MPI is the new standard for multicomputer and cluster message passing introduced by the Message-Passing Interface Forum (MPIF) in April 1994. This paper describes the current inter-communicator interface found in MPI and the reasons for its current design. We also motivate the need for additional inter-communicator operations and introduce the extensions we have included in MPIX (MPI eXtension Library), a library of extensions to MPI that we are currently developing. Inter-communicators may be used for a variety of purposes such as in client/server applications (i.e., I/O and graphics servers) and for process management in dynamic process environments and multi-protocol implementations; MPI's definitions are unnecessarily restrictive, so we extend them here. We discuss the inter-communicator collective operations defined in MPIX and illustrate their use. We also discuss additional inter-communicator construction routines not in the original MPI interface, but that are provided in MPIX....

Simulating transients in semiconductor devices involves numerically solving the time-dependent drift-diffusion equations, usually in two or three space dimensions. Because of the computation cost of these simulations, methods that perform careful domain decomposition so as to exploit parallel processing have received much recent attention. In this paper, we describe using accelerated waveform relaxation (WR) to perform parallel device transient simulation using both clusters of workstations and the IBM SP-2. The accelerated WR algorithms are compared to pointwise direct and iterative methods, and it is shown that the accelerated WR method is competitive on a single processor. In addition, it is shown that with a domain decomposition chosen for rapid iterative method convergence rather than parallel efficiency, the pointwise methods parallelize poorly but the WR mcthod achieves near linear speedup (with respect to the number of processors) on the IBM SP-2.

The draft of the MPI (Message-Passing Interface) standard was released at Supercomputing '93, November 1993. The final MPI document is expected to be released in mid-April of 1994. Language bindings for C and FORTRAN were included in the draft; however, a language binding for C++ was not included. MPI provides support for datatypes and topologies that is needed for developing numerical libraries, however the interfaces described in the MPI draft document for these two features have several disadvantages. In this paper we describe and offer examples of a C++ interface to MPI that can be used to build scalable, object-oriented numerical libraries in a language that directly supports objectoriented programming. We also introduce several ideas from the Zipcode message-passing library related to datatypes and topologies, then compare and contrast these to the MPI approach. The best ideas from both approaches are combined to make a better interface to datatypes and topologies. Work support...

This paper presents a new kind of portability system, Unify, which modifies the PVM message passing system to provide (currrently a subset of) the Message Passing Interface (MPI) standard notation for message passing. Unify is designed to reduce the effort of learning MPI while providing a sensible means to make use of MPI libraries and MPI calls while applications continue to run in the PVM environment. We are convinced that this strategy will reduce the costs of porting completely to MPI, while providing a gradual environment within which to evolve. Furthermore, it will permit immediate use of MPI-based parallel libraries in applications, even those that use PVM for user code. We describe several paradigms for supporting MPI and PVM message passing notations in a single environment, and note related work on MPI and PVM implementations. We show the design options that existed within our chosen paradigm (which is an MPI interface added to the base PVM system), and why we chose that par...

. In this paper, we present a portable, object-oriented, pure Java implementation of the Message-Passing Interface (MPI), called jmpi. jmpi is a class library of Java-routines for specifying and coordinating parallel codes. Our pure Java implementation is distinguished from earlier implementation efforts that pervasively use native methods and provide a Java wrapper functionality to some specific traditional MPI implementations. While bringing in a consistent MPI object model suitable for Java, we also follow the standard MPI Application Programming Interface (API) definitions as closely as possible to keep the learning curve short for experienced MPI programmers. We tested the performance of jmpi, by running a set of benchmark programs written in Java with calls to jmpi library routines on a cluster of SUN UltraSparc workstations. 1#Introduction The use of a collection of general-purpose heterogeneous computer systems interconnected by existing networks and support services as a sing...

We introduce the all-software, standard C++-based Aurora distributed shared data system. As with related systems, it provides a shared data abstraction on distributed memory hardware. An innovation in Aurora is the use of scoped behaviour for per-context data sharing optimizations (i.e., portion of source code, such as a loop or phase). With scoped behaviour, a new language scope (e.g., nested braces) can be used to optimize the data sharing behaviour of the selected source code. Different scopes and different shared data can be optimized in different ways. Thus, scoped behaviour provides a novel level of flexibility to incrementally tune the parallel performance of an application. 1. Introduction Parallel programming systems based on shared memory and shared data models are becoming more popular and widespread. Accessing local and remote data using the same programming interface (i.e., reads and writes) is often more convenient than mixing local accesses with message passing. Conseq...

We describe a number of early efforts to make use of the Message Passing Interface (MPI) standard in applications, based on an informal survey conducted in May-June, 1994. Rather than a definitive statement of all MPI development work, this paper addresses initial successes, progress, and impressions that application developers have with MPI, according to the responses received. We summarize the important aspects of each survey response, and draw conclusions about the spread of MPI into applications. An understanding of message-passing, and access to the MPI standard are prerequisites for appreciating this paper. Some background material is provided to ease this requirement. Skjellum, et al. Early MPI: : : 3 1 Introduction In this paper, we describe a number of early efforts to make use of the Message Passing Interface (MPI) standard in real applications (Forum 1994a; Forum 1994b). An informal survey of efforts is reported here, together with our commentary. We summarize the respon...

This report addresses the issues arising from the use of parallel machines and considers the various techniques used by members of the consortium in this context. Before considering the algorithms in detail, we describe, in section 2, the main types of parallel architecture and survey various attempts at providing a taxonomy. Then, in section 3, we address the difficult issue of the measurement of processor performance in order to quantify any enhancement obtained by implementing an algorithm in parallel. Section 4 presents the main features of in general, and PVM ( ) in particular. The latter is an application that is used to generate distributed versions of sequential algorithms for use on networks of workstations. The parallel implementations of the GA toolkit, , and the associated simulated annealing toolkit, , both developed at UEA, have been produced using PVM. Exact algorithms will always find the optimal solution to a problem given enough time and space. Subject to these constraints, they must always be the preferred method of solution. In practice, the time and space constraints can prevent the use of an exact algorithm and thus the potential of parallelism to reduce these factors becomes an important factor. Total enumeration is embarassingly parallel. With processors it is reasonable to expect an-fold reduction in time to undertake such a thorough search. Such a saving is seldom sufficient to make the method viable so we will concentrate on other exact methods here. In section 5, we review parallel branchand -bound, reprinting a survey paper written by the UEA partners in the consortium and previously published in [1]. Because of the interest in interior point methods for the CALMA project and its widely cited potential for parallelisation, this provides the ...

Many image processing tasks exhibit a high degree of data locality and parallelism and map quite readily to specialized massively parallel computing hardware. However, as workstation clusters are becoming a viable and economical parallel computing resource, it is important to understand how to use these environments for parallel image processing as well. In this paper we discuss our implementation of a parallel image processing software library (the Parallel Image Processing Toolkit). The library is easily extensible and hides most parallelism from the user. Inside the Toolkit, a message-passing model of parallelism is designed around the Message Passing Interface (MPI) standard. Experimental results are presented to demonstrate the parallel speedup obtained with the Parallel Image Processing Toolkit in a typical workstation cluster with some common image processing tasks. We also discuss load balancing and the potential for parallelizing portions of image processing tasks that seem to...

process naming to allow libraries to describe their communication in terms suitable to their own data structures and algorithms, ffl The ability to "adorn" a set of communicating processes with additional user-defined attributes, such as extra collective operations. This mechanism should provide a means for the user or library writer effectively to extend a message-passing notation. In addition, a unified mechanism or object is needed for conveniently denoting communication context, the group of communicating processes, to house abstract process naming, and to store adornments. 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 134 CHAPTER 5. GROUPS, CONTEXTS, AND COMMUNICATORS 5.1.2 MPI's Support for Libraries The corresponding concepts that MPI provides, specifically to support robust libraries, are as follows: ffl Contexts of communication, ffl Groups of processes, ffl Virtual topolo...