Thursday, September 7

9:00 – 9:40       Invited lecture I : Erik Boman

                         Combinatorial Algorithms for Parallel Sparse Matrix Distributions (presentation)

 

9:40 – 10:30     Preconditioners I

·       On finding approximate supernodes for an efficient ILU(k) factorization  (presentation)

      Pascal Hénon, Pierre Ramet, Jean Roman

·       Numerical experiments with additive Schwarz preconditioner for non­overlapping domain decomposition in 3D  (presentation)

      Luc Giraud, Azzam Haidar, Shane Mulligan

·       Support Graph Preconditioners for the Finite Element Method [CANCELLED]

      Vivek Sarin, Meiqiu Wang

 

10:30 – 11:00         Coffee break

11:00 – 12:30   Direct Methods for Dense and Sparse Matrices

·       Modeling the LU Factorization for SMP Clusters  (presentation)

      Jack Dongarra, Emmanuel Jeannot, Julien Langou

·       A Preliminary Analysis of the Out­of­Core Solution Phase of a parallel Multifrontal Approach  (presentation)

      Patrick Amestoy, Iain Duff, Abdou Guermouche, Tzvetomila Slavova

·       Efficiently solving large sparse linear systems on a distributed and heterogeneous grid by using the multisplitting­direct method (presentation)

      Sylvain Contassot-Vivier, Raphaël Couturier, Christophe Denis, Fabienne Jézéquel

·       Schur complements and direct solvers applied on a simple domain: trend and behavior of a monolevel strategy  (presentation)

      Jean­Paul Boufflet, Emmanuel Lefrançois, Michel Vayssade

 

12:40 – 14:15         Lunch

14:15 – 14:55   Invited lecture II : Gérard Meurant

                         Algebraic Multilevel Preconditioners on Massively Parallel Computers (presentation)

 

14:55 – 16:10   Preconditioners II

·       A new implementation of the CMRH method for solving dense linear systems (presentation)

Mohammed Heyouni, Hassan Sadok

·       A Parallel Implementation with a Comparative Study of GMRES(m) Preconditioned by Multiplicative Schwarz Methods (presentation)

      Emmanuel Kamgnia, Guy Antoine Atenkeng Kahou, Bernard Philippe, Masha Sosonkina

·       Convergence of iterative solvers for the method coupling Finite elements and integral representation  (presentation)

      Faker Ben Belgacem, Nabil Gmati, Faten Jelassi, Bernard Philippe

 

 

16:10 – 16:30         Coffee break

16:30 – 18:10   Parallel sessions

 

I. Grid and eigenvalue computations

Minisymposium proposed by Serge Petiton

·       Solving symmetric eigenproblems on grid and global computing environment (presentation)

         Laurent Choy, Serge Petiton, Mitsuhisa Sato

·       Dynamic load balancing of the power method on a Grid (presentation)

      Clovis Dongmo Jiogo, Pierre Manneback

·       Efficient Parallel Implementation of Classical Gram-Schmidt Orthogonalization Using Matrix Multiplication (presentation)

      Takuya Yokozawa, Daisuke Takahashi, Taisuke Boku, Mitsuhisa Sato

·       Resolution of Large Scale Hermitian Eigenproblem using YML Workflow Framework (presentation)

      Nahid Emad, Tetsuya Sakurai

 

***********************

         II. Graph Partitioning

·       A Partitioning Algorithm for Block­Diagonal Matrices with Overlap  (presentation)

      Guy Antoine Atenekeng Kahou, Laura Grigori, Masha Sosonkina

·       PT­Scotch: A tool for efficient parallel graph ordering (presentation)

      Cédric Chevalier, Francois Pellegrini

·       Using hypergraph partitioning for iterative linear system solution techniques (presentation)

      Masha Sosonkina, Yousef Saad, Bora Uçar

·       Parallel algorithms for downdating the least squares estimator of the regression mode (presentation)

      Petko Ivanov Yanev, Erricos John Kontoghiorghes

 

18:10                       End of the sessions

18:30                       Reception at the Town Hall of Rennes

 

Friday, September 8

8:00 – 8:40       Invited lecture III : Erik Elmroth

                         Using recursion to improve performance of dense linear algebra software (presentation)

 

8:40 – 9:55       Parallel algorithms I

·       Parallel synchronous and asynchronous iterative algorithms for Procrustes­type problems (presentation)

Giorgos Kollias, Efstratios Gallopoulos, Daniel Szyld

·       Parallel Quasi­Monte Carlo Algorithms for Matrix Computations (presentation)

Aneta Karaivanova

·       Algorithm RaPTI for solving evolution problems in a time-parallel way (presentation)

Nabil Nassif, Noha Makhoul Karam, Yeran Soukiassian.

 

9:55 – 10:25           Coffee break

10:25 – 12:05   Parallel sessions

 

I. Parallel Matrix Algorithms in Computational Science

Minisymposium proposed by Ahmed Sameh

·       Solving large scale eigenvalue problems in electronic structure calculations (presentation)

      Yousef Saad

·       Computational Nanoelectronics (presentation)

      Eric Polizzi

·       Molecular Dynamics Simulations of Reactive Systems (presentation)

      Sagar Pandit, Metin Aktugla, Ananth Grama

·       A Parallel System Solver for the Navier-Stokes Equations (presentation)

      Ahmed Sameh, Murat Manguoglu, Faisal Saied

 

***********************

II. Second Session

SVD based methods

·       Balanced Truncation Model Reduction of Large and Sparse Descriptor Linear Systems (presentation)

      José Manuel Badía, Peter Benner, Rafael Mayo, Enrique S. Quintana­Ortí, Gregorio Quintana-Ortí, Alfredo Remón

·       New Class of Block Matrix Orderings for the Parallel Two­Sided Jacobi SVD Algorithm (presentation)

      Gabriel Okša, Ondrej Sýkora, Marián Vajteršic

 

Parallel environments

·       The Standards and Data Structures of Dense Linear Algebra Software (presentation)

      Fred Gustavson, Jerzy Waśniewski

·       An unified linear equation solvers interface for industrial softwares (presentation)

Bernard Sécher, Michel Belliard

 

 

12:05 – 13:45         Lunch

13:45 – 14:25   Invited lecture IV : Peter Arbenz

                         Multi­level mu­Finite Element Analysis for Human Bone Structures (presentation)

 

14:25 – 15:40   Parallel Algorithms II

·       Combining Building Blocks for Parallel Multilevel Matrix Multiplication  (presentation)

Sascha Hunold, Thomas Rauber, Gudula Rünger

·       Matrix Multiplication on Three Heterogeneous Processors  (presentation)

Brett Becker, Alexey Lastovetsky

·       Parallel processing of matrix operations on the Pulsyr configurable architecture  (presentation)

Andriy Lutsyk, Bohdan Rusyn, Oleksiy Lutsyk

 

 

15:40 – 16:00         Coffee break

16:00 – 16:50   Applications I

·       Efficient technique for solving low Mach number compressible multiphase problems  (presentation)

Benjamin Braconnier, Boniface Nkonga, Mikaël Papin, Pierre Ramet, Mario Ricchiuto, Jean Roman, Rémi Abgrall

·       Parallel Multistage Stochastic Programming Model Used in Portofolio Management (presentation)

Maria Lucka, Igor Melichercik, Ladislav Halada.

 

16:50                       End of the sessions

17:00                       Departure for the Mont Saint Michel excursion and conference dinner

 

Saturday, September 9

8:45 – 9:25       Invited lecture V : Jari Toivanen

                         A Parallel Partial Solution Variant of Cyclic Reduction and Applications (presentation)

 

9:25 – 10:40     Domain Decomposition

                         Minisymposium proposed by Damien Tromeur-Dervout

·       A new cement to glue nonconforming grids with Robin interface conditions in the finite element case (presentation)

      Caroline Japhet, Yvon Maday, Frédéric Nataf

·       Aitken-Schwarz DDM to solve Darcy flow in heterogeneous underground mediums (presentation)

      Jocelyne Erhel, Andrea Frullone, Damien Tromeur-Dervout, Jean-Raynal de Dreuzy

·       Domain Decomposition methods for Stiff ODEs/DAEs (presentation)

      David Guibert, Damien Tromeur-Dervout

 

 

10:40 – 11:10         Coffee break

11:10 – 12:00   Applications II

·       ab initio Electronic Structure Calculations on the BG/L Supercomputer (presentation)

      Constantinos Bekas, Alessandro Curioni, Wanda Andreoni

·       Multilevel domain decomposition for electronic structure calculations (presentation)

      Maxime Barrault, Guy Bencteux, Eric Cancès, William Hager, Claude Le Bris

 

 

12:00                       End of the Conference

12:10                       Lunch