4 Hits in 4.9 sec

Parallel Iterative Solvers of GeoFEM with Selective Blocking Preconditioning for Nonlinear Contact Problems on the Earth Simulator

Kengo Nakajima
2003 Proceedings of the 2003 ACM/IEEE conference on Supercomputing - SC '03  
An efficient parallel iterative method with selective blocking preconditioning has been developed for symmetric multiprocessor (SMP) cluster architectures with vector processors such as the Earth Simulator  ...  This method provides robust and smooth convergence and excellent vector and parallel performance in 3D geophysical simulations with contact conditions performed on the Earth Simulator.  ...  Acknowledgments This study is a part of the Solid Earth Platform for Large-Scale Computation project funded by the Ministry of Education, Culture, Sports, Science and Technology, Japan through Special  ... 
doi:10.1145/1048935.1050164 dblp:conf/sc/Nakajima03 fatcat:4dgn2uzpqzgv7cv6ryws6loh6e

Parallel Preconditioning Methods with Selective Fill-Ins and Selective Overlapping for Ill-Conditioned Problems in Finite-Element Methods [chapter]

Kengo Nakajima
2007 Lecture Notes in Computer Science  
In the present paper, parallel preconditioning methods with selective fill-ins and selective overlapping for contact problems have been developed and evaluated on PC clusters with 64 cores.  ...  Finally, the proposed methods have been applied to ill-conditioned problems with heterogeneity and robustness, and their efficiency has been evaluated on TSUBAME super-cluster with up to 512 cores.  ...  The author would like to thank Professor Satoshi Matsuoka and members of the Global Scientific Information and Computing Center, Tokyo Institute of Technology, for use of TSUBAME super-cluster.  ... 
doi:10.1007/978-3-540-72588-6_173 fatcat:ud5vnojdunhgfjjaafcirzrs2q

Strategies for Preconditioning Methods of Parallel Iterative Solvers for Finite-Element Applications in Geophysics [chapter]

Kengo Nakajima
Lecture Notes in Earth Sciences  
In particular, for simulation codes based on the finite-element method (FEM), most of the computational time is devoted to solving linear equation systems with sparse coefficient matrices.  ...  In contrast, iterative methods are memory scalable. Iterative methods are therefore the only choice for large-scale simulations by massively parallel computers.  ...  Acknowledgements This work is supported by the 21st Century Earth Science COE Program at the University of Tokyo, and CREST/Japan Science and Technology Agency.  ... 
doi:10.1007/978-3-540-85879-9_3 fatcat:pn3hnf5lxvg5rmcth442l46syq

Efficient solution of sparse linear systems arising in engineering applications on vector hardware [article]

Sunil Reddy Tiyyagura, Universität Stuttgart, Universität Stuttgart
Block-based Linear Iterative Solver (BLIS) is a scalable software library for solving large sparse linear systems, especially arising from engineering applications.  ...  The two main performance problems with the public domain solvers are addressed in this work. The first problem of short vector length is solved by introducing a vector specific sparse storage format.  ...  GeoFEM project One successful project using the largest installation of NEC SX-6 (Earth Simulator project in Japan) was GeoFEM, which was active between 1997-2001, funded by The Science and Technology  ... 
doi:10.18419/opus-1878 fatcat:72ce3l7dojaphdzyecqsmdcopy