5th International Workshop on

Parallel Matrix Algorithms and Applications (PMAA'08)

20-22 June 2008, Neuchâtel, Switzerland



Plenary talk 1 Friday, 20.06.2008 11:20 - 12:15 Room: GGA
A parallel GMRES method preconditioned by a Multiplicative Schwarz iteration
Speaker: Bernard Philippe Chair: Yousef Saad
Plenary talk 2 Saturday, 21.06.2008 17:20 - 18:15 Room: GGA
Exploiting nonnegativity in matrix and tensor factorizations to improve topic detection and tracking in text mining
Speaker: Michael W. Berry Chair: Ahmed Sameh


Parallel session G Friday, 20.06.2008 14:00 - 16:00

Session PS07 Room: GGA
Robust multilevel methods and parallel algorithms - 1 Chair: Johannes Kraus
  #25:  U. Yang, R. Falgout, J. Brannick
  Compatible relaxation in parallel algebraic multigrid
  #19:  C. Flaig, P. Arbenz, C. Bekas, H. van Lenthe, U. Mennel, R. Muller, M. Sala
  A scalable multi-level preconditioner for matrix-free micro-finite element analysis of human bone structures
  #27:  E. Karer, J. Kraus
  On the strength of nodal dependence in amg for vector-field problems
  #24:  H. Yang, W. Zulehner
  An algebraic multigrid (AMG) solver for a finite element (FEM) discretization of the Stokes/Navier-Stokes system on hybrid meshes and its parallelization
Session PS11 Room: GPA
Krylov space methods and applications Chair: Mario Arioli
  #47:  P. Jiranek, M. Rozloznik, M. Gutknecht
  On the numerical behavior of Simpler GMRES and GCR
  #65:  M. Arioli, I. Duff
  Backward stability of FGMRES
  #64:  J. Hogg, J. Scott
  On the use of mixed precision for the fast and robust solution of sparse symmetric linear systems.
  #68:  S. Djungu, P. Manneback
  Block algorithms for computing PageRank by sites
Parallel session H Friday, 20.06.2008 16:25 - 18:30

Session PS03 Room: GGA
Parallel preconditioners Chair: Ahmed Sameh
  #17:  A. Grama, M. Manguoglu, M. Koyuturk, A. Sameh
  Parallel banded preconditioners for non-symmetric linear system solvers
  #10:  A. Basermann, J. Schmidt
  Block incomplete LU preconditioning on modern hardware platforms
  #9:  A. Haidar, L. Giraud, S. Pralet
  Algebraic preconditioners for parallel hybrid solvers
  #22:  A. Martin-Huertas, Jose I. Aliaga, E. Quintana-Orti, Matthias Bollhoefer
  Parallel multilevel ILU preconditioners
  #13:  A. Sameh, M. Naumov
  A parallel hybrid banded solver and its generalization to sparse linear systems
Session PS04 Room: GPA
Parallel combinatorial scientific computing Chair: Costas Bekas
  #80:  J. Her, F. Pellegrini
  Efficient and scalable parallel graph partitioning
  #44:  R. Bisseling, T. van Leeuwen, U. Catalyurek
  A hybrid two-dimensional method for sparse matrix partitioning
  #42:  M. Sosonkina, Y. Saad
  Hypergraph partitioning for sparse linear systems: a case study with discontinuous PDEs
  #14:  C. Bekas, A. Curioni, P. Arbenz
  Very large scale graph partitioning problems in micro finite element analyses of human bone structure
Session PS09 Room: GB1
Large-scale sparse matrix computations Chair: Pasqua D'Ambra and Daniela di Serafino
  #28:  B. Ucar,C. Aykanat, U.V. Catalyurek
  A matrix partitioning interface to patoh in matlab
  #30:  S. Filippone, P. D'Ambra, D. di Serafino
  MLD2P4: A package of scalable algebraic multilevel Schwarz preconditioners
  #32:  A. Buttari, P. Amestoy, J. L'excellent
  Towards a parallel analysis phase for a multifrontal sparse solver.
  #37:  L. Grigori, J. Demmel, H. Xiang
  Communication avoiding Gaussian elimination
  #58:  R. Vuduc,S. Williams, L. Oliker, J. Shalf, J. Demmel, K. Yelick
  Programming models and techniques for sparse matrix kernels on multicore platforms
Parallel session I Saturday, 21.06.2008 09:00 - 11:00

Session PS13 Room: GGA
Robust multilevel methods and parallel algorithms - 2 Chair: Svetozar Margenov
  #23:  J. Kraus, I. Georgiev, S. Margenov
 Hierarchical multilevel splittings for discontinuous Galerkin approximations of elliptic problems with high-frequency-high-contrast coefficients
  #26:  S. Tomar, J. Kraus
  Multilevel preconditioning in H(div) and applications to a posteriori error estimates
  #35:  R. Blaheta, P. Byczanski, R. Kohut, J. Stary
  Parallel Schwarz type solvers for THM modelling
  #36:  J. Stary, R. Blaheta, R. Kohut, A. Kolcun, S. Margenov
  Micro FEM analysis of geocomposites
  #21:  Y. Vutov, R. Blaheta, S. Margenov
 Parallel PCG algorithms for numerical homogenization of voxel structures
Session PS14 Room: GPA
SVD and Jacobi methods Chair: Gabriel Oksa
  #16:  N. Bosner, J. Barlow, Z. Drmac
  Parallel versions of one-sided bidiagonalization
  #57:  S. Singer, S. Singer, V. Hari, K. Bokulic, D. Davidovic, M. Juresic, A. Uscumlic
  Parallel implementations of the one--sided indefinite block Jacobi methods
  #52:  V. Hari, V. Zadelj-Martic
  Convergence to diagonal form of general Jacobi-type processes
  #48:  G. Oksa, M. Becka, L. Grigori, M. Vajtersic
  Optimal data distribution in the preconditioned parallel two-sided block Jacobi SVD algorithm
Parallel session J Saturday, 21.06.2008 11:20 - 13:20

Session PS01 Room: GGA
Algebraic preconditioning of iterative methods Chair: Miroslav Tuma
  #8:  O. Schenk
  Multilevel preconditioning for large-scale nonconvex PDE-constrained optimization
  #6:  M. Bollhofer
  Recent advances in preconditioing large-scale symmetric indefinite systems
  #4:  J. Mayer
  Symmetric permutations for I-matrices to avoid small pivots during incomplete factorization
  #7:  J. Duintjer Tebbens, M. Tuma
  On the usage of triangular preconditioner updates in matrix-free environment.
  #3:  T. Huckle
  Frobenius norm minimization and probing
Session PS06 Room: GPA
Parallel dense numerical linear algebra Chair: Peter Arbenz
  #18:  B. Parlett
  The envelope method
  #59:  L. Karlsson, B. Kagstrom
  Dynamic node-scheduling of a multishift QR sweep algorithm
  #81:  R. Alfredo, E. Quintana-Orti, G. Quintana-Orti, R. van de Geijn
  Parallel factorization of band matrices
  #60:  A. Buluc, C. Budak, J. Gilbert
 Gaussian elimination based algorithms on the GPU
  #83:  W. Gansterer, H. Schabauer, A. Gruber
 Tridiagonalizing complex symmetric matrices
Parallel session K Saturday, 21.06.2008 15:00 - 17:00

Session PS10 Room: GGA
Parallel eigensolvers and applications Chair: Jose E. Roman and Paulo Vasconcelos
  #45:  C. Campos, R. Ralha, V. Hernandez, D. Guerrero
  Towards a parallel code without communication for the eigenvalues of symmetric tridiagonals
  #34:  O. Marques
  Experiences in the computation of interior eigenvalues for electronic structure calculations
  #73:  R. Dusseaux, K. Ait Braham, N. Emad
  Eigenvalue system for the scattering from rough surfaces saving in computation time by a physical approach
  #33:  J. Roman, M. Kammerer, F. Merz, F. Jenko
  Fast eigenvalue calculations in a massively parallel plasma turbulence code
  #82:  P. Vasconcelos, F. Almeida, J. Roman
  A parallel code for computing eigenvalues of integral operators
Session PS15 Room: GPA
Parallelizing iterative methods Chair: Matthias Bollhoefer
  #55:  M. Bolten
  Highly scalable multigrid method for circulant and Toeplitz matrices using non-Galerkin coarse grid operators
  #40:  M. Emans
  AMG for equation systems in commercial fluid dynamics software
  #41:  J. Mas, J. Cerdan, J. Marin
  Symmetric low rank updates of ISM based preconditioners
  #54:  E. Varnik, U. Naumann
  Exploiting constant Jacobian entries in seed matrix construction
Parallel session M Sunday, 22.06.2008 08:40 - 10:40

Session PS17 Room: B103
Linear systems - 1 Chair: Denis Trystram
  #71:  M. Faverge, X. Lacoste, P. Ramet
  A NUMA aware scheduler for a parallel sparse direct solver
  #63:  H. Hasegawa, H. Kotakemori
  A mixed precision iterative method with fast quadruple precision arithmetic operation tuned by SSE2
  #56:  A. Hokpunna, M. Manhart
  Interface-splitting algorithm for parallelisation of diagonal dominant tridiagonal systems with multiple right-hand sides.
  #78:  A. Goldman, Y. Ngoko, D. Trystram
  Combining numerical iterative solvers
Session PS18 Room: B104
Automatic parallelization tuning and GPU programming Chair: Costas Bekas
  #46:  F. de Mesmay, F. Franchetti, Y. Voronenko, M. Pueschel
  Automatic generation of multithreaded vectorized adaptive libraries for matrix multiplication
  #72:  J. Elble, P. Vouzis, N. Sahinidis
  GPU supercomputing case study: the Kaczmarz algorithm
  #79:  W. Nasri, S. Achour, L. Steffenel
  Integrating performance models and adaptive approaches for efficient parallel algorithms
  #39:  K. Naono, T. Sakurai, M. Egi
  Research trends on automatic tuning methods for matrix computations and proposal of a new run-time automatic tuning method
Session PS19 Room: B013
HPC applications in physics Chair: Nahid Emad
  #62:  O. Certik, J. Vackar, M. Tuma
  Solving many-body Schrodinger equation using density functional theory and finite elements
  #61:  F. Dupros, F. De Martin, E. Foerster, D. Komatitsch, J. Roman
  High-performance finite-element simulations of seismic wave propagation in 3D inelastic media
  #67:  B. Lathuiliere, M. Barrault, P. Ramet, J. Roman
  A domain decomposition method applied to large eigenvalue problems in neutron physics
  #5:  T. Sheorey
  Implicit domain decomposition algorithm for multiphase flow through random porous formation
  #38:  N. Missirlis
  Towards optimum diffusion for load balancing in heterogeneous torus
Parallel session N Sunday, 22.06.2008 11:00 - 13:00

Session PS02 Room: B103
Nonnegative matrix factorization algorithms and applications Chair: Michael Berry
  #77:  E. Gallopoulos, V. Kalofolias, G. Kollias, D. Zeimpekis, C. Boutsidis
 Step up 2 NNDSVD: better initial approximations for NMF
  #76:  A. Pascual Montano
  Using the non-smooth NMF model in bioinformatics
  #12:  E. Tjioe, M. Berry, R. Homayouni, K. Heinrich
  Using a literature-based NMF model for discovering gene functional relationships
Session PS05 Room: B013
Parallelizing matrix computations Chair: Thomas Huckle
  #15:  M. Schnupp, R. Jacob
  A lower bound for sparse-matrix dense-vector multiplication using packet communication
  #69:  M. Bader
  Parallel dense and sparse matrix multiplication in a Quantum control problem
  #70:  K. Waldherr
  Chebyshev methods for the matrix exponential
  #29:  O. Fortmeier
  Parallel matrix computations arising in two-phase flow problems
Session PS16 Room: B104
Linear systems - 2 Chair: Pascal Henon
  #49:  X. Pinel, H. Calandra, I. Duff, S. Gratton, X. Vasseur
  Solution of three-dimensional Helmholtz equation in the frequency domain, using Krylov methods preconditioned by multigrid
  #51:  S. Fujino, K. Somehara
  A proposal of preconditioning based on AGM (arithmetic-geometric mean) of eigenvalues and its estimation
  #53:  T. Collignon, M. van Gijzen
  Efficient iterative solution of large sparse linear systems on a cluster of geographically separated clusters
  #66:  P. Henon, J. Gaidamour
  HIPS: a parallel hybrid direct/iterative solver based on a Schur complement approach.

Created by Computing & Statistics 2007