TY - JOUR
T1 - Parallel elliptic preconditioners
T2 - Fourier analysis and performance on the connection machine
AU - Chan, Tony F.
AU - Jay Kuo, C. C.
AU - Tong, Charles
N1 - Funding Information:
* This work was supported in part by the Department of Energy under contract DE-F003-87ER25037, the National Science Foundation under contracts NSF-DMS87-14612 and BBS 87 14206, the Army Research Office under contract DAALO3-88-K-0085 and by the Research Institute for Ad-vanced Computer Science, NASA Ames.
PY - 1989/5
Y1 - 1989/5
N2 - We study the performance of several widely used preconditioners for 2D elliptic partial differential equations (SSOR, ILU, MILU and polynomial preconditioners) with the natural and red-black orderings implemented on the Connection Machine (CM). Their performance is primarily influenced by two factors: the rate of convergence and the ease of parallelization. The convergence rate is analyzed by Fourier analysis and confirmed with experimental results. Although the naturally ordered SSOR and MILU preconditioners have convergence rates one order faster than the other preconditioners, the experiments show that the red-black ordered SSOR, ILU, MILU, polynomial preconditioners takes less execution time than their naturally ordered counterparts. This is partially due to the fact that the red-blavk ordering provides more parallelism than the natural ordering.
AB - We study the performance of several widely used preconditioners for 2D elliptic partial differential equations (SSOR, ILU, MILU and polynomial preconditioners) with the natural and red-black orderings implemented on the Connection Machine (CM). Their performance is primarily influenced by two factors: the rate of convergence and the ease of parallelization. The convergence rate is analyzed by Fourier analysis and confirmed with experimental results. Although the naturally ordered SSOR and MILU preconditioners have convergence rates one order faster than the other preconditioners, the experiments show that the red-black ordered SSOR, ILU, MILU, polynomial preconditioners takes less execution time than their naturally ordered counterparts. This is partially due to the fact that the red-blavk ordering provides more parallelism than the natural ordering.
UR - http://www.scopus.com/inward/record.url?scp=0024663831&partnerID=8YFLogxK
U2 - 10.1016/0010-4655(89)90163-X
DO - 10.1016/0010-4655(89)90163-X
M3 - Article
AN - SCOPUS:0024663831
SN - 0010-4655
VL - 53
SP - 237
EP - 252
JO - Computer Physics Communications
JF - Computer Physics Communications
IS - 1-3
ER -