HPC in Java: Experiences in Implementing the NAS Parallel Benchmarks
نویسندگان
چکیده
This paper reports on the design, implementation and benchmarking of a Java version of the Nas Parallel Benchmarks. We first briefly describe the implementation and the performance pitfalls. We then compare the overall performance of the Fortran MPI (PGI) version with a Java implementation using the ProActive middleware for distribution. All Java experiments were conducted on virtual machines with different vendors and versions. We show that the performance varies with the type of computation but also with the Java Virtual Machine, no single one providing the best performance in all experiments. We also show that the performance of the Java version is close to the Fortran one on computational intensive benchmarks. However, on some communications intensive benchmarks, the Java version exhibits scalability issues, even when using a high performance socket implementation (JFS). Key–Words: Benchmarks, Java, HPC, NAS Benchmarks, ProActive, SCI
منابع مشابه
Performance and Scalability of the NAS Parallel Benchmarks in Java
Several features make Java an attractive choice for scientific applications. In order to gauge the applicability of Java to Computational Fluid Dynamics (CFD), we have implemented the NAS Parallel Benchmarks in Java. The performance and sealability of the benchmarks point out the areas where improvement in Java compiler technology and in Java thread implementation would position Java closer to ...
متن کاملParallel Languages and Compilers: Perspective From the Titanium Experience
We describe the rationale behind the design of key features of Titanium—an explicitly parallel dialect of Java for high-performance scientific programming—and our experiences in building applications with the language. Specifically, we address Titanium’s Partitioned Global Address Space model, SPMD parallelism support, multi-dimensional arrays and arrayindex calculus, memory management, immutab...
متن کاملPreliminary Evaluation of Chapel capabilities with NAS Parallel Benchmarks
Chapel is a new high-level programming language aimed at the programmability aspect required for HPC. This document shows a preliminary comparison between the OMP implementation of several NAS Benchmarks and their corresponding versions written in Chapel.
متن کاملNAS Experience with the Cray X1
A Cray X1 computer system was installed at the NASA Advanced Supercomputing (NAS) facility at NASA Ames Research Center in 2004. An evaluation study of this unique high performance computing (HPC) architecture, from the standpoints of processor and system performance, ease of use, and production computing readiness tailored to the needs of the NAS scientific community, was recently completed. T...
متن کاملBenchmarking Java on the IBM SP-2
In this paper we present results and analysis of benchmark tests that directly compare the performance of JavaMPI, Fortran and C routines on the 128-node IBM SP-2 HPC system, at the San Diego Supercomputer Center. The suite of tests include scalability studies of the NAS Embarrassingly Parallel and Integer Sort Benchmarks for 2 to 128 processors, along with results from basic operations and com...
متن کامل