Many numerical algorithms and scientific computations on unstructured meshes can be viewed as the independent application of a local operation everywhere on a mesh. This local operation is often called a computational kernel and its independent application lends itself naturally to parallel computation. An unstructured mesh can be described by sets of entities (vertices, edges, cells) and the connectivity between those sets forming the topology of the mesh. PyOP2 is a domain-specific language (DSL) for the parallel executions of computational kernels on unstructured meshes or graphs.
Keywords for this software
References in zbMATH (referenced in 5 articles )
Showing results 1 to 5 of 5.
- Luporini, Fabio; Ham, David A.; Kelly, Paul H.J.: An algorithm for the optimization of finite element integration loops (2017)
- Robert C. Kirby, Lawrence Mitchell: Solver composition across the PDE/linear algebra barrier (2017) arXiv
- Lange, Michael; Mitchell, Lawrence; Knepley, Matthew G.; Gorman, Gerard J.: Efficient mesh management in firedrake using PETSc DMPlex (2016)
- Mitchell, Lawrence; Müller, Eike Hermann: High level implementation of geometric multigrid solvers for finite element problems: applications in atmospheric modelling (2016)
- Fabio Luporini, Ana Lucia Varbanescu, Florian Rathgeber, Gheorghe-Teodor Bercea, J. Ramanujam, David A. Ham, Paul H.J. Kelly: COFFEE: an Optimizing Compiler for Finite Element Local Assembly (2014) arXiv