检索结果-内蒙古大学图书馆

Compile-time estimation of communication costs for data parallel programs

JOURNAL OF parallel AND DISTRIBUTED COMPUTING 1996年第1期39卷 46-65页

作者： Fahringer, T Institute for Software Technology and Parallel Systems University of Vienna Liechtensteinstrasse 22 A-1090 Vienna Austria

Most of the current compiler projects for distributed memory architectures leave the critical and time-consuming problem of finding performance-efficient data distributions and profitable program transformations for a given parallel program almost entirely to the programmer. Performance estimators provide critical performance information to both programmers and parallelizing compilers, the most crucial part of which involves determining the communication overhead induced by a program. In this paper, we present a very practical approach to the problem of compile-time estimation of communication costs for regular codes that includes analytical methods to model the number of messages exchanged, data volume transferred, transfer time, and network contention. In order to achieve high estimation accuracy, our estimator aggressively exploits compiler analysis and optimization information. It is assumed that machine parameters and problem size are known at compile time. We conducted a variety of experiments to validate the estimation accuracy and the ability to support both the programmer and compiler in the effort of performance tuning of parallel programs. We believe that our approach can be automatically applied to a large class of regular codes. (C) 1996 Academic Press, Inc.

关键词：

来源：评论

学校读者我要写书评

暂无评论

Vienna Fortran and the path towards a standard parallel language

引用

IEICE TRANSACTIONS ON INFORMATION AND systems 1997年第4期E80D卷 409-416页

作者： Chapman, BM Mehrotra, P Zima, HP Institute for Software Technology and Parallel Systems University of Vienna A-1090 Vienna Liechtensteinstr. 22 Austria

Highly parallel scalable multiprocessing systems (HMPs) are powerful tools for solving large-scale scientific and engineering problems. However, these machines are difficult to program since algorithms must exploit locality in order to achieve high performance. Vienna Fortran was the first fully specified data-parallel language for HMPs that provided features for the specification of data distribution and alignment at a high level of abstraction. In this paper we outline the major elements of Vienna Fortran and compare it to High Performance Fortran (HPF), a de-facto standard in this area. A significant weakness of HPF is its lack of support for many advanced applications, which require irregular data distributions and dynamic load balancing. We introduce HPF+, an extension of HPF based on Vienna Fortran, that provides the required functionality.

关键词： high performance computing distributed-memory architectures data distribution, parallelization Fortran extensions

来源：评论

学校读者我要写书评

暂无评论

SUPERB AND VIENNA-FORTRAN

引用

parallel COMPUTING 1994年第10-11期20卷 1487-1517页

作者： ZIMA, HP BREZANY, P CHAPMAN, BM Institute for Software Technology and Parallel Systems University of Vienna Br&uuml nner Strasse 72 A-1210 Vienna Austria

Distributed-memory systems are powerful tools for solving large-scale scientific and engineering problems. However these machines are difficult to program since the data have to be distributed across the processors and message-passing operations must be inserted for communicating non-local data. In this paper, we discuss SUPERB and Vienna Fortran, two related developments with the objective of providing the user with a higher level programming paradigm while not sacrificing target code performance. The parallelization system SUPERB was developed in the German supercomputer project SUPRENUM from 1985 to 1989. It is based on the Single-Program-Multiple-Data (SPMD) paradigm, allows the use of global addresses, and automatically inserts the necessary communication statements, given a user-supplied data distribution. SUPERB was the first implemented system that translated sequential Fortran 77 into explicitly parallel message-passing Fortran. As a result of the experiences with SUPERB and related research, the language Vienna Fortran was designed within the ESPRIT project GENESIS, in a joint effort of the University of Vienna and ICASE, Nasa Langley Research Center. Vienna Fortran is a machine-independent language extension to Fortran, which includes a broad range of features for the high-level support of advanced application development for distributed-memory multiprocessors. It has significantly influenced the development of High Performance Fortran, a first attempt of language standardization in this area.

关键词： DISTRIBUTED-MEMORY MULTIPROCESSOR systems NUMERICAL COMPUTATION DATA parallel ALGORITHMS DATA DISTRIBUTION PROGRAM ANALYSIS OPTIMIZATION

来源：评论

学校读者我要写书评

暂无评论

VFC: the Vienna Fortran compiler

引用

Scientific Programming 1999年第1期7卷 67-81页

作者： Benkner, Siegfried Institute for Software Technology and Parallel Systems University of Vienna Liechtenstein Strasse 22 A‐1090 Vienna Austria

High Performance Fortran (HPF) offers an attractive high-level language interface for programming scalable parallel architectures providing the user with directives for the specification of data distribution and delegating to the compiler the task of generating an explicitly parallel program. Available HPF compilers can handle regular codes quite efficiently, but dramatic performance losses may be encountered for applications which are based on highly irregular, dynamically changing data structures and access patterns. In this paper we introduce the Vienna Fortran Compiler (VFC), a new source-to-source parallelization system for HPF+, an optimized version of HPF, which addresses the requirements of irregular applications. In addition to extended data distribution and work distribution mechanisms, HPF+ provides the user with language features for specifying certain information that decisively influence a program's performance. This comprises data locality assertions, non-local access specifications and the possibility of reusing runtime-generated communication schedules of irregular loops. Performance measurements of kernels from advanced applications demonstrate that with a high-level data parallel language such as HPF+ a performance close to hand-written message-passing programs can be achieved even for highly irregular codes.

关键词： High Performance Fortran data parallel programming data distribution parallelizing compilers communication schedule reuse

来源：评论

学校读者我要写书评

暂无评论

High-performance VLSI model elliptic solvers

High-performance VLSI model elliptic solvers

引用

International Conference and Exhibition on High-Performance Computing and Networking

作者： Vajtersic, M Institute for Software Technology and Parallel Systems University of Vienna Liechtensteinstr. 22 Vienna A-1092 Austria

ISBN: (纸本)3540593934

VLSI parallel algorithms for a solution of fundamental elliptic problems with Laplace operators (Dirichlet and first boundary value problem for Poisson and biharmonic equation respectively) on a rectangular N x N grid are proposed. A standard multigrid algorithm is adopted for Poisson equation which allows a parallel solution of this problem in T = O(logN) parallel steps. A special network consisting of N x N processor elements and of O(NlogN) interconnection lines in each direction results in a design the area of which is A = O(N(2)log(2)N). AT(2) estimation for a complexity of this Poisson solver is O(N(2)log(4)N) which improves the best result known until now by a factor of O(N/logN). This VLSI multigrid Poisson solver is applied to the semidirect method for solving the biharmonic equation. The parallel time of the algorithm is O(root-log(2)N) and the area needed is A = O(N(3)logN). The total complexity for such VLSI semidirect solver is AT(2) = O(N(4)log(5)N).

关键词： Boundary value problems

来源：评论

学校读者我要写书评

暂无评论

An introduction to HPF+ project 2nd

引用

2nd International Symposium on High Performance Computing, ISHPC 1999

作者： Zima, Hans P. University of Vienna Institute for Software Technology and Parallel Systems Liechtensteinstrasse 22 Vienna Austria

来源：评论

学校读者我要写书评

暂无评论

Simulation software engineering: Experiences and challenges 3

Simulation software engineering: Experiences and challenges

引用

3rd International Workshop on software Engineering for High Performance Computing in Computational Science and Engineering, SEHPCCSE 2015 - co-located with Supercomputing 2015

作者： Wagner, Stefan Pflüger, Dirk Mehl, Miriam Institute of Software Technology University of Stuttgart Germany Institute for Parallel and Distributed Systems University of Stuttgart Germany

ISBN: (纸本)9781450340120

Using software for large-scale simulations has become an important research method in many disciplines. With increasingly complex simulations, simulation software becomes a valuable assest. Yet, the quality of many simulation codes is worrying. In this paper, we want to collect and structure the challenges for a systematic simulation software engineering as a reference and the basis for further research. We describe our own experiences with developing simulation software and collaborating with non-computer-scientists. We complement our experienced challenges with a brief literature review. We structured the challenges for simulation software engineering into six areas: motivation and recognition;education and training;developer turnover;software length of life;verification, validation and debugging;and efficiency vs. maintainability. Overcoming these challenges needs efforts from research agencies, scientific computing researchers as well as software engineering researchers. © 2015 ACM.

关键词： Verification

来源：评论

学校读者我要写书评

暂无评论

Advanced optimizations for parallel irregular out-of-core programs 3rd

引用

3rd International Workshop on Applied parallel Computing in Industrial Problems and Optimization, PARA 1996

作者： Brezany, Peter Dang, Minh Institute for Software Technology and Parallel Systems University of Vienna Liechtensteinstrasse 22 ViennaA-1090 Austria

ISBN: (纸本)3540620958

Large scale irregular applications involve data arrays and other data structures that are too large to fit in main memory and hense reside on disks. This paper presents a method for implementing this kind of applications on distributed-memory systems. The method is based on a runtime system that has been built on top of the CHAOS library. © Springer-Verlag Berlin Heidelberg 1996.

关键词： Computers

来源：评论

学校读者我要写书评

暂无评论

Supercompilers for massively parallel architectures 1

Supercompilers for massively parallel architectures

引用

1st Aizu International Symposium on parallel Algorithms/Architecture Synthesis, AISPAS 1995

作者： Chapman, Barbara Pantano, Mario Zima, Hans Institute for Software Technology and Parallel Systems University of Vienna Brunner Strasse 72 ViennaA-1210 Austria

ISBN: (纸本)081867038X

We describe the design of a compilation system, which translates Fortran programs automatically into explicitly parallel programs for a massively parallel architecture. Such a compiler must automatically generate data distributions and derive program transformation strategies based on a combination of heuristics and analysis features. Programs are subjected to an iterative tuning process, involving three major components: a restructuring system, a performance analysts subsystem, and a parallelization support environment. Many of the components in this environment have already been implemented in the framework of the Vienna Fortran compilation system. © 1995 IEEE.

关键词： parallel architectures

来源：评论

学校读者我要写书评

暂无评论

Processing array statements and procedure interfaces in the prepare high PERFORMANCE fortran compiler 5th

Processing array statements and procedure interfaces in the ...

引用

5th International Conference on Compiler Construction, CC 1994

作者： Benkner, Siegfried Brezany, Peter Zima, Hans Institute for Software Technology and Parallel Systems University of Vienna Brünnerstrasse 72 ViennaA-1210 Austria

ISBN: (纸本)9783662466629

Recently, a standard set of extensions for Fortran 90, called High Performance Fortran (HPF), has been developed which would provide a portable interface to a wide variety of parallel architectures. HPF focuses mainly on issues of distributing data across the memories of a distributed memory multiprocessor. This paper proposes techniques for processing HPF data distribution and data alignment specifications, array statements and procedure interfaces in the HPF compiler which is being developed in the ESPRIT project Prepare. © Springer-Verlag Berlin Heidelberg 1994.

关键词： Data handling

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：