Frontal methods are an efficient and pop
ular means of Gauss elimination of matrix
equations that arise in finite element
analysis. Nested dissection of a computa
tional domain makes possible high-level
parallelism in a widely used frontal algo
rithm for unsymmetric systems. A concur
rent, highly vectorized, multifrontal, finite
element analysis of axisymmetric liquid
drop oscillations with 2,210 equations runs
on the CRAY X-MP/48 with factors of 1.9
and 2.9 reduction in elapsed time on two
and four processors, respectively. On an
ELXSI 6400 (which has an additional
memory level, local processor cache, ig
nored in the algorithm's design for the
CRAY), implementation of the same
problem initially achieved a speedup of
only 1.4 on four processors. Modification
of the concurrent algorithm, to take ad
vantage of the cache and frontwidth re
duction by element reordering, doubled the
concurrent speedup on the ELXSI to 2.8 on
four processors.
Get full access to this article
View all access options for this article.
References
1.
Berger, M.J., and Bokhari, S.H.1985. A partitioning strategy for PDE's across multiprocessors . Proc. 1985 Internat. Canf. Parallel Proc. Washington, DC: IEEE Computer Society, pp. 166-170.
2.
Bixler, N.E., and Benner, R.E.1985. Finite element analysis of axisymmetric oscillations of sessile liquid drops. Proc. 4th Internal. Conf. Numer. Methods Lam. Turb. Flow. Swansea, U.K.: Pineridge Press, pp. 1336-1347.
3.
Boudouvis, A.G., and Scriven, L.E.1985. Explicidy vectorized frontal routine for hydrodynamic stability and bifurcation analysis by Galerkin/finite element methods. Proc. Supercomputer Appl. Symp. New York: Plenum, pp. 197-213.
4.
Duff, I.S.1985. Parallel implementation of multifrontal schemes. ANL/ MCS-TM-49Argonne, Illinois: Argonne National Laboratory.
5.
Duff, I.S., and Reid, J.K.1982. Experience of sparse matrix codes on the CRAY-1 . Comput. Phys. Comm.26:293-302.
6.
Duff, I.S., and Reid, J.K.1983. The multifrontal solution of indefinite sparse linear equations. ACM Trans. Math. Software9:302-325.
7.
Duff, I.S., and Reid, J.K.1984. The multifrontal solution of unsymmetric sets of linear equations. SIAM J. Sci. Statist. Comput.5:633-641.
8.
Englemann, M.S., Strang, G., and Bathe, K.J.1981. The application of quasi-Newton methods in fluid mechanics . Internat. J. Numer Methods Engrg.17:707-718.
9.
Flower, J.W., Otto, S.W., and Salama, M.C.1986. A preprocessor for irregular finite element problems. C3P-292Pasadena: California Institute of Technology.
10.
Fox, G.C., and Furmanski, W.1986. Load balancing by a neural network. C3P-363Pasadena: California Institute of Technology .
11.
Gresho, P.M., Lee, R.L., and Sani, R.C.1979. On the time dependent solution of the incompressible Navier-Stokes equations in two and three dimensions. Recent Adv. Numer. Methods Fluids1:27-81.
12.
Hood, P.1976. Frontal solution program for unsymmetric matrices . Internat. J. Numer. Methods Engrg.10:379-399.
13.
Irons, R.M.1970. A frontal solution program for finite element analysis . Internat. J. Numer. Methods. Engrg.2:5-32.
14.
Jalby, W., and Meier, U.1986. Optimizing matrix operations on a parallel multiprocessor with a memory heirarchy. CSRD Report. Urbana-Champaign : University of Illinois.
15.
Montry, G.R.1987. A portable system for simulating distributed memory parallel processors on shared memory machines. Hypercube Multiprocessors 1987 . Philadephia: SIAM, pp. 276-282.
16.
Montry, G.R., and Benner, R.E.1987. Parallel processing on an ELXSI 6400. Proc. 2nd Internat. Conf. Supercomputing, St. Petersburg, Florida: ISI Inc., 2:64-71.
17.
Silvester, P.P., Suda, H.A., and Stone, G.D.1984. A memory-economic frontwidth reduction algorithm . Internat. J. Numer. Methods Engrg.20:733-743.
18.
Strang, G., and Fix, G.J.1973. An analysis of the finite element method. Englewood Cliffs, New Jersey: Prentice-Hall .
