Matrix-free approaches for GPU acceleration of a high-order finite element hydrodynamics application using MFEM, Umpire, and RAJA

Abstract

With the introduction of advanced heterogeneous computing architectures based on GPU accelerators, large-scale production codes have had to rethink their numerical algorithms and incorporate new programming models and memory management strategies in order to run efficiently on the latest supercomputers. In this work we discuss our co-design strategy to address these challenges and achieve performance and portability with MARBL, a next-generation multi-physics code in development at Lawrence Livermore National Laboratory. We present a two-fold approach, wherein new hardware is used to motivate both new algorithms and new abstraction layers, resulting in a single source application code suitable for a variety of platforms. Focusing on MARBL’s ALE hydrodynamics package, we demonstrate scalability on different platforms and highlight that many of our innovations have been contributed back to open-source software libraries, such as MFEM (finite element algorithms) and RAJA (kernel abstractions).

Review of Literature on Paralleling Finite Element Schemes for Solving Nonlinear Structure Deformation Problems On Super Computers

“…С появлением передовых архитектур гетерогенных вычислений на основе ускорителей GPU разработчикам крупномасштабных производственных программ пришлось переосмыслить свои численные алгоритмы и включить новые модели программирования и стратегии управления памятью, чтобы эффективно использовать новейшие суперкомпьютеры. A. Vargas et al в [40] обсуждают стратегию решения этих проблем, повышения производительности и достижения переносимости с помощью мультифизического программного кода нового поколения MARBL, разрабатываемого в Ливерморской национальной лаборатории им. Э. Лоуренса.…”

Water Environmental Pollution Risk Assessment and Water Pollution Dispersion Simulation by Integrating GIS and Random Forest Algorithm

“…However, it can only be used for vector data spaces. Furthermore, depending on the regularity of the spatial events, raster data or weakly structured data cannot be expressed [13][14].…”