Implementation Of Lanczos Algorithms on Vector Computers

  • B. N. Parlett
  • B. Nour-Omid
  • J. Natvig


We report on two recent studies [17],[18] in which the Lanczos algorithm was transported to a vector machine. The results suggest that our present implementations cannot exploit the full power of Class VI machines, nor even half of it. Yet the nature of the Lanczos algorithm lends itself to vectorization.


Vector Length Operation Count Lanczos Algorithm Page Fault Double Eigenvalue 
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]
    K.J.Bathe and E.L.Wilson, “Numerical Methods in Finite Element Analysis”, (Prentice Hall, Englewood Cliffs, New Jersey, 1977).Google Scholar
  2. 2.
    K.J.Bathe and E.L.Wilson, Large eigenvalue problems in dynamic analysis,A.S.C.E.,Journal of Engineering Mechanics Division99 (1973), 467–479.Google Scholar
  3. 3.
    CDC CYBER 200 FORTRAN Version 2 Reference Manual, Publication No. 60485000 (CDC, 1981).Google Scholar
  4. 4.
    CDC CYBER 200 Model 205 - Technical Description (CDC, 1980).Google Scholar
  5. 5.
    T.Ericsson, A.Ruhe, The spectral transformation Lanczos method for the numerical solution of large sparse generalized symmetric eigenvalue problems,Math. Comp. 34 (1980) 1251–1268.Google Scholar
  6. 6.
    B.S.Garbow etal., ‘Lecture Notes in Computer Science’, vol. 51, (Springer-Verlag, 1977).Google Scholar
  7. 7.
    M.J.Kascic, Jr., ‘Vector Processing on the CYBER 205’, (CDC CYBER 205 Workshop, Fort Collins, Colorado, February 1983).Google Scholar
  8. 8.
    C.Lanczos, An iteration method for the solution of the eigenvalue problem of linear differential and integral operators, >J. Res. Nat. Bur. Standard45 (1950) 255–282.Google Scholar
  9. 9.
    B.Nour-Omid, B.N.Parlett, R.L.Taylor, Lanczos versus subspace iteration for solution of eigenvalue problems,International Journal for Numerical Methods in Engineering19 (1983) 859–871.CrossRefGoogle Scholar
  10. 10.
    B.N.Parlett, ‘The Symmetric Eigenvalue Problem’, (Prentice-Hall, Englewood Cliffs, N.J., 1980).Google Scholar
  11. 11.
    B.N.Parlett, B.Nour-Omid, The use of refined error bounds when updating eigenvalues of tridiagonals, Technical Report PAM-175, (Center for Pure & Applied Math., Univ. of California, Berkeley, 1983).Google Scholar
  12. 12.
    Price Schedule for CYBER 205 Services, (CSU, Computer Center, Fort Collins, Colorado, 1982).Google Scholar
  13. 13.
    H.Rutishauser, Simultaneous iteration method for symmetric matrices, in: J.H.Wilkinson, C.H.Reinsch, eds., ‘Handbook for Automatic Computation (Linear Algebra) ’, (Springer-Verlag, New York, 1971) 284–302.Google Scholar
  14. 14.
    D.S.Scott, Block Lanczos software for symmetric eigenvalue problems, Report ORNL/CSD-48, UC-32 (Union Carbide Corporation, 1979).Google Scholar
  15. 15.
    H.D.Simon, The Lanczos algorithm for solving symmetric linear systems, Technical Report PAM-74 (Center for Pure & Applied Math., Univ. of California, Berkeley, 1982)Google Scholar
  16. 16.
    O.C. Zienkiewics, ‘The Finite Element Method’, 3rd edition, (McGraw-Hill, London, 1977).Google Scholar
  17. 17.
    John G. Lewis, Horst D. Simon, Numerical experience with the spectral transformation Lanczos method, Technical Report MM-TR-16, Boeing Computer Services, Seattle, WA 98124 (April 1984).Google Scholar
  18. 18.
    Jon Natvig, B.Nour-Omid, B.N.Parlett, Effect of the CYBER 205 on methods for computing natural frequencies of structures, Technical Report PAM-218 (Center for Pure & Applied Math., Univ. of California, Berkeley, March 1984).Google Scholar

Copyright information

© Plenum Press, New York 1985

Authors and Affiliations

  • B. N. Parlett
    • 1
  • B. Nour-Omid
    • 1
  • J. Natvig
    • 1
  1. 1.Center for Pure and Applied MathematicsUniversity of CaliforniaBerkeleyUSA

Personalised recommendations