With the end of Moore's law and increasing concerns about operating costs, high performance computing (HPC) systems are undergoing drastic changes which are making traditional parallelization strategies suboptimal. In this seminar, we aim to discuss some of the broad design constraints and trends for these modern HPC systems. Our talk will specifically focus on addressing the growth in concurrency using the C++ runtime library, High Performance ParallelX (HPX). We will present scaling results comparing HPX's performance to a MPI parallelization for a discontinuous Galerkin Kernel. Results will be shown on NERSC's Cori, with a more traditional Haswell architecture and TACC's Stampede2, with the Knights Landing architecture.