ISSN:1040-3108    

DOI:10.1002/cpe.4330010102

出版商:John Wiley&Sons, Ltd    

年代:1989

数据来源: WILEY

摘要:

AbstractIn this paper the performance of the Intel iPSC/2 hypercube multiprocessor is analyzed. Computation and communication performance for a number of benchmarks are presented. We derive some fundamental performance parameters of the machine. Further, we investigate the difference between several communication schemes. Using the results of our measurements we can highlight some features and peculiarities in the iPSC/2 hardware and software. Where possible we make a comparison with the iPSC/1 and Ncube hypercubes.

ISSN:1040-3108    

DOI:10.1002/cpe.4330010103

出版商:John Wiley&Sons, Ltd    

年代:1989

数据来源: WILEY

摘要:

AbstractThe development of computing systems with large numbers of processors has been motivated primarily by the need to solve large, complex problems more quickly than is possible with uniprocessor systems. Traditionally, multiprocessor systems have been uniprogrammed, i.e., dedicated to the execution of a single set of related processes, since this approach provides the fastest response for an individual program once it begins execution. However, if the goal of a multiprocessor system is to minimize average response time or to maximize throughput, then multiprogramming must be considered.In this paper, a model of a simple multiprocessor system with a two‐program workload is reviewed; the model is then applied to an Intel iPSC/2 hypercube multiprocessor with a workload consisting of parallel wavefront algorithms for solving triangular systems of linear equations. Throughputs predicted by the model are compared with throughputs obtained experimentally from an actual system. The results provide validation for the model and indicate that significant performance improvements for multiprocessor systems are possible through multiprogrammin

ISSN:1040-3108    

DOI:10.1002/cpe.4330010104

出版商:John Wiley&Sons, Ltd    

年代:1989

数据来源: WILEY

摘要:

AbstractThis paper analyzes the performance of a discrete‐event combat simulation executed on a parallel processor under control of the Time Warp Operating System. Time Warp is in a class of distributed simulation methods calledOptimistic methodswhich have proven to be useful over a wide range of simulations. The combat simulation used for this performance study, called STB88, is a division‐corps model incorporating a number of different types of computations. The speed‐up for three versions of this model on the Caltech/JPL Mark III Hypercube and the BBN Butterfly parallel processors was measured relative to an efficient sequential execution of the same model on the same hardware. The results indicate that STB88 version 1 achieves a speed‐up of 28.6 on 60 Mark III processors, while STB88 version 2 achieves a speed‐up of 36.8 on 100 Butterfly processors. Version 3 of STB88 achieved a speed‐up of 38.5 on 128 Mark III processors. The versions differed only in their interface to Time Warp. On the Butterfly, the sequential execution completed in 2 hours, while the 100 processor execution completed in

ISSN:1040-3108    

DOI:10.1002/cpe.4330010105

出版商:John Wiley&Sons, Ltd    

年代:1989

数据来源: WILEY

摘要:

AbstractAn algorithm has been implemented for time‐accurate solutions of the two‐dimensional compressible Euler equations using an irregular triangular mesh. The code runs on distributed or shared memory or sequential machines, and is written using the Distributed Irregular Mesh Environment (DIME). DIME is a programming environment for calculations with such meshes, with adaptive mesh refinement and dynamic load balancing. Results are presented for an example of a Mach 3 flow over a step, computed with a 32‐processor NCUBE hype

ISSN:1040-3108    

DOI:10.1002/cpe.4330010106

出版商:John Wiley&Sons, Ltd    

年代:1989

数据来源: WILEY

摘要:

AbstractParallel supercomputers are now in regular use at Caltech for several major scientific calculations. We use this experience to abstract a set of lessons for applications, decomposition, performance, hardware and software. We consider hypercubes, transputer arrays and the SIMD Connection Machine CM‐2 and AMT DAP. These are contrasted, where possible, with CRAY and other high performance conventional computers. Applications covered are lattice gauge theory, plasma physics, statistical and condensed matter physics, astronomical data analysis, quantum chemistry, graphics ray tracing, string dynamics, grain dynamics, astrophysical particle dynamics, computer chess and Kalman filter

ISSN:1040-3108    

DOI:10.1002/cpe.4330010107

出版商:John Wiley&Sons, Ltd    

年代:1989

数据来源: WILEY

摘要:

AbstractWe describe a novel methodology for translating some PARLOG programs to occam2. Our methodology directs the translation based on a process level description generated from the PARLOG program. The translation of several example PARLOG programs is included. These examples feature: dynamic creation of processes; the one‐to‐many communication capabilities of PARLOG; and use of unbounded, asynchronous communication. The examples are then used to consider the translation of a previously developed programming application from the field of Artificial Intelligence, namely a PARLOGchart parser.We see two main interests in such a methodology: firstly, as a means of obtaining high‐performance execution of PARLOG programs–PARLOG programs translated to occam2 can be executed on the high‐performance Transputer device; secondly, PARLOG may have a use as a declarative specification language for the development of occam2

ISSN:1040-3108    

DOI:10.1002/cpe.4330010108

出版商:John Wiley&Sons, Ltd    

年代:1989

数据来源: WILEY

ISSN:1040-3108    

DOI:10.1002/cpe.4330010101

出版商:John Wiley&Sons, Ltd    

年代:1989

数据来源: WILEY