Plenary Lecture

Sparse Approximate Inverse Preconditioners on High Performance GPU Platforms

Professor Daniele Bertaccini
Department of Mathematics
Universita' di Roma "Tor Vergata"
Italy
E-mail: bertaccini@mat.uniroma2.it

Abstract: The search for e cient preconditioners in factorized form for Krylov subspace methods is a crucial theme underlying numerical linear algebra research for large problems over the last decades. Recent developments, especially in computing hardware, have renewed the interest in updates for approximate inverse preconditioners for large and sparse linear systems, because this can mitigate the high computational cost required for forming the underlying factors. Comparative runs are proposed focusing on the approximate inverse preconditioners proposed by Benzi and Tuma from 1996 and the sparsication and inversion proposed by van Duin in 1999. Computational costs, reorderings and implementations with some other issues are considered both on conventional and innovative computing architectures like GPUs.

Brief Biography of the Speaker: Daniele Bertaccini (Ph.D in Mathematics) is an Associate Professor of Numerical Analysis at Universita' di Roma "Tor Vergata", Facolta' di Scienze, Dipartimento di Matematica).
Bertaccini's work in matrix computation devises and analyzes algorithms for solving numerical multidimensional problems arising in science and engineering.
He develops and analyzes fast algorithms for solving large scale systems in
- numerical time integration for evolutionary partial differential equations
- image restoration
- advection-diffusion-reaction systems from life sciences
- optimization
- ordinary differential equations

Bulletin Board

Currently:

The Conference Program is online.

The Conference Guide is online.

The paper submission deadline has expired. Please choose a future conference to submit your paper.

Plenary Speakers

WSEAS Main Site

Publication Ethics and Malpractice Statement