Design and Implementation of the HPCS Graph Analysis Benchmark on Symmetric MultiprocessorsReportar como inadecuado


Design and Implementation of the HPCS Graph Analysis Benchmark on Symmetric Multiprocessors


Design and Implementation of the HPCS Graph Analysis Benchmark on Symmetric Multiprocessors - Descarga este documento en PDF. Documentación en PDF para descargar gratis. Disponible también para leer online.

Graph theoretic problems are representative of fundamental computations in traditional and emerging scientific disciplines like scientific computing and computationalbiology, as well as applications in national security. We present our design and implementation of a graph theory application that supports the kernels from the Scalable Synthetic Compact Applications SSCA benchmark suite, developed under theDARPA High Productivity Computing Systems HPCS program. This syntheticbenchmark consists of four kernels that require irregular access to a large, directed,weighted multi-graph. We have developed a parallel implementation of this bench-mark in C using the POSIX thread library for commodity symmetric multiprocessorsSMPs. In this paper, we primarily discuss the data layout choices and algorithmicdesign issues for each kernel, and also present execution time and benchmark validationresults.



School of Computational Science and Engineering Technical Reports - College of Computing Technical Reports -



Autor: Bader, David A. - Madduri, Kamesh - -

Fuente: https://smartech.gatech.edu/







Documentos relacionados