Event

Scalable Methods for Nonlocal Models

Sandia National Laboratories

Abstract: The naive discretization of nonlocal integral operators leads to matrices with significant density compared to the sparse systems arising from partial differential equations. This makes the efficient solution of nonlocal models a challenging task.  I will discuss ongoing research into efficient hierarchical matrix assembly and multilevel solvers (geometric/algebraic multigrid and domain decomposition) that are suitable for nonlocal models.

Speaker’s Bio: Christian Glusa is a staff member in the Center for Computing Research at Sandia National Laboratories.  His interests include scalable linear solvers, such as multigrid and domain decomposition methods, and their application to problems in electromagnetics and fractional order equations.  Christian received a Ph.D. in applied mathematics from Brown University in 2017, and undergraduate degrees from Karlsruhe Institute of Technology and Ecole Polytechnique in 2012.

Last Updated: June 29, 2023 - 7:33 am