in a given OOL in a clean and elegant fashion, using only already existing concepts - i.e. without modifying the OOL syntax and semantics - to exploit the potential power of massively parallel systems. To illustrate our ideas, we present an application of this approach to a well known parallel paradigm, along with experimental performance results.

关键词： MASSIVELY parallel SYSTEMS parallel programming paradigms OBJECT ORIENTED LANGUAGES

来源：评论

学校读者我要写书评

暂无评论

The Fork95 parallel programming language: Design, implementation, application

引用

INTERNATIONAL JOURNAL OF parallel programming 1997年第1期25卷 17-50页

作者： Kessler, CW Seidl, H Fachbereich IV-Informatik Universität Trier Trier Germany

Fork95 is an imperative parallel programming language intended to express algorithms for synchronous shared memory machines (PRAMs). It is based on ANSI C and offers additional constructs to hierarchically divide processor groups into subgroups and manage shared and private address subspaces. Fork95 makes the assembly-level synchronicity of the underlying hardware available to the programmer at the language level. Nevertheless, it supports locally asynchronous computation where desired by the programmer. We present a one pass compiler, fee, which compiles Fork95 and C programs to the SB-PRAM machine. The SB-PRAM is a lock-step synchronous, massively parallel multiprocessor currently being built at Saarbrucken University, with a physically shared memory and uniform memory access time. We examine three important types of parallel computation Frequently used for the parallel solution of real-world problems. While farming and parallel divide-and-conquer are directly supported by Fork95 language constructs, pipelining can be easily expressed using existing language Features;an additional language construct for pipelining is not required.

关键词： Fork95 parallel programming language synchronous program execution PRAM parallel programming paradigms

来源：评论

学校读者我要写书评

暂无评论

Performance models for the processor farm paradigm

引用

IEEE TRANSACTIONS ON parallel AND DISTRIBUTED SYSTEMS 1997年第5期8卷 475-489页

作者： Wagner, AS Sreekantaswamy, HV Chanson, ST NORTEL NO TELECOM STN C OTTAWA ON K1Y 4H7 CANADA HONG KONG UNIV SCI & TECHNOL KOWLOON HONG KONG

In this paper, we describe the design, implementation, and modeling of a runtime kernel to support the processor farm paradigm on multicomputers. We present a general topology-independent framework for obtaining performance models to predict the performance of the start-up, steady-state, and wind-down phases of a processor farm. An algorithm is described, which for any interconnection network determines a tree-structured subnetwork that optimizes farm performance. The analysis technique is applied to the important case of k-ary tree topologies. The models are compared with the measured performance on a variety of topologies using both constant and varied task sizes.

关键词： parallel programming paradigms performance evaluation processor farm tree networks message passing architecture network flow master-slave

来源：评论

学校读者我要写书评

暂无评论

EXTENSIBILITY AND REUSE OF OBJECT-ORIENTED SYNCHRONIZATION COMPONENTS 4th

EXTENSIBILITY AND REUSE OF OBJECT-ORIENTED SYNCHRONIZATION C...

引用

4TH INTERNATIONAL PARLE CONF ( parallel ARCHITECTURES AND LANGUAGES EUROPE )

作者： THOMAS, L Centre de Recherche en lnformatique de Nancy (CRIN) Campus Scientifique Boîte Postale No. 239 Vandceuvre-lès-Nancy 54506 France

ISBN: (纸本)3540555994

Synchronization components in an object oriented framework are hard to express because of their conflicting and interfering aims with the reuse and extension concepts. After surveying the state of the art, we identify two problems: the synchronization problem deals with acceptance of an activity based solely on the current executing activities and those wishing to do so, and the behavior problem deals with the availability of methods based solely on the state of the object. The allowance for an activity to execute is then expressed by two separated denotations, each handled separately with respect to the extensibility and reusabilty mechanism, whatever it may be. © Springer-Verlag Berlin Heidelberg 1992.

关键词： parallel programming paradigms parallel LANGUAGE EXTENSIONS OBJECT-ORIENTED LANGUAGE CONCURRENCY SYNCHRONIZATION INHERITANCE REUSE

来源：评论

学校读者我要写书评

暂无评论

Automatic parallelization Tools

Automatic Parallelization Tools

引用

World Congress on Engineering and Computer Science

作者： Qian, Ying King Abdullah Univ Sci & Technol KAUST Supercomp Lab Thuwal 239556900 Saudi Arabia

ISBN: (纸本)9789881925169

ln recent years, much has been made of the computing industry's widespread shift to parallel computing. Nearly all consumer computers will ship with multicore processors. parallel computing will no longer be only relegated to exotic supercomputers or mainframes, moreover, electronic devices such as mobile phones and other portable devices have begun to incorporate parallel computing capabilities. High performance computing becomes increasingly important. To maintain high quality solutions, programmers have to efficiently parallelize and map their algorithms. This task is far from trivial, especially for different existing parallel computer architectures and different parallel programming paradigms. To reduce the burden on programmers, automatic parallelization tools are introduced. The purpose of this paper is to discuss different automatic parallelization tools for different parallel architectures.

关键词： parallel processing automatic parallelization tool parallel programming paradigms

来源：评论

学校读者我要写书评

暂无评论

RELIABLE COMMUNICATION IN VPL 4th

RELIABLE COMMUNICATION IN VPL

引用

4TH INTERNATIONAL PARLE CONF ( parallel ARCHITECTURES AND LANGUAGES EUROPE )

作者： KUHN, E PUNTIGAM, F University of Technology Vienna Institute of Computer Languages Argentinierstr. 8 Vienna A-1040 Austria

ISBN: (纸本)3540555994

We compare different degrees of architecture abstraction and communication reliability in distributed programming languages. A nearly architecture independent logic programming language and system with reliable communication, called VPL (Vienna parallel Logic) is presented. We point out the contradiction between complete architecture independence and reliable high-level communication in programming languages. The description of an implementation technique of VPL’s reliable communication on shared memory and distributed architectures comprises a major part of this paper. The reliable communication is based on the specification of consistent system states that must survive system failures. © Springer-Verlag Berlin Heidelberg 1992.

关键词： parallel programming paradigms parallel LANGUAGE EXTENSIONS IMPLEMENTATION OF parallel LANGUAGES

来源：评论

学校读者我要写书评

暂无评论

建议与咨询留下您的常用邮箱和电话号码，以便我们向您反馈解决方案和替代方法

时间限定

文献类型

馆藏选择

核心期刊

语言

文献类型

帮助

文字说明：

检索规则说明：

检索范例：

分类表

所选分类

限定检索结果

文献类型

馆藏范围

日期分布

学科分类号

主题

机构

作者

语言

请选择保存的检索档案：

请选择收藏分类：

通借通还

建议与咨询 留下您的常用邮箱和电话号码，以便我们向您反馈解决方案和替代方法

时间限定

文献类型

馆藏选择

核心期刊

语言

文献类型

帮助

文字说明：

检索规则说明：

检索范例：

分类表

所选分类

限定检索结果

文献类型

馆藏范围

日期分布

学科分类号

主题

机构

作者

语言

请选择保存的检索档案： 新增检索档案 确定 取消

请选择收藏分类： 新增自定义分类 确定 取消

通借通还

建议与咨询留下您的常用邮箱和电话号码，以便我们向您反馈解决方案和替代方法

请选择保存的检索档案：

请选择收藏分类：