Wolfgang Schröder-preikschat, Forschungszentrum Informationstechnik Gmbh (gmd, Alexandre Navaux et al. · 2007
Motivation The parallel computing community has been using clusters of commodity workstations as an alternative to expensive massively parallel processors (MPP) for several years. While MPPs can rely on custom hardware to achieve high performance, their development follows a slow pace, mainly due to the small production scale. Clusters, in the other hand, benefit from the frenetic pace with which the workstation market evolves. As a matter of fact, when an MPP comes to the market, it is very likely that processor, memory and interconnection systems with similar features will already be available at the commodity market. Therefore, it seems evident that both technologies are going to converge into a single one. However, when we compare the performance of parallel applications running on clusters and on MPPs, the figures show a quite different scenario: clusters are still far behind their expensive relatives. The Numerical Aerospace Simulation Facility from NASA ha
Cited by 0
David J. Pritchard · ePrints Soton (University of Southampton) · 1994
Cited by 0
Gregor von Laszewski · OSTI OAI (U.S. Department of Energy Office of Scientific and Technical Information) · 2009
This article introduces an interactive parallel programming environment (IPPE) that simplifies the generation and execution of parallel programs. One of the tasks of the environment is to generate message-passing parallel programs for homogeneous and heterogeneous computing platforms. The parallel programs are represented by using visual objects. This is accomplished with the help of a graphical programming editor that is implemented in Java and enables portability to a wide variety of computer platforms. In contrast to other graphical programming systems, reusable parts of the programs can be stored in a program library to support rapid prototyping. In addition, runtime performance data on different computing platforms is collected in a database. A selection process determines dynamically the software and the hardware platform to be used to solve the problem in minimal wall-clock time. The environment is currently being tested on a Grand Challenge problem, the NASA four-dimensional data assimilation system.
Cited by 3Open access
E.J. Cameron, Daniel Cohen, B. Gopinath et al. · ACM SIGCOMM Computer Communication Review · 1988
The IC* project is an effort to create an environment for the design, specification and development of complex systems such as communication protocols, parallel machines and distributed systems. The basis of this work is the IC* model of parallel computation. In this model a system is specified by a set of invariant expressions which describe its behavior in time. The novel features of the model include: temporal and structural constraints, inherent parallelism, explicit modeling of time, nondeterministic evolution and dynamic activation. The project also includes the construction of a parallel computer specifically designed to support the model of computation. The model of computation, the status of the project and the user environment is the content of this paper.
Cited by 4
Nihat Adar, Selçuk Canbek, Erol Seke et al. · Lecture notes in computational science and engineering · 2008
Cited by 0
José Luiz Ribeiro Filho · UCL Discovery (University College London) · 1995
This thesis investigates an object-oriented programming environment for building parallel applications based on genetic algorithms (GAs). It describes the design of the Genetic Algorithms Manipulation Environment (GAME), which focuses on three major software development requirements: flexibility, expandability and portability. Flexibility is provided by GAME through a set of libraries containing pre-defined and parameterised components such as genetic operators and algorithms. Expandability is offered by GAME'S object-oriented design. It allows applications, algorithms and genetic operators to be easily modified and adapted to satisfy diverse problem's requirements. Lastly, portability is achieved through the use of the standard C++ language, and by isolating machine and operating system dependencies into low-level modules, which are hidden from the application developer by GAME'S application programming interfaces. The development of GAME is central to the Programming Environment for Applications of PArallel GENetic Algorithms project (PAPAGENA). This is the principal European Community (ESPRIT III) funded parallel genetic algorithms project. It has two main goals: to provide a general-purpose tool kit, supporting the development and analysis of large-scale parallel genetic algorithms (PGAs) applications, and to demonstrate the potential of applying evolutionary computing in diverse problem domains. The research reported in this thesis is divided in two parts: i) the analysis of GA models and the study of existing GA programming environments from an application developer perspective; ii) the description of a general-purpose programming environment designed to help with the development of GA and PGA-based computer programs. The studies carried out in the first part provide the necessary understanding of GAs' structure and operation to outline the requirements for the development of complex computer programs. The second part presents GAME as the result of combining development requirements, relevant features of existing environments and innovative ideas, into a powerful programming environment. The system is described in terms of its abstract data structures and sub-systems that allow the representation of problems independently of any particular GA model. GAME's programming model is also presented as general-purpose object-oriented framework for programming coarse-grained parallel applications. GAME has a modular architecture comprising five modules: the Virtual Machine, the Parallel Execution Module, the Genetic Libraries, the Monitoring Control Module, and the Graphic User Interface. GAME's genetic-oriented abstract data structures, and the Virtual Machine, isolates genetic operators and algorithms from low-level operations such as memory management, exception handling, etc. The Parallel Execution Module supports GAME's object- oriented parallel programming model. It defines an application programming interface and a runtime library that allow the same parallel application, created within the environment, to run on different hardware and operating system platforms. The Genetic Libraries outline a hierarchy of components implemented as parameterised versions of standard and custom genetic operators, algorithms and applications. The Monitoring Control Module supports dynamic control and monitoring of simulations, whereas the Graphic User Interface defines a basic framework and graphic 'widgets' for displaying and entering data. This thesis describes the design philosophy and rationale behind these modules, covering in more detail the Virtual Machine, the Parallel Execution Module and the Genetic Libraries. The assessment discusses the system's ability to satisfy the main requirements of GA and PGA software development, as well as the features that distinguish GAME from other programming environments.
Cited by 4
Jos'e Nagib Cotrim, Adam Beguelin, Bruce Lowekamp et al. · 1995
Writing parallel programs for distributed multi-user computing environments is a difficult task. The Distributed object migration environment (Dome) addresses three major issues of parallel computing in an architecture independent manner: ease of programming, dynamic load balancing, and fault tolerance. Dome programmers, with modest effort, can write parallel programs that are automatically distributed over a heterogeneous network, dynamically load balanced as the program runs, and able to survive compute node and network failures. This paper provides the motivation for and an overview of Dome, including a preliminary performance evaluation of dynamic load balancing for distributed vectors. Dome programs are shorter and easier to write than the equivalent programs written with message passing primitives. The performance overhead of Dome is characterized, and it is shown that this overhead can be recouped by dynamic load balancing in imbalanced systems. Finally, we show that a parallel ...
Cited by 0
Kai Tan, Edmonton Ab, Kai Tan et al. · 2003
copies of this thesis and to lend or sell such copies for private, scholarly or scientific research purposes only. The author reserves all other publication and other rights in association with the copyright in the thesis, and except as herein before provided, neither the thesis nor any substantial portion thereof may be printed or otherwise reproduced in any material form whatever without the author’s prior written permission. Date:
Cited by 0
Özalp Babaoglu, Alessandro Amoroso, Luigi Alberto Giachini et al. · Defense Technical Information Center (DTIC) · 1991
Modern distributed systems consisting of powerful workstations and high-speed interconnection networks are an economical alternative to special-purpose super computers. The technical issues that need to be addressed in exploiting the parallelism inherent in a distributed system include heterogeneity, high-latency communication, fault tolerance and dynamic load balancing. Current software systems for parallel programming provide little or no automatic support towards these issues and require users to be experts in fault-tolerant distributed computing. The Paralex system is aimed at exploring the extent to which the parallel application programmer can be liberated from the complexities of distributed systems. Paralex is a complete programming environment and makes extensive use of graphics to define, edit, execute and debug parallel scientific applications. All of the necessary code for distributing the computation across a network and replicating it to achieve fault tolerance and dynamic load balancing is automatically generated by the system. In this paper we give an overview of Paralex and present our experiences with a prototype implementation.
Cited by 10
Éric Violard · Parallel Processing Letters · 1994
This paper presents the main concepts of the mathematical theory PEI for parallel programming and emphasizes its derivation power. The mathematical basis of this theory leads to a nice implementation in CENTAUR 7 of an environment whose purpose is to transform parallel programs. It is illustrated by two similar examples: the convolution sum and the Dirichlet product. The second one uses non-affine dependencies that can be easily taken into account using PEI.
Cited by 4
Waldemar Ruggiero Júnior · LA Referencia (Red Federada de Repositorios Institucionales de Publicaciones Científicas) · 2007
Esse trabalho propõe uma solução de um Data Warehouse distribuído para implementação de um sistema de CRM, Customer Relationship Management, com o objetivo de uso no ambiente bancário possibilitando que ações de relacionamento com clientes sejam planejadas e implementadas. É proposta uma arquitetura para implementação de um Data Warehouse em ambiente distribuído, utilizando programação paralela. Com o aumento no volume de dados armazenados nos Data Warehouse, as arquiteturas tradicionais exigem processadores e sistemas de entrada e saída cada vez mais robustos em termos de desempenho. Isso fica mais bem caracterizado quando são executadas procuras complexas (Ad hoc). A utilização de ambientes distribuídos em conjunto com programação paralela é uma alternativa para redução de custo e aumento de desempenho. É apresentada, aqui, uma proposta de arquitetura de Data Warehouse distribuído, integrado ao uso de programação paralela.
Cited by 0
Our Ref, Mark Debbage, Mark B. Hill · 1992
Introduction Parallel programming is in a state of flux. The choice of parallel programming paradigms and the presentation of concurrency to the applications programmer are still matters of intense research. Fortunately the cost of prototyping and experimenting with programming environments can be dramatically reduced using a low-level toolkit. Furthermore, standard low-level interfaces enable the rapid porting of higher-level code to new platforms. ParaPET, the Parallel Programming Environment Toolkit, has been developed to fulfill this role. It is based upon the lightweight communications infrastructure of the Virtual Channel Router[1] (VCR) package. The VCR system is readily available from the University of Southampton Parallel Application Centre and is distributed for a nominal cost to cover administration and media overheads 1 . VCR presents a distributed occam programming environment which dramatically simplifies the configuration of con
Cited by 0
G. T. Nguyen, V. D. Tran, M. Kotocova · 2002
Abstract. In this paper, fault-tolerant feature of TOPAS parallel programming environment for distributed systems is presented. TOPAS automatically analyzes data dependence among tasks and synchronizes data, which reduces the time needed for parallel program developments. TOPAS also provides supports for scheduling, load balancing and fault tolerance. The main topics of this paper is to present the solution for transparent recovery of asynchronous distributed computation on clusters of workstations without hardware spare when a fault occurs on a node. Experiments show simplicity and efficiency of parallel programming in TOPAS environment with fault-tolerant integration, which provides graceful performance degradation and quick reconfiguration time for application recovery. 1.
Cited by 0
Aldinucci M., Danelutto M., Teti P. · ISTI Open Portal · 2003
In this work we present Lithium, a pure Java structured parallel programming environment based on skeletons (common, reusable and efficient parallelism exploitation patterns). Lithium is implemented as a Java package and represents both the first skeleton based programming environment in Java and the first complete skeleton based Java environment exploiting macro data flow implementation techniques. Lithium supports a set of user code optimizations which are based on skeleton rewriting techniques. These optimizations improve both absolute performance and resource usage with respect to original user code. Parallel programs developed using the library run on any network of workstations provided the workstations support plain JRE. The paper describes the library implementation, outlines the optimization techniques used and eventually presents the performance results obtained on both synthetic and real applications.
Cited by 0
Reiner W. Hartenstein, Jürgen Becker, Michael Herz et al. · 2002
The paper presents a partitioning and parallelizing programming environment for a novel parallel architecture. This universal embedded accelerator is based on a reconfigurable datapath hardware. The partitioning and parallelizing programming environment accepts C programs and carries out both a profiling-driven host/accelerator partitioning for performance optimization in a first step, and in a second step a resource-driven sequential/structural partitioning of the accelerator source code to optimize the utilization of its reconfigurable resources.
Cited by 8
José Nagib Cotrim Árabe, Adam Beguelin, Bruce Lowekamp et al. · 1995
Writing parallel programs for distributed multi-user computing environments is a difficult task. The Distributed object migration environment (Dome) addresses three major issues of parallel computing in an architecture independent manner: ease of programming, dynamic load balancing, and fault tolerance. Dome programmers, with modest effort, can write parallel programs that are automatically distributed over a heterogeneous network, dynamically load balanced as the program runs, and able to survive compute node and network failures. This paper provides the motivation for and an overview of Dome, including a preliminary performance evaluation of dynamic load balancing for distributed vectors. Dome programs are shorter and easier to write than the equivalent programs written with message passing primitives. The performance overhead of Dome is characterized, and it is shown that this overhead can be recouped by dynamic load balancing in imbalanced systems. Finally, we show that a parallel program can be made failure resilient through Dome''s architecture independent checkpoint and restart mechanisms.
Cited by 90
G. Hurteau, V. Van Dongen, G. Gao et al. · 1995
EPPP is an Environment for Portable Parallel Programming (EPPP) for current and future generation parallel computers. It is portable in the sense that the user can develop and tune his/her application on single workstations and rapidly port it and run it efficiently on a variety of parallel distributed-memory machines. In order to achieve this goal, EPPP consists of mainly four components that are integrated together: a programming language based on the data-parallel programming paradigm, an advanced compilation technology, a simulator, and a performance debugger. The topic of this paper is to present an overview of these components, their current status and their future development. 1 Background A program is said to be portable when it can easily be compiled and executed efficiently on various platforms. Portability is known to be a major concern for people like scientists and engineers who have developped very large software over the years and are pursuing their development for furt...
Cited by 0
Francisco Heron de Carvalho-Junior, Rafael Dueire Lins, Ricardo Massa Ferreira Lima · 2001
In this paper, we present the parallelization of a sequential functional implementation of a Monte Carlo Transport Problem, called MCP- Haskell[Hammes et al., 1995], using Haskell# . This experiment gave us important feedback for evaluating Haskell# features, helping us to answer some questions, like how expressive is Haskell# for representing known parallel computational patterns, how easy it is to build large scale parallel programs in an elegant and concise way, and how efficient are Haskell# programs. Based on our conclusions, we suggest new features to be incorporated in Haskell# to improve its expressiveness and performance. We also present the performance figures for the MCP-Haskell# benchmark.
Cited by 0
James C. Browne, T. Lee, John Werth · IEEE Transactions on Software Engineering · 1990
Reports on the initial experimental evaluation of ROPE (reusability-oriented parallel programming environment), a software component reuse system. ROPE helps the designer find and understand components by using a new classification method called structured relational classification. ROPE is part of a development environment for parallel programs which uses a declarative/hierarchical graphical programming interface. This interface allows use of components with different levels of abstraction, ranging from design units to actual code modules. ROPE supports reuse of all the component types defined in the development environment. Programs developed with the aid of ROPE were found to have error rates far less than those developed without ROPE.< <ETX xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">></ETX>
Cited by 47
Johan Ringström, Peter Fritzson, Johan Fagerström · Microprocessing and Microprogramming · 1992
Cited by 2