PVM parallel programming model provides a convenient methodology of creating dynamic master/worker applications. In this paper, we introduce the benefits from the use of KappaPi tool for automatic analysis of master/w...
详细信息
ISBN:
(纸本)3540410104
PVM parallel programming model provides a convenient methodology of creating dynamic master/worker applications. In this paper, we introduce the benefits from the use of KappaPi tool for automatic analysis of master/worker applications. First, by the automatic detection of the master/worker paradigm in the application. And second, by the performance analysis of the application focusing on the performance bottlenecks and the limitations of this master/worker collaboration.
parallel programming with distributed object technology becomes increasingly popular but shared-memory programming is still a common way of utilizing parallel machines. In fact, both models can coexist fairly well and...
详细信息
ISBN:
(纸本)354067442X
parallel programming with distributed object technology becomes increasingly popular but shared-memory programming is still a common way of utilizing parallel machines. In fact, both models can coexist fairly well and software DSM systems can be constructed easily using distributed object systems. In this paper, we describe the construction of a hybrid programming platform based on the ARTS distributed object system. We describe how an object-oriented design approach provides a compact and flexible description of the system components. A sample implementation demonstrates that three classes of less than 100 lines of code each suffice to implement sequential consistency.
作者:
Benkner, SBrandes, TUniv Vienna
Inst Software Engn A-1090 Vienna Austria GMD
German Natl Res Ctr Informat Technol SCAI D-53754 St Augustin Germany
The OpenMP Application Program Interface supports parallel programming on scalable symmetric multiprocessor machines (SMP) with a shared memory by providing the user with simple work-sharing directives for C/C++ and F...
详细信息
ISBN:
(纸本)3540679561
The OpenMP Application Program Interface supports parallel programming on scalable symmetric multiprocessor machines (SMP) with a shared memory by providing the user with simple work-sharing directives for C/C++ and Fortran so that the compiler can generate parallel programs based on thread parallelism. However, the lack of language features for exploiting data locality often results in poor performance since the non-uniform memory access times on scalable SMP machines cannot be neglected. HPF, the de-facto standard for data parallel programming, offers a rich set of data distribution directives in order to exploit data locality, but has mainly been targeted towards distributed memory machines. In this paper we describe an optimized execution model for HPF programs on SMP machines that avails itself with the mechanisms provided by OpenMP for work sharing and thread parallelism while exploiting data locality based on user-specified distribution directives. This execution model has been implemented in the ADAPTOR HPF compilation system and experimental results verify the efficiency of the chosen approach.
This paper presents a flexible and effective model for object-oriented parallel programming in both local and wide area contexts and its implementation as a Java package. Blending remote evaluation and active messages...
详细信息
ISBN:
(纸本)354067442X
This paper presents a flexible and effective model for object-oriented parallel programming in both local and wide area contexts and its implementation as a Java package. Blending remote evaluation and active messages, our model permits programmers to express asynchronous, complex interactions, so overcoming some of the limitations of the models based on message passing and RPC and reducing communication costs.
The Skel-BSP methodology provides an adaptive support for skeleton programs aiming to achieve performance portability for parallel programming. The adaptivity is obtained by choosing templates and implementation param...
详细信息
ISBN:
(纸本)3540675531
The Skel-BSP methodology provides an adaptive support for skeleton programs aiming to achieve performance portability for parallel programming. The adaptivity is obtained by choosing templates and implementation parameters according to the target machine characteristics. Each choice is made using an optimization theorem demonstrated using EdD-BSP cost predictions. The work presents several strategies to optimize the pure data parallel subset of Skel-BSP. Data parallel programs can be written in Skel-BSP combining map, reduce and scan skeletons. Analogous results have been already derived for the stream parallel subset of the language.
We describe a parallel solution to the problem of reliably plotting a plane algebraic curve. The sequential program is implemented in the software library CASA on top of the computer algebra system Maple. The parallel...
详细信息
ISBN:
(纸本)3540679561
We describe a parallel solution to the problem of reliably plotting a plane algebraic curve. The sequential program is implemented in the software library CASA on top of the computer algebra system Maple. The parallel version is based on Distributed Maple, a parallel programming extension written in Java. We evaluate its performance on a cluster of workstations and PCs, on a massively parallel multiprocessor, and on a cluster that couples workstations and multiprocessor.
A theory of programming starts with a complete Boolean Algebra of specifications, and defines healthiness conditions which exclude in feasibility of implementation These are expressed as algebraic laws useful for tran...
详细信息
ISBN:
(纸本)0769509150;0769509169
A theory of programming starts with a complete Boolean Algebra of specifications, and defines healthiness conditions which exclude in feasibility of implementation These are expressed as algebraic laws useful for transformation and optimisation of design. programming notations and languages must be restricted to those preserving all the healthiness conditions. We have explored a wide range of programming paradigms, including nondeterministic, sequential, parallel, logical and probabilistic. In all cares, we have found a single healthiness condition, formalised by constructions due to Karoubi and to Kleisli. The uniformity maintains for all paradigms a single notion of correctness throughout the chain that leads from specification through designs to programs that are proved to meet the original specification.
The Message Passing Interface 1.1 (MPI 1.1) standard defines a library of message-passing functions for parallel and distributed computing. We have developed a new software tool called C++2MPI which can automatically ...
详细信息
ISBN:
(纸本)076950759X
The Message Passing Interface 1.1 (MPI 1.1) standard defines a library of message-passing functions for parallel and distributed computing. We have developed a new software tool called C++2MPI which can automatically generate MPI derived datatypes for a specified C++ class. C++2MPI can generate data types for derived classes, for partially and fully-specialized templated classes, and for classes with private data members. Given one or more user-provided classes as input, C++2MPI generates, compiles and archives a function for creating the MPI derived datatype. When the generated function is executed, it builds the derived MPI datatype if the datatype does nor already exist, and returns the value of an MPI handle for referencing the datatype. PGMT (Processing Graph Method Tool) is a set of program interfaces for porting the Graph Method (PGM), a parallel programming method, to diverse networks of processors. C++2MPI was developed as a component of PGMT, bur can be used as a stand-alone tool.
Several generalizations of the flat data parallel model have been proposed. Their aim is to allow the capability of nested parallel invocations, combining the easiness of programming of the data parallel model with th...
详细信息
ISBN:
(纸本)3540410104
Several generalizations of the flat data parallel model have been proposed. Their aim is to allow the capability of nested parallel invocations, combining the easiness of programming of the data parallel model with the efficiency of the control parallel model. We examine the solutions provided to this issue by two standard parallel programming platforms, OpenMP and MPI. Both their expression capacity and their efficiency are compared on a Sun HPC 3500 and a SGI Origin 2000. The two considered architectures are shared memory and, consequently, more suitable for their exploitation under OpenMP. In spite of this, the results prove that, under the use of the methodology proposed for MPI in this paper, not only the performances of the two platforms are similar but, more remarkably, the effort invested in software development is also the same.
Network based distributed computing has been gaining popularity over the past decade. Many parallel programming languages and related parallel programming modes are becoming widely accepted. However, the execution of ...
详细信息
ISBN:
(纸本)0769507719
Network based distributed computing has been gaining popularity over the past decade. Many parallel programming languages and related parallel programming modes are becoming widely accepted. However, the execution of parallel applications on distributed systems has been hampered by the high communication overhead. To reduce the communication overhead and the completion time of a parallel application, we propose a Key Message model for parallel computing on Network of Workstations (NOWs). In the Key Message model, all messages generated in a key message path are prioritized. A Key Message path in a task graph is defined as the path that is optimized by the Key Message algorithm. All messages generated in a key message path are prioritized. Besides, the Key Message algorithm automatically finds the Key message paths. In this paper, we first describe the algorithm that identifies the Ley message to be prioritized in a parallel application, then analyze the cost of the algorithm, and finally evaluate the performance of the algorithm in a simulation. Our preliminary analysis of the algorithm shows improvement oner the system which does not use prioritization scheme.
暂无评论