SS 5: Advances in preconditioned iterative methods: theory, implementation and applications (B) Applications and High-Performance Computing

Organizers

Abstract

The special session is devoted to various aspects of high performance preconditioning both from theoretical and practical point of view. This covers but is not limited to recent developments in algebraic multilevel, multigrid, auxiliary space, and domain decomposition algorithms for adaptive finite element methods and nonstandard discretizations. The main focus of the session is on the construction of efficient and robust preconditioners for ill-conditioned and nearly singular problems as well as on the performance, scalability, and convergence analysis of preconditioned iterative methods. Contributions in the field of quantum physics and continuum mechanics, for example problems in geophysics, are especially encouraged.

List of speakers

  • Carmen Rodrigo (Zaragoza, Spain):
    Poroelasticity: stable numerical discretization and multi-grid solution
  • Pasqua D’Ambra (Naples, Italy):
    Performance and scalability analysis of an AMG with a prescribed convergence rate
  • Neil Budko (Delft, Netherlands):
    Convergence of Krylov methods for Schroedinger equation with random potential
  • James Brannick (State College, USA):
    Multigrid preconditioning of the overlap operator in lattice quantum chromodynamics

< Back