The Parallelization of an Adaptive Multigrid Eigenvalue Solver with LPARX
نویسندگان
چکیده
We have developed a parallel adaptive eigenvalue solver and applied it to a model problem in theoretical materials science. Our method combines adaptive mesh reenement techniques with a novel multigrid eigenvalue algorithm. By exploiting adaptivity, we have reduced computation time and memory consumption by more than two orders of magnitude. We have implemented our solver using the LPARX parallel programming system, which considerably simpliied the programming and enabled us to run the same code on a diversity of high performance parallel architectures.
منابع مشابه
Using Skeletons to Implement a Parallel Multigrid Method with Overlapping Adaptive Grids
Algorithmic skeletons are polymorphic higher-order functions that represent common parallelization patterns. They can be used as the building blocks of parallel applications by integrating them into a sequential language. In this paper we present a skeleton-based approach to manage overlapping of distributed grids, which occur in parallel adaptive multigrid algorithms. Overlapping is necessary ...
متن کاملA Parallel Solver for Extreme Eigenpairs1
In this paper a parallel algorithm for nding a group of extreme eigenvalues is presented. The algorithm is based on the well known Davidson method for nding one eigenvalue of a matrix. Here we incorporate knowledge about the structure of the subspace through the use of an arrowhead solver which allows more parallelization in both the original Davidson and our new version. In our numerical resul...
متن کاملA Robust Parallel Programming Model forDynamic Non - Uniform Scienti c
LPARX provides eecient run-time support for dynamic , non-uniform scientiic calculations running on MIMD distributed memory architectures. It extends HPF's data decomposition model to provide support for dynamic, block irregular data structures. LPARX represents data decompositions as rst-class objects and expresses data dependencies in a manner which is logically independent of data decomposit...
متن کاملHigh-Level versus Low-Level DO-Loop Parallelization: Results for One Testcase of a Multi-Block Solver on a Shared Memory Parallel Vector Computer
Within the NICE project, supported by the Dutch Foundation HPCN, the multi-block Navier-Stokes flow solver ENSOLV is being parallelized. In this article first results of the block-loop parallelization of ENSOLV are presented. We discuss the performance results of this parallelization on a 16-processor NEC SX-4, a shared memory parallel vector computer. The results are compared to those of the l...
متن کاملParallel Multigrid in an Adaptive PDE Solver Based on Hashing
Partial diierential equations can be solved eeciently by adaptive multigrid methods on a parallel computer. We report on the concept of hash-table storage techniques to set up such a code. The code requires substantial less amount of memory and is easier to code in the sequential case. The parallelization takes place by a space lling curve domain decomposition intimately connected to the hash t...
متن کاملذخیره در منابع من
با ذخیره ی این منبع در منابع من، دسترسی به آن را برای استفاده های بعدی آسان تر کنید
عنوان ژورنال:
دوره شماره
صفحات -
تاریخ انتشار 1995