GPU-Accelerated Discontinuous Galerkin Methods on Polytopic Meshes

Discontinuous Galerkin (dG) methods on meshes consisting of polygonal/polyhedral (henceforth, collectively termed as polytopic) elements have received considerable attention in recent years. Due to the physical frame basis functions used typically and quadrature challenges involved, matrix-assembly step for these is often computationally cumbersome. To address this important practical issue, work proposes two parallel assembly implementation algorithms Compute Unified Device Architecture--enabled graphics cards interior penalty dG method polytopic various classes linear PDE problems. We are concerned with both single processing unit (GPU) parallelization, well distributed GPU nodes. The results included showcase almost scalability respect number cores since no communication needed step. In turn, can justify claim that be implemented extremely efficiently, any computing time overhead compared finite “standard” simplicial or box-type effectively circumvented by proposed algorithms.