Advertisement

Vectorized Parallel Solver for Tridiagonal Toeplitz Systems of Linear Equations

  • Beata DmitrukEmail author
  • Przemysław Stpiczyński
Conference paper
  • 149 Downloads
Part of the Lecture Notes in Computer Science book series (LNCS, volume 12043)

Abstract

The aim of this paper is to present two versions of a new divide and conquer parallel algorithm for solving tridiagonal Toeplitz systems of linear equations. Our new approach is based on a recently developed algorithm for solving linear recurrence systems. We discuss how to reduce the number of necessary synchronizations and show proper data layout that allows to use cache memory and SIMD extensions of modern processors. Numerical experiments show that our new implementations achieve very good seedup on multicore and manycore architectures. Moreover, they are more energy efficient than a simple sequential algorithm.

Keywords

Tridiagonal Toeplitz systems Parallel algorithms Vectorization SIMD extensions OpenMP Energy efficiency 

Notes

Acknowledgements

The use of computer resources installed at Maria Curie-Skłodowska University in Lublin is kindly acknowledged.

References

  1. 1.
    Belhaj, S., Dridi, M.: A fast algorithm of two-level banded Toeplitz systems of linear equations with application to image restoration. New Trends Math. Sci. 2, 277–283 (2017).  https://doi.org/10.20852/ntmsci.2017.178CrossRefGoogle Scholar
  2. 2.
    Chung, K.L., Yan, W.M.: Parallel B-spline surface fitting on mesh-connected computers. J. Parallel Distrib. Comput. 35, 205–210 (1996).  https://doi.org/10.1006/jpdc.1996.0082CrossRefGoogle Scholar
  3. 3.
    Chung, K.L., Yan, W.M.: Vectorized algorithms for solving special tridiagonal systems. Comput. Math. Appl. 32, 1–14 (1996).  https://doi.org/10.1016/S0898-1221(96)00203-9CrossRefMathSciNetzbMATHGoogle Scholar
  4. 4.
    Du, L., Sogabe, T., Zhang, S.L.: A fast algorithm for solving tridiagonal quasi-Toeplitz linear systems. Appl. Math. Lett. 75, 74–81 (2018).  https://doi.org/10.1016/j.aml.2017.06.016CrossRefMathSciNetzbMATHGoogle Scholar
  5. 5.
    Garey, L., Shaw, R.: A parallel method for linear equations with tridiagonal Toeplitz coefficient matrices. Comput. Math. Appl. 42(1), 1–11 (2001).  https://doi.org/10.1016/S0898-1221(01)00125-0CrossRefMathSciNetzbMATHGoogle Scholar
  6. 6.
    Khan, K.N., Hirki, M., Niemi, T., Nurminen, J.K., Ou, Z.: RAPL in action: experiences in using RAPL for power measurements. ACM Trans. Model. Perform. Eval. Comput. Syst. 3(2), 9:1–9:26 (2018).  https://doi.org/10.1145/3177754CrossRefGoogle Scholar
  7. 7.
    McNally, J.M., Garey, L.E., Shaw, R.E.: A split-correct parallel algorithm for solving tridiagonal symmetric Toeplitz systems. Int. J. Comput. Math. 75(3), 303–313 (2000).  https://doi.org/10.1080/00207160008804986CrossRefMathSciNetzbMATHGoogle Scholar
  8. 8.
    McNally, J.M., Garey, L., Shaw, R.: A communication-less parallel algorithm for tridiagonal Toeplitz systems. J. Comput. Appl. Math. 212, 260–271 (2008).  https://doi.org/10.1016/j.cam.2006.12.001CrossRefMathSciNetzbMATHGoogle Scholar
  9. 9.
    Rojo, O.: A new method for solving symmetric circulant tridiagonal systems of linear equations. Comput. Math. Appl. 20, 61–67 (1990).  https://doi.org/10.1016/0898-1221(90)90165-GCrossRefMathSciNetzbMATHGoogle Scholar
  10. 10.
    Stpiczyński, P.: Solving linear recurrence systems using level 2 and 3 BLAS routines. In: Wyrzykowski, R., Dongarra, J., Paprzycki, M., Waśniewski, J. (eds.) PPAM 2003. LNCS, vol. 3019, pp. 1059–1066. Springer, Heidelberg (2004).  https://doi.org/10.1007/978-3-540-24669-5_137CrossRefzbMATHGoogle Scholar
  11. 11.
    Stpiczyński, P., Potiopa, J.: Piecewise cubic interpolation on distributed memory parallel computers and clusters of workstations. In: Fifth International Conference on Parallel Computing in Electrical Engineering (PARELEC 2006), Bialystok, Poland, 13–17 September 2006, pp. 284–289. IEEE Computer Society (2006).  https://doi.org/10.1109/PARELEC.2006.68
  12. 12.
    Stpiczyński, P., Potiopa, J.: Solving a kind of boundary-value problem for ordinary differential equations using Fermi—the next generation CUDA computing architecture. J. Comput. Appl. Math. 236, 384–393 (2011).  https://doi.org/10.1016/j.cam.2011.07.028CrossRefMathSciNetzbMATHGoogle Scholar
  13. 13.
    Terekhov, A.V.: A highly scalable parallel algorithm for solving Toeplitz tridiagonal systems of linear equations. J. Parallel Distrib. Comput. 87, 102–108 (2016).  https://doi.org/10.1016/j.jpdc.2015.10.004CrossRefGoogle Scholar
  14. 14.
    Vidal, A.M., Alonso, P.: Solving systems of symmetric Toeplitz tridiagonal equations: Rojo’s algorithm revisited. Appl. Math. Comput. 219, 1874–1889 (2012).  https://doi.org/10.1016/j.amc.2012.08.030CrossRefMathSciNetzbMATHGoogle Scholar
  15. 15.
    Wang, H.H.: A parallel method for tridiagonal equations. ACM Trans. Math. Softw. 7(2), 170–183 (1981).  https://doi.org/10.1145/355945.355947CrossRefMathSciNetzbMATHGoogle Scholar

Copyright information

© Springer Nature Switzerland AG 2020

Authors and Affiliations

  1. 1.Institute of Computer ScienceMaria Curie–Skłodowska UniversityLublinPoland

Personalised recommendations