Implications of Memory Performance for Highly Efficient Supercomputing of Scientific Applications

  • Akihiro Musa
  • Hiroyuki Takizawa
  • Koki Okabe
  • Takashi Soga
  • Hiroaki Kobayashi
Part of the Lecture Notes in Computer Science book series (LNCS, volume 4330)


This paper examines the memory performance of the vector-parallel and scalar-parallel computing platforms across five applications of three scientific areas; electromagnetic analysis, CFD/heat analysis, and seismology. Our evaluation results show that the vector platforms can achieve the high computational efficiency and hence significantly outperform the scalar platforms in the areas of these applications. We did exhaustive experiments and quantitatively evaluated representative scalar and vector platforms using real applications from the viewpoint of the system designers and developers. These results demonstrate that the ratio of memory bandwidth to floating-point operation rate needs to reach 4-bytes/flop to preserve the computational performance with hiding the memory access latencies by pipelined vector operations in the vector platforms. We also confirm that the enough number of memory banks to handle stride memory accesses leads to an increase in the execution efficiency. On the scalar platforms, the cache hit rate needs to be almost 100% to achieve the high computational efficiency.


Direct Numerical Simulation Memory Access Ground Penetrating Radar Memory Reference Memory Bandwidth 
These keywords were added by machine and not by the authors. This process is experimental and the keywords may be updated as the learning algorithm improves.


Unable to display preview. Download preview PDF.

Unable to display preview. Download preview PDF.


  1. 1.
    Shingu, S., et al.: A 26.58 Tflops Global Atmospheric Simulation with the Spectral Transform Method on the Earth Simulator. In: Proceedings of the ACM/IEEE SC 2002 conference (2002)Google Scholar
  2. 2.
    Yokokawa, M., et al.: 16.4-Tflops Direct Numerical Simulation of Turbulence by a Fourier Spectral Method on the Earth. In: Proceedings of the ACM/IEEE SC 2002 conference (2002)Google Scholar
  3. 3.
    Oliker, L., et al.: Evaluation of Cache-based Superscalar and Cacheless Vector Architectures for Scientific Computations. In: Proceedings of the ACM/IEEE SC 2003 conference (2003)Google Scholar
  4. 4.
    Oliker, L., et al.: Scientific Computations on Modern Parallel Vector System. In: Proceedings of the ACM/IEEE SC 2004 conference (2004)Google Scholar
  5. 5.
    Fatoohi, R.A.: Vector Performance Analysis of Three Supercomputers: Cray-2, Cray Y-MP, and ETA10-Q. In: Proceedings of Supercomputing 1989 (1989)Google Scholar
  6. 6.
    Fatoohi, R.A.: Vector Performance Analysis of The NEC SX-2. In: Proceedings of Supercomputing 1990 (1990)Google Scholar
  7. 7.
    Shan, H., et al.: Performance Characteristics of the Cray X1 and Their Implications for Application Performance Tuning. In: Proceedings of the ICS 2004 (2004)Google Scholar
  8. 8.
    Kitagawa, K., et al.: A Hardware Overview of SX-6 and SX-7 Supercomputer. NEC Research & Development 44, 2–7 (2003)Google Scholar
  9. 9.
    Senta, T., et al.: Itanium2 32-way Server System Architecture. NEC Research & Development 44, 8–12 (2003)Google Scholar
  10. 10.
    Kobayashi, T., et al.: FDTD simulation on array antenna SAR-GPR for land mine detection. In: Proceeding of SSR 2003: 1st International Symposium on Systems and Human Science, Osaka, Japan, November 2003, pp. 279–283 (2003)Google Scholar
  11. 11.
    Kunz, K.S., Luebbers, R.J.: The Finite Difference Time Domain Method for Electromagnetics. CRC Press, Boca Raton (1993)Google Scholar
  12. 12.
    Takagi, Y., et al.: Study of High Gain and Broadband Antipodal Fermi Anenna with Corrugation. In: 2004 International Symposium on Antennas and Propagation, vol. 1, pp. 69–72 (2004)Google Scholar
  13. 13.
    Tsuboi, K., Masuya, G.: Direct Numerical Simulations for Instabilities of Remixed Planar Flames. In: The Fourth Asia-Pacific Conference on Combustion, Nanjing, China (November 2003)Google Scholar
  14. 14.
    Nakajima, M., et al.: Numerical Simulation of Three-Dimensional Separated Flow and Heat Transfer around Staggerd Surface-Mounted Rectangular Blocks in a Channel. Numerical Heat Transfer, Part A 47, 691–708 (2005)CrossRefGoogle Scholar
  15. 15.
    Ariyoshi, K., et al.: Spatial variation in propagation speed of postseismic slip on the subducting plate boundary. In: 2nd Water Dynamics, vol. B-30, Sendai, Japan (2004)Google Scholar

Copyright information

© Springer-Verlag Berlin Heidelberg 2006

Authors and Affiliations

  • Akihiro Musa
    • 1
    • 2
  • Hiroyuki Takizawa
    • 1
  • Koki Okabe
    • 1
  • Takashi Soga
    • 3
  • Hiroaki Kobayashi
    • 1
  1. 1.Tohoku UniversitySendaiJapan
  2. 2.NEC CorporationTokyoJapan
  3. 3.NEC System TecnologiesOsakaJapan

Personalised recommendations