We have evaluated the performance of
vectorized nuclear codes used at the
Japan Atomic Energy Institute. Forty nu
clear codes were restructured for vectori
zation, with an average vectorization of
83%. The vector to scalar ratio for modi
fied code performance is about 5. Vectori
zation improved performance in most
cases, but certain codes, including LOCA
and Monte Carlo codes, required consid
erable effort to vectorize, without high
performance gains. We discuss the bar
riers to achieving high performance, and
the need to consider applications charac
teristics when evaluating system
performance.
Get full access to this article
View all access options for this article.
References
1.
Asai, K., Ishiguro, M., and Matsuura, T.1983. Adaptability of supercomputers to nuclear computations . J. Atomic Energy Soc. Japan25(3):164-171 (in Japanese).
2.
Asai, K., Higuchi, K., and Katakura, J.1986. Vectorization of the KENO-IV code. Nuclear Sci. Engrg.92(1):298-307.
3.
Björck, A., and Anderson, N.1972. Numerical methods. Englewood Cliffs, New Jersey: Prentice-Hall.
4.
Hockney, R.W.1970. The potential calculation and some applications. In Methods in computational physics, Vol. 9. Alder, B. and Fernbach, S. eds. New York: Academic Press .
5.
Hockney, R.W., and Jesshope, C.R.1981. Parallel computers. Bristol, England: Adam Hilger Ltd.
6.
Ishiguro, M., Harada, H., and Shinozawa, N.1986. Vectorization of the light water reactor transient analysis code RELAP5. Nuclear Sci. Engrg.92(1):126-135.
7.
Lubeck, O., Moore, J., and Mendez, R.1985. A benchmark comparison of three supercomputers: Fujitsu VP-200, Hitachi S810/20, and Cray X-MP/2. IEEE Computer . 18(12):10-24.
8.
Tamura, H., Kamiya, S., and Ishigai, T.1985. FACOM VP-100/200: Supercomputers with ease of use . Parallel Comput.2(2):87-107.
9.
Wilkinson, J.H., and Reinsch, C.1971. Linear algebra: Handbook for automatic computation , Vol. 2. New York: Springer-Verlag.
