Exploring weak scalability for FEM calculations on a GPU-enhanced cluster

被引:69
作者
Goeddeke, Dominik [1 ]
Strzodka, Robert [2 ]
Mohd-Yusof, Jamaludin [3 ]
McCormick, Patrick [3 ]
Buijssen, Sven H. M. [1 ]
Grajewski, Matthias [1 ]
Turek, Stefan [1 ]
机构
[1] Univ Dortmund, Inst Appl Math, D-44227 Dortmund, Germany
[2] Stanford Univ, Max Planck Ctr, Stanford, CA 94305 USA
[3] Los Alamos Natl Lab, Comp Conputat & Stat Sci Div, Los Alamos, NM 87545 USA
关键词
graphics processors; heterogeneous computing; parallel multigrid solvers; commodity based clusters; finite elements; COMPUTATION;
D O I
10.1016/j.parco.2007.09.002
中图分类号
TP301 [理论、方法];
学科分类号
080201 [机械制造及其自动化];
摘要
The first part of this paper surveys co-processor approaches for commodity based clusters in general, not only with respect to raw performance, but also in view of their system integration and power consumption. We then extend previous work on a small GPU cluster by exploring the heterogeneous hardware approach for a large-scale system with up to 160 nodes. Starting with a conventional commodity based cluster we leverage the high bandwidth of graphics processing units (GPUs) to increase the overall system bandwidth that is the decisive performance factor in this scenario. Thus, even the addition of low-end, out of date GPUs leads to improvements in both performance- and power-related metrics. (c) 2007 Elsevier B.V. All rights reserved.
引用
收藏
页码:685 / 699
页数:15
相关论文
共 35 条
[1]
*AMD INC, 2006, TORR TECHN
[2]
[Anonymous], FGCS FUTURE GENER CO
[3]
[Anonymous], **NON-TRADITIONAL**
[4]
[Anonymous], CS07594 U TENN COMP
[5]
[Anonymous], CRAY XD1 SUP
[6]
BECKER C, 2007, THESIS U DORTMUND FA
[7]
Brook for GPUs: Stream computing on graphics hardware [J].
Buck, I ;
Foley, T ;
Horn, D ;
Sugerman, J ;
Fatahalian, K ;
Houston, M ;
Hanrahan, P .
ACM TRANSACTIONS ON GRAPHICS, 2004, 23 (03) :777-786
[8]
BUTTARI A, 2007, UTCS07595 U TENN KNO
[9]
*CLEARSP TECHN INC, 2006, CLEARSP ADV ACC BOAR
[10]
DAGRACA G, 2006, 7 C REAL NUMB COMP R