Abstract
Towards exascale supercomputing, both academia and industry have started to investigate the future HPC technologies. One of the most difficult challenges is the enhancement of energy efficiency of the computer system. We discuss the energy efficiency of existing architecture in this paper. With the analysis of the performance of the dense matrix–matrix multiplication (DGEMM), we propose the DGEMM-specialized Vector-SIMD architecture that only requires the small number of processor cores and low memory bandwidth. The DGEMM-specialized Vector-SIMD architecture can outperform existing architectures with respect to several metrics, as far as it is dedicated to limited usages, such as the DGEMM calculation. We conclude that this type of discussion will be essential in designing the future computer architecture.
This is a preview of subscription content, log in via an institution to check access.
Access this chapter
Tax calculation will be finalised at checkout
Purchases are for personal use only
Similar content being viewed by others
Notes
- 1.
Each fused multiply and add operation requires two memory accesses (16B).
References
Gotoblas library. http://www.tacc.utexas.edu/tacc-projects/gotoblas2/.
Top500 supercomputer sites. http://www.top500.org/.
Kazushige Goto and Robert A. van de Geijn. Anatomy of high-performance matrix multiplication. ACM Trans. Math. Softw., 34(3):12:1–12:25, May 2008.
P. Kogge, K. Bergman, S. Borkar, D. Campbell, W. Carson, W. Dally, M. Denneau, P. Franzon, W. Harrod, K. Hill, and Others. Exascale computing study: Technology challenges in achieving exascale systems. Technical report, University of Notre Dame, CSE Dept., 2008.
Peter M. Kogge and Timothy J. Dysart. Using the top500 to trace and project technology and architecture trends. In Proceedings of 2011 International Conference for High Performance Computing, Networking, Storage and Analysis, SC ’11, pages 28:1–28:11, New York, NY, USA, 2011. ACM.
Kuniaki Koike, Ken Fujino, Toshiyuki Fukushige, Hiroshi Daisaka, Yutaka Sugawara, Mary Inaba, Kei Hiraki, and Junnichiro Makino. Gravitational n-body simulation and lu decomposition with the multi purpose accelerator grape-dr (in japanese). Technical Report, 2009(26):1–11, 2009-07-28.
Yunsup Lee, Rimas Avizienis, Alex Bishara, Richard Xia, Derek Lockhart, Christopher Batten, and Krste Asanović. Exploring the tradeoffs between programmability and efficiency in data-parallel accelerators. In Proceedings of the 38th annual international symposium on Computer architecture, ISCA ’11, pages 129–140, New York, NY, USA, 2011. ACM.
Naoya Maruyama, Masaaki Kondo, Yasuo Ishii, Akihiro Nomura, Hiroyuki Takizawa, Takahiro Katagiri, Reiji Suda, and Yutaka Ishikawa. Technical Roadmap for Exascale Supercomputing. Technical report, SDHPC, 2008.
Takashi Soga, Akihiro Musa, Youichi Shimomura, Ryusuke Egawa, Ken’ichi Itakura, Hiroyuki Takizawa, Koki Okabe, and Hiroaki Kobayashi. Performance evaluation of nec sx-9 using real science and engineering applications. In Proceedings of the Conference on High Performance Computing Networking, Storage and Analysis, SC ’09, pages 28:1–28:12, New York, NY, USA, 2009. ACM.
Guangming Tan, Linchuan Li, Sean Triechle, Everett Phillips, Yungang Bao, and Ninghui Sun. Fast implementation of dgemm on fermi gpu. In Proceedings of 2011 International Conference for High Performance Computing, Networking, Storage and Analysis, SC ’11, pages 35:1–35:11, New York, NY, USA, 2011. ACM.
David W. Wall. Limits of instruction-level parallelism. In Proceedings of the fourth international conference on Architectural support for programming languages and operating systems, ASPLOS-IV, pages 176–188, New York, NY, USA, 1991. ACM.
Wm. A. Wulf and Sally A. McKee. Hitting the memory wall: implications of the obvious. SIGARCH Comput. Archit. News, 23(1):20–24, March 1995.
Author information
Authors and Affiliations
Corresponding author
Editor information
Editors and Affiliations
Rights and permissions
Copyright information
© 2013 Springer-Verlag Berlin Heidelberg
About this paper
Cite this paper
Ishii, Y. (2013). Architectural Considerations for Exascale Supercomputing. In: Resch, M., Wang, X., Bez, W., Focht, E., Kobayashi, H. (eds) Sustained Simulation Performance 2012. Springer, Berlin, Heidelberg. https://doi.org/10.1007/978-3-642-32454-3_2
Download citation
DOI: https://doi.org/10.1007/978-3-642-32454-3_2
Published:
Publisher Name: Springer, Berlin, Heidelberg
Print ISBN: 978-3-642-32453-6
Online ISBN: 978-3-642-32454-3
eBook Packages: Mathematics and StatisticsMathematics and Statistics (R0)