ScaLAPACK: A Portable Linear Algebra Library for Distributed Memory Computers – Design Issues and Performance

J. Choi, J. Demmel, Inderjit Dhillon, J. Dongarra, S. Ostrouchov, A. Petitet, K. Stanley, D. Walker, R. Whaley

Abstract:   This paper outlines the content and performance of ScaLAPACK, a collection of mathematical software for linear algebra computations on distributed memory computers. The importance of developing standards for computational and message passing interfaces is discussed. We present the different components and building blocks of ScaLAPACK. This paper outlines the difficulties inherent in producing correct codes for networks of heterogeneous processors. We define a theoretical model of parallel computers dedicated to linear algebra applications: the Distributed Linear Algebra Machine (DLAM). This model provides a convenient framework for developing parallel algorithms and investigating their scalability, performance and programmability. Extensive performance results on various platforms are presented and analyzed with the help of the DLAM. Finally, this paper briefly describes future directions for the ScaLAPACK library and concludes by suggesting alternative approaches to mathematical libraries, explaining how ScaLAPACK could be integrated into efficient and user-friendly distributed systems.

Download: pdf

Citation

  • ScaLAPACK: A Portable Linear Algebra Library for Distributed Memory Computers – Design Issues and Performance (pdf)
    J. Choi, J. Demmel, I. Dhillon, J. Dongarra, S. Ostrouchov, A. Petitet, K. Stanley, D. Walker, R. Whaley.
    Computer Physics Communications 97, pp. 1-15, August 1996.

    Bibtex: