COTS Clusters vs. the Earth Simulator: An Application Study Using IMPACT-3D
April 2005
Page 88.2
Abstract
In 2002, Japan announced the Earth Simulator-a supercomputer based on low-volume vector processors and a custom network-and reported that computational scientists had used it to achieve 14.9 TFLOPS with the IMPACT-3D code, which is written in High Performance Fortran (HPF). Of particular interest was that they had achieved this level of performance using a high-level parallel programming model. There has been considerable concern in the U.S. about the appropriateness of its hardware and software investments in supercomputing technology. To help assess the U.S. strategy of building systems from commodity-off-the-shelf (COTS) components, we explored using a combination of HPF and scalar compiler technology to tailor IMPACT-3D to microprocessor-based supercomputers and evaluated its performance and scalability on the AlphaServer-based Lemieux cluster at the Pittsburgh Supercomputer Center (PSC). On the Earth Simulator, IMPACT-3D achieved 45% of peak performance on 4096 processors; on 1024 processors of PSC's Lemieux, we achieved 17.29% of peak performance.
References
[1]
V. Adve, G. Jin, J. Mellor-Crummey, and Q. Yi. High Performance Fortran compilation techniques for parallelizing scientific codes. In Proceedings of the 1998 ACM/IEEE Conference on Supercomputing, Orlando, FL, Nov. 1998. IEEE Computer Society Press.
[2]
V. Adve and J. Mellor-Crummey. Using integer sets for data-parallel program analysis and optimization. In Proceedings of the SIGPLAN '98 Conference on Programming Language Design and Implementation, Montreal, Canada, June 1998.
[3]
ANSI. Myrinet-on-VME Protocol Specification (ANSI/VITA 26-1998). American National Standard Institute, 1998.
[4]
D. Chavarría-Miranda. Advanced Data-Parallel Compilation. PhD thesis, Dept. of Computer Science, Rice University, Dec. 2003.
[5]
D. Chavarría-Miranda and J. Mellor-Crummey. An evaluation of data-parallel compiler support for line-sweep applications. The Journal of Instruction-Level Parallelism, 5, Feb. 2003. http://www.jilp.org/vol5. Special issue with selected papers from: The Eleventh International Conference on Parallel Architectures and Compilation Techniques, Sept. 2002. Guest Editors: Erik Altman and Sally McKee.
[6]
DOE Office of Science. Reasserting U.S. leadership in scientific computation. Office of Science Occasional Paper, http://www.ultrasim.info/doe_docs/ Occasional-Science-Computation.pdf. Last accessed Oct. 8, 2004.
[7]
A. Harten. High resolution schemes for hyperbolic conservation laws. Journal of Computational Physics, 49:357-393, 1983.
[8]
Intel Developer Forum. InfiniBand Architecture, June 2001.
[9]
G. Jin and J. Mellor-Crummey. On reducing storage requirement of scientific applications. In Proceedings of the Los Alamos Computer Science Institute Second Annual Symposium, Santa Fe, NM, Oct. 2003.
[10]
D. J. Kerbyson, A. Hoisie, and H. J. Wasserman. A comparison between the Earth Simulator and Alphaserver systems using predictive application performance models. In Proceedings of the 2003 International Parallel and Distributed Processing Symposium, Nice, France, Apr. 2003.
[11]
C. Lazou. The Japanese Earth Simulator: A challenge and an opportunity. Primeur Monthly, June 2002. http://www.hoise.com/primeur/02/ articles/monthly/CL-PR-06-02-1.html.
[12]
J. Mellor-Crummey, V. Adve, B. Broom, D. Chavarría-Miranda, R. Fowler, G. Jin, K. Kennedy, and Q. Yi. Advanced optimization strategies in the Rice dHPF compiler. Concurrency and Computation: Practice and Experience, 14(8-9):741-767, 2002.
[13]
L. Oliker, A. Canning, J. Carter, and J. Shalf. Scientific computations on modern parallel vector systems. In Proceedings of the 2004 ACM/IEEE Conference on Supercomputing, Pittsburgh, PA, Nov. 2004. IEEE Computer Society Press.
[14]
PARAllel Total Energy Code. http://http://www. nersc.gov/projects/paratec.
[15]
F. Petrini, W. Feng, A. Hoisie, S. Coll, and E. Frachtenberg. The Quadrics network: high performance clustering technology. IEEE Micro, 22(1):46-57, Jan.-Feb. 2002.
[16]
A. Qasem, G. Jin, and J. Mellor-Crummey. Improving performance with integrated program transformations. Technical Report TR03-419, Dept of Computer Science, Rice University, Houston, TX, Oct. 2003.
[17]
H. Sakagami, H. Murai, Y. Seo, and M. Yokokawa. 14.9 TFLOPS three-dimensional fluid simulation for fusion science with HPF on the Earth Simulator. In Proceedings of the 2002 ACM/IEEE Conference on Supercomputing, pages 1-14, Baltimore, Maryland, Nov. 2002. IEEE Computer Society Press.
[18]
T. Sato, S. Kitawaki, and M. Yokokawa. Earth Simulator running. http://www.ultrasim.info/sato.pdf, 2002.
[19]
Y. Seo, H. Iwashita, H. Ohta, and S. Takahashi. HPF/JA: HPF extensions for real-world parallel applications. In Proceedings of the 2th Annual HPF User Group meeting, Porto, Portugal, June 1998.
[20]
S. Shingu, H. Takahara, H. Fuchigami,M. Yamada, Y. Tsuda, W. Ohfuchi, Y. Sasaki, K. Kobayashi, T. Hagiwara, S. ichi Habata, M. Yokokawa, H. Itoh, and K. Otsuka. A 26.58 TFLOPS global atmospheric simulation with the spectral transform method on the Earth Simulator. In Proceedings of the 2002 ACM/IEEE Conference on Supercomputing, Baltimore, Maryland, Nov. 2002. IEEE Computer Society Press.
[21]
M. Yokokawa, K. Itakura, A. Uno, T. Ishihara, and Y. Kaneda. 16.4 TFLOPS direct numerical simulation of turbulence by a Fourier spectral method on the Earth Simulator. In Proceedings of the 2002 ACM/IEEE Conference on Supercomputing, Baltimore, Maryland, Nov. 2002. IEEE Computer Society Press.
Index Terms
- COTS Clusters vs. the Earth Simulator: An Application Study Using IMPACT-3D
Recommendations
Three-level hybrid vs. flat MPI on the Earth Simulator: Parallel iterative solvers for finite-element method
6th IMACS International symposium on iterative methods in scientific computingAn efficient parallel iterative method for finite element method has been developed for symmetric multiprocessor (SMP) cluster architectures with vector processors such as the Earth Simulator. The method is based on a three-level hybrid parallel ...
Software system of the earth simulator
The Earth Simulator (ES) is a large scale, distributed memory, parallel computer system consisting of 640 processor nodes (PN) with shared memory vector multiprocessors (64GFLOPS/PN, 5120 APs in total, AP: arithmetic processor). All the nodes are ...
Optimizing parallel performance of unstructured volume rendering for the Earth Simulator
Parallel graphics and visualisationA scalable and high-performance parallel visualization subsystem has been developed in GeoFEM for the Earth Simulator (ES). As part of the ES project in Japan, the proposed sub-system is effective for the visualization of huge-scale unstructured ...
Comments
Information & Contributors
Information
Published In
April 2005
ISBN:0769523129
Publisher
IEEE Computer Society
United States
Publication History
Published: 04 April 2005
Qualifiers
- Article
Contributors
Other Metrics
Bibliometrics & Citations
Bibliometrics
Article Metrics
- 0Total Citations
- 0Total Downloads
- Downloads (Last 12 months)0
- Downloads (Last 6 weeks)0