Abstract
In this paper we discuss design principles to implement a set of linear algebra subroutines in a portable libraty for parallel computers. Our design supports reuse of code and easy adaption to new parallel programming paradigms or network configurations.
The routines are supposed to be used in self-verifying algorithms. They therefore have to deliver a validated result of high accuracy.
Abstract
Обсуждаются принципы построения и реализации набора линейно-алгебраических подпрограмм переносимой библиотеки для параллельных компьютеров. Поддерживается повторное использование кода. Подпрограммы могуг легко адаптироваться к новым парадигмам параллельного программирования и сетевым конфигураниям.
Подпрограммы предназначены для использования в самоверифицирующих алгоритмах, позтому они дают достоверные результаты высокой точности.
Similar content being viewed by others
References
Bohlender, G. and Wolff von Gudenberg, J.Accurate matrix multiplication on the array processor AMT-DAP. In: Kaucher, Markov, and Mayer (eds) “Computer Arithmetic, Scientific Computation and Mathematical Modelling”, IMACS Annals on Computing and Applied Mathematics 12, Baltzer, Basel, 1992.
Choi, J., Dongarra, J., and Walker, D.LAPACK working note 57: PUMMA: parallel universal matrix multiplication algorithms on distributed memory concurrent computers. University of Tenessee, TR CS-93-187, 1993.
Dongarra, J., Pozo, R., and Walker, D.LAPACK working note 61: an object oriented design for high performance linear algebra on distributed memory architectures. University of Tenessee, TR CS-93-200, 1993.
Jézéquel, J. M., Bergheul, F., and André, F.Programming massively parallel arcitectures with sequential object oriented languages. FGCS10 (1) (1994), pp. 59–70.
Klatte, R., Kulisch, U., Neaga, M., Ratz, D., and Ullrich, Ch.PASCAL-XSC-language reference with examples. Springer, Berlin, 1992.
Klatte, R., Kulisch, U., Lawo, C., Rauch, M. and Wiethoff, A.C-XSC, a C++ class library for extended scientific computing. Springer, Berlin, 1993.
Reith, R.Wissenschaftliches Rechnen auf Multicomputern—BLAS-Routinen und die Lösung linearer Gleichungssysteme mit Fehlerkontrolle. Dissertation, Universität Basel, 1993.
Wolff von Gudenberg, J.Modelling SIMD-type parallel arithmetic operations in Ada. In: Christodoulakis, D. (ed.) “Ada: The Choice for ’92”, LNCS 499, Springer, Berlin, 1991.
Wolff von Gudenberg, J.Accurate matrix operations on hypercube computers. In: Herzberger, J. and Atanassova, L. (eds) “Computer Arithmetic and Enclosure Methods”, North-Holland, Amsterdam, 1992.
Wolff von Gudenberg, J.Parallel accurate linear algebra subroutines. Reliable Computing1 (2) (1995), pp. 189–199.
Author information
Authors and Affiliations
Rights and permissions
About this article
Cite this article
von Gudenberg, J.W. Design of a parallel linear algebra library for verified computation. Reliable Comput 1, 411–419 (1995). https://doi.org/10.1007/BF02391686
Received:
Revised:
Issue Date:
DOI: https://doi.org/10.1007/BF02391686