Characterization of Scientific Workloads on Systems with Multi-Core Processors

Author

Alam, Sadaf R. ; Barrett, Richard F. ; Kuehn, Jeffery A. ; Roth, Philip C. ; Vetter, Jeffrey S.

Author_Institution

Comput. Sci. & Math. Div., Oak Ridge Nat. Lab., TN

fYear

2006

fDate

25-27 Oct. 2006

Firstpage

225

Lastpage

236

Abstract

Multi-core processors are planned for virtually all next-generation HPC systems. In a preliminary evaluation of AMD Opteron Dual-Core processor systems, we investigated the scaling behavior of a set of micro-benchmarks, kernels, and applications. In addition, we evaluated a number of processor affinity techniques for managing memory placement on these multi-core systems. We discovered that an appropriate selection of MPI task and memory placement schemes can result in over 25% performance improvement for key scientific calculations. We collected detailed performance data for several large-scale scientific applications. Analyses of the application performance results confirmed our micro-benchmark and scaling results

Keywords

message passing; multiprocessing systems; storage management; AMD Opteron Dual-Core processor systems; MPI task; all next-generation HPC systems; memory placement management; multicore processors; multicore systems; processor affinity; scientific workloads; Application software; Bandwidth; Concurrent computing; Government; Kernel; Memory management; Multicore processing; Parallel processing; Scientific computing; Sockets; AMD Opteron; Multi-core processor; Performance characterization; micro-benchmarking; scientific applications;

fLanguage

English

Publisher

ieee

Conference_Titel

Workload Characterization, 2006 IEEE International Symposium on

Conference_Location

San Jose, CA

Print_ISBN

1-4244-0508-4

Electronic_ISBN

1-4244-0509-2

Type

conf

DOI

10.1109/IISWC.2006.302747

Filename

4086151