Journal Article

PreJuSER-393

http://join2-wiki.gsi.de/foswiki/pub/Main/Artwork/join2_logo100x88.png Performance measurement and analysis of large-scale parallel applications on leadership computing systems

Wylie, B. J. N.FZJ* ; Geimer, M.FZJ* ; Wolf, F.FZJ*

2008
John Wiley & Sons New York, NY

This record in other databases:  

Please use a persistent id in citations: doi:10.3233/SPR-2008-0255

Abstract: Developers of applications with large-scale computing requirements are currently presented with a variety of high-performance systems optimised for message-passing, however, effectively exploiting the available computing resources remains a major challenge. In addition to fundamental application scalability characteristics, application and system peculiarities often only manifest at extreme scales, requiring highly scalable performance measurement and analysis tools that are convenient to incorporate in application development and tuning activities. We present our experiences with a multigrid solver benchmark and state-of-the-art real-world applications for numerical weather prediction and computational fluid dynamics, on three quite different multi-thousand-processor supercomputer systems - Cray XT3/4, MareNostrum & Blue Gene/L - using the newly-developed SCALASCA toolset to quantify and isolate a range of significant performance issues.

Keyword(s): J ; Large-scale parallel applications and systems (auto) ; performance measurement and analysis (auto)

Note: This work was partially funded under the German Helmholtz Association Young Investigators Program under Contract No. VNG-118. Collaboration with the developers of the XNS application from RWTH Aachen University was undertaken using resources of the John von Neumann Institute for Computing as part of the first Blue Gene/L Scaling Workshop in Julich.

---

Contributing Institute(s):

1. Jülich Supercomputing Centre (JSC)
2. Jülich-Aachen Research Alliance - Simulation Sciences (JARA-SIM)

Research Program(s):

1. Scientific Computing (P41)
2. ATMLPP - ATML Parallel Performance (ATMLPP) (ATMLPP)

Appears in the scientific report 2008

---