This Article Full Text (PDF) References Alert me when this article is cited Alert me if a correction is posted Services Email this article to a friend Similar articles in this journal Similar articles in Web of Science Alert me to new issues of the journal Add to Saved Citations Download to citation manager Request Permissions Request Reprints Add to My Marked Citations Citing Articles Citing Articles via Google Scholar Citing Articles via Scopus Google Scholar Articles by de Supinski, B. R. Articles by Çatalyürek, U. Search for Related Content Social Bookmarking                         What's this?

BlueGene/L applications: Parallelism On a Massive Scale

LAWRENCE LIVERMORE NATIONAL LABORATORY, LIVERMORE, CA 94551, USA, BRONIS{at}LLNL.GOV

Martin Schulz

LAWRENCE LIVERMORE NATIONAL LABORATORY, LIVERMORE, CA 94551, USA

Vasily V. Bulatov

LAWRENCE LIVERMORE NATIONAL LABORATORY, LIVERMORE, CA 94551, USA

William Cabot

LAWRENCE LIVERMORE NATIONAL LABORATORY, LIVERMORE, CA 94551, USA

Bor Chan

LAWRENCE LIVERMORE NATIONAL LABORATORY, LIVERMORE, CA 94551, USA

Andrew W. Cook

LAWRENCE LIVERMORE NATIONAL LABORATORY, LIVERMORE, CA 94551, USA

Erik W. Draeger

LAWRENCE LIVERMORE NATIONAL LABORATORY, LIVERMORE, CA 94551, USA

James N. Glosli

LAWRENCE LIVERMORE NATIONAL LABORATORY, LIVERMORE, CA 94551, USA

Jeffrey A. Greenough

LAWRENCE LIVERMORE NATIONAL LABORATORY, LIVERMORE, CA 94551, USA

Keith Henderson

LAWRENCE LIVERMORE NATIONAL LABORATORY, LIVERMORE, CA 94551, USA

Alison Kubota

LAWRENCE LIVERMORE NATIONAL LABORATORY, LIVERMORE, CA 94551, USA

Steve Louis

LAWRENCE LIVERMORE NATIONAL LABORATORY, LIVERMORE, CA 94551, USA

Brian J. Miller

LAWRENCE LIVERMORE NATIONAL LABORATORY, LIVERMORE, CA 94551, USA

Mehul V. Patel

LAWRENCE LIVERMORE NATIONAL LABORATORY, LIVERMORE, CA 94551, USA

Thomas E. Spelce

LAWRENCE LIVERMORE NATIONAL LABORATORY, LIVERMORE, CA 94551, USA

Frederick H. Streitz

LAWRENCE LIVERMORE NATIONAL LABORATORY, LIVERMORE, CA 94551, USA

Peter L. Williams

LAWRENCE LIVERMORE NATIONAL LABORATORY, LIVERMORE, CA 94551, USA

Robert K. Yates

LAWRENCE LIVERMORE NATIONAL LABORATORY, LIVERMORE, CA 94551, USA

Andy Yoo

LAWRENCE LIVERMORE NATIONAL LABORATORY, LIVERMORE, CA 94551, USA

George Almasi

IBM THOMAS J. WATSON RESEARCH CENTER

Gyan Bhanot

IBM THOMAS J. WATSON RESEARCH CENTER

Alan Gara

IBM THOMAS J. WATSON RESEARCH CENTER

John A. Gunnels

IBM THOMAS J. WATSON RESEARCH CENTER

Manish Gupta

IBM THOMAS J. WATSON RESEARCH CENTER

Jose Moreira

IBM THOMAS J. WATSON RESEARCH CENTER

James Sexton

IBM THOMAS J. WATSON RESEARCH CENTER

Bob Walkup

IBM THOMAS J. WATSON RESEARCH CENTER

Charles Archer

IBM SYSTEMS AND TECHNOLOGY GROUP

Francois Gygi

UNIVERSITY OF CALIFORNIA, DAVIS

Timothy C. Germann

LOS ALAMOS NATIONAL LABORATORY

Kai Kadau

LOS ALAMOS NATIONAL LABORATORY

Peter S. Lomdahl

LOS ALAMOS NATIONAL LABORATORY

Charles Rendleman

LAWRENCE BERKELEY NATIONAL LABORATORY

Michael L. Welcome

LAWRENCE BERKELEY NATIONAL LABORATORY

William McLendon

SANDIA NATIONAL LABORATORIES

Bruce Hendrickson

SANDIA NATIONAL LABORATORIES

Franz Franchetti

CARNEGIE MELLON UNIVERSITY

Stefan Kral

VIENNA UNIVERSITY OF TECHNOLOGY

Jürgen Lorenz

VIENNA UNIVERSITY OF TECHNOLOGY

Christoph W. Überhuber

VIENNA UNIVERSITY OF TECHNOLOGY

Edmond Chow

D. E. SHAW RESEARCH AND DEVELOPMENT

Ümit Çatalyürek

OHIO STATE UNIVERSITY

BlueGene/L (BG/L), developed through a partnership between IBM and Lawrence Livermore National Laboratory (LLNL), is currently the world's largest system both in terms of scale, with 131,072 processors, and absolute performance, with a peak rate of 367 Tflop/s. BG/L has led the last four Top500 lists with a Linpack rate of 280.6 Tflop/s for the full machine installed at LLNL and is expected to remain the fastest computer in the next few editions. However, the real value of a machine such as BG/L derives from the scientific breakthroughs that real applications can produce by successfully using its unprecedented scale and computational power. In this paper, we describe our experiences with eight large scale applications on BG/ L from several application domains, ranging from molecular dynamics to dislocation dynamics and turbulence simulations to searches in semantic graphs. We also discuss the challenges we faced when scaling these codes and present several successful optimization techniques. All applications show excellent scaling behavior, even at very large processor counts, with one code even achieving a sustained performance of more than 100 Tflop/s, clearly demonstrating the real success of the BG/L design.

Key Words: massively parallel architectures • BlueGene/L • application scalability • performance study and optimization

International Journal of High Performance Computing Applications, Vol. 22, No. 1, 33-51 (2008)
DOI: 10.1177/1094342007085025


CiteULike    Complore    Connotea    Del.icio.us    Digg    Reddit    Technorati    Twitter    What's this?