Parallel Calculation of Properties of Magnetic Impurities in Metais

  • Eloiza Sonoda USP
  • Gonzalo Travieso USP

Resumo


The computation of physical properties of dilute magnetic alloys is a computationally intensive task. In principie, it consists on the diagonalization of a huge matrix. Through the use of a suitable base and of the renormalization group method, the diagonalization can be simplified to the iterative diagonalization of a set of smaller matrices (of the order of some hundreds of lines and columns). The resulting problem is nevertheless still too large if enough precision is desired. We describe here the parallel implementation of an algorithm for this problem in a duster. The results show that the implementation achieves good reduction of execution times, but lacks speedup and scalability due to load balancing problems. We analyze these problems and suggest paths to their resolution.

Palavras-chave: parallel processing, clusters, message passing, Anderson model, renormalization group

Referências

S. Alexander and P. W. Anderson. Interaction between localized states in metals. Phys Rev, 133(6), 1964.

T. E. Anderson. D. E. Culler, and D. A. Patterson. A case for now. IEEE Micro, 1(15), 1996.

P. W. Anderson. Localized magnetic states in metals. Phys Rev, 124(1), 1961.

L. Blackford, J. Choi, A. Cleary, E. D'Azevedo, J. Demmel, I. Dhillon. J. Dongarra. S. Hammarling, G. Henry, A. Petitet, K. Stanley, D. Walker, and R. C. Whaley. ScalApack User's Guide. SIAM, 1997.

S. C. Costa, C. A. Paula, V. L. Libero, and L. N. Oliveira. Numerical renormalization-group computation of specific heats. Phys. Rev. 8, 55(1), 1997.

David E. Culler. Jaswinder Pal Sing, and Anoop Gupta. Parallel Computer Architecture. A Hardware/Software Approach. Morgan Kaufmann. 1999.

C. A. de Paula. Densidade espectral para o modelo de Anderson de duas impurezas. PhD thesis, IFSC-USP, 1998.

lan Foster. Designing and Building Parallel Programs. Addison Wesley. 1995.

Free Software Foundation. The gnu project, http://www.gnu.org. visited 2001/05/04.

Willian Gropp and Edwing Lusk. User's Guide for MPICH, A Portable lmplementation of MPI. Argonne National Laboratory. 1996.

The Linux Kernel. http://www.kernel.org. visited 2001/05/04.

H. R. Krishna-Munhy, J. W. Wilkins, and K. G. Wilson. Renormalization-group approach to the anderson model of dilute magnetic alloys. Phys Rev 8, 21(3}, 1980.

Message Passing Interface Forum. MPI: A Message-Passing lnterface Standard, 1995.

L. N. Oliveira. The numerical renormalization group and the problem of impurities in metals. Braz. J. of Phys., 22(3). 1992.

T. Sterling. Beowulf: a parallel workstation for scientific computing. In Proc. of the 1995 Intl. Conf. on Parallel Processing, 1995.

Bjame Stroustrup. The C++ Programming Language. Addison Wesley, 3rd edition edition, 1997.

K. G. Wilson. The renormalization group: critical phenornena and the kondo problem. Rev of Mod Physics. 47(4), 1975.

M. Yoshida, M. A. Whitaker, and L. N. Oliveira. Renormalization-group calculation of excitation propenies for impurity models. Phys. Rev. B, 41(3). 1990.
Publicado
10/09/2001
SONODA, Eloiza; TRAVIESO, Gonzalo. Parallel Calculation of Properties of Magnetic Impurities in Metais. In: INTERNATIONAL SYMPOSIUM ON COMPUTER ARCHITECTURE AND HIGH PERFORMANCE COMPUTING (SBAC-PAD), 13. , 2001, Pirenópolis. Anais [...]. Porto Alegre: Sociedade Brasileira de Computação, 2001 . p. 38-43. DOI: https://doi.org/10.5753/sbac-pad.2001.22190.