A parallel element‐by‐element solution scheme

A parallel element-by-element scheme is developed for finite-element solution of elliptic boundary-value problems. It is shown that, for a broad class of computational grids, the solution algorithm is totally parallelizable. Moreover, the conversion of an existing serial EBE code to a parallel code is seen to be quite simple. The method is implemented on the Alliant FX/8 and Sequent Balance parallel computers and ‘speedup’ performance studies are conducted for a representative elliptic PDE in two dimensions. The present procedure can be applied quite generally to other finite-element applications, and the associated programs can be modified in a straightforward way to implement the method.

[1]  T. Hughes,et al.  An element-by-element solution algorithm for problems of structural and solid mechanics , 1983 .

[2]  Iain S. Duff,et al.  Parallel implementation of multifrontal schemes , 1986, Parallel Comput..

[3]  G. Carey,et al.  Element‐by‐element linear and nonlinear solution schemes , 1986 .

[4]  Thomas C. Oppe,et al.  Vectorized iterative methods for partial differential equations , 1986 .

[5]  Dennis Gannon,et al.  On the Impact of Communication Complexity on the Design of Parallel Numerical Algorithms , 1984, IEEE Transactions on Computers.

[6]  C. Micchelli,et al.  Polynomial Preconditioners for Conjugate Gradient Calculations , 1983 .

[7]  Geoffrey C. Fox,et al.  Algorithms for concurrent processors , 1984 .

[8]  Linda J. Hayes,et al.  A vectorized version of a sparse matrix‐vector multiply , 1986 .

[9]  Oliver A. McBryan,et al.  Parallel algorithms for elliptic equations , 1985 .

[10]  G. Carey.  Parallelism in finite element modelling , 1986 .

[11]  Graham F. Carey,et al.  Finite Element Analysis Using Advanced Processors , 1986, FJCC.

[12]  R. Courant.  Variational methods for the solution of problems of equilibrium and vibrations , 1943 .

[13]  Garry H. Rodrigue.  Some ideas for decomposing the domain of elliptic partial differential equations in the schwarz process , 1986 .

[14]  R. Fletcher.  Conjugate gradient methods for indefinite systems , 1976 .

[15]  Loyce Adams,et al.  Reordering computations for parallel execution , 1986 .

[16]  Mark Seager.  Overhead considerations for parallelizing conjugate gradient , 1986 .