chapter  7
18 Pages

Mixed-Precision GPU-Multigrid Solvers with Strong Smoothers

ByDominik Go¨ddeke and Robert Strzodka

In this chapter, we present efficient fine-grained parallelization techniques for robust multigrid solvers, in particular for numerically strong, inherently sequential smoothing operators. We apply them to sparse ill-conditioned linear systems of equations that arise from grid-based discretization techniques like finite differences, volumes and elements. Our exemplary results demonstrate both the numerical and runtime performance of these techniques, as well as significant speedups over conventional CPUs.