Load balancing for the parallel adaptive solution of partial differential equations
The authors are concerned with a problem that arises when one wishes to exploit the advantages both of parallel computation and of adaptive finite element strategies. The problem is that adaptivity leads to a restructuring of the finite element mesh, with a consequent redistribution of processor loading. This can lead to processor load imbalances which militate against an efficient solution procedure. The authors discount recursive bisection methods as a procedure for partitioning the domain into subdomains with balanced loading, due to the expense that would be entailed in using this approach in conjunction with an adaptive method. Instead, they propose three strategies for resolving the problem. The first, tiling, is based on a dynamic load balancing technique due to \textit{E. Leiss} and \textit{H. N. Reddy} [Distributed load balancing: design and performance analysis. W. M. Keck Research Computation Laboratory 5, 205-270 (1989)]. It is applicable to two-dimensional structured meshes. For unstructured meshes or for three-dimensional problems it is not a feasible method since it can be expensive. Thus an alternative procedure, redistribution through pairwise exchanges, is introduced to deal with such problems. It builds on tiling, but exploits graph theoretical ideas systematically and effectively. The last alternative is octree decomposition, which is also suitable for three-dimensional unstructured meshes, and which exploits the properties of the underlying tree structure. The three methods proposed in this paper are all tested on a wide range of examples, which illustrate very clearly their performance in practice.
- Adaptive mesh refinement for hyperbolic partial differential equations
- An adaptive mesh-moving and local refinement method for time-dependent partial differential equations
- An efficient communications strategy for finite element methods on the Connection Machine CM-5 system
- An expert system for the optimal mesh design in thehp-version of the finite element method
- An h-p adaptive finite element method for the numerical simulation of compressible flow
- An Improved Spectral Graph Partitioning Algorithm for Mapping Parallel Computations
- Automatic three‐dimensional mesh generation by the finite octree technique
- High-order adaptive methods for parabolic systems
- High-Order Finite Element Methods for Singularly Perturbed Elliptic and Parabolic Problems
- scientific article; zbMATH DE number 3885493 (Why is no real title available?)
- scientific article; zbMATH DE number 4089080 (Why is no real title available?)
- scientific article; zbMATH DE number 107951 (Why is no real title available?)
- scientific article; zbMATH DE number 822241 (Why is no real title available?)
- scientific article; zbMATH DE number 3417498 (Why is no real title available?)
- Parallel, adaptive finite element methods for conservation laws
- Partitioning Sparse Matrices with Eigenvectors of Graphs
- The Runge-Kutta Local Projection Discontinuous Galerkin Finite Element Method for Conservation Laws. IV: The Multidimensional Case
- TVB Runge-Kutta Local Projection Discontinuous Galerkin Finite Element Method for Conservation Laws II: General Framework
- TVB Runge-Kutta local projection discontinuous Galerkin finite element method for conservation laws. III: One-dimensional systems
- A multi-level diffusion method for dynamic load balancing
- The Runge-Kutta discontinuous Galerkin method for conservation laws. I: Multidimensional systems
- Parallel structures and dynamic load balancing for adaptive finite element computation
- A load balancing package on distributed memory systems and its application to particle-particle particle-mesh (P3M) methods
- Scalability and load balancing in adaptive algorithms for multidimensional integration
- Parallel computing with load balancing on heterogeneous distributed systems.
- An adaptive load balancing method for parallel molecular dynamics simulations
- Dynamic load balancing in computational mechanics
- Shape-optimized mesh partitioning and load balancing for parallel adaptive FEM
- Parallel load balancing for dynamic execution environments
- A balanced accumulation scheme for parallel PDE solvers
- Simultaneous mesh generation and partitioning for Delaunay meshes
- Multithreaded model for the dynamic load-balancing of parallel adaptive PDE computations
- Parallel adaptive \(hp\)-refinement techniques for conservation laws
- A prefix code matching parallel load-balancing method for solution-adaptive unstructured finite element graphs on distributed memory multicomputers
- A versatile load balancing framework for parallel applications based on domain decomposition
- A new paradigm for parallel adaptive meshing algorithms
- Error estimates and adaptive finite element methods. A bibliography (1990--2000)
- scientific article; zbMATH DE number 4215280 (Why is no real title available?)
- Sur la modélisation des ondes élastiques dans les structures : calculs adaptatif et parallèle avec équilibrage de charges
- Load balancing for adaptively refined grids
- scientific article; zbMATH DE number 1183041 (Why is no real title available?)
- Load Balancing for Adaptive Multigrid Methods
- Parallel adaptive finite element analysis of viscous flows based on a combined compressible‐incompressible formulation
- An algorithm for domain partitioning with load balancing
- scientific article; zbMATH DE number 833722 (Why is no real title available?)
- scientific article; zbMATH DE number 934528 (Why is no real title available?)
- scientific article; zbMATH DE number 1446089 (Why is no real title available?)
- Euro-Par 2004 Parallel Processing
- scientific article; zbMATH DE number 956667 (Why is no real title available?)
- Parallel dynamic load balancing strategies for adaptive irregular applications
- A new metric for dynamic load balancing
- Guaranteed-quality parallel Delaunay refinement for restricted polyhedral domains
- Load balancing using heterogeneous processors for continuum problems on a mesh
This page was built for publication: Load balancing for the parallel adaptive solution of partial differential equations
Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q1344328)
