Pierre Ramet
- Université de Bordeaux (UB)
- Centre Inria de l'Université de Bordeaux
- Laboratoire Bordelais de Recherche en Informatique (LaBRI)
- Institut Universitaire de Technologie - Bordeaux (IUT Bordeaux)
- Outils et Optimisations pour le Calcul Haute Performance et l'Apprentissage (TOPAL)
Présentation
Ma problématique de recherche concerne les problèmes du calcul haute performance, en particulier l'algèbre linéaire creuse. Je développe la bibliothèque PaStiX (https://solverstack.gitlabpages.inria.fr/pastix) qui implémente un solveur direct creux haute performance.
Domaines de recherche
Compétences
Publications
Publications
Algebraic hierarchical partitioning to improve H-matrix compressionPP 2026 - SIAM Conference on Parallel Processing for Scientific Computing, SIAM, Mar 2026, Berlin, Germany |
|
|
|
Scalable and portable LU factorization with partial pivoting on top of runtime systemsIPDPS25 - 39th IEEE International Parallel and Distributed Processing Symposium, Jun 2025, Milan, Italy |
|
|
Batching the tasks of the LU factorization with partial pivoting on top of runtime systemsCOMPAS 2025 - Conférence francophone d'informatique en Parallélisme, Architecture et Système, Jun 2025, Bordeaux, France |
|
|
Slides 18th CMCIM 2024, Toward an algebraic multigrid method for the indefinite Helmholtz equation18th Copper Mountain Conference On Iterative Methods, Apr 2024, Frisco (Co), United States |
|
|
Enhancing sparse direct solver scalability through runtime system automatic data partitionWAMTA 2024 - Workshop on Asynchronous Many-Task Systems and Applications 2024, Feb 2024, Knoxville, United States. pp.105-110, ⟨10.1007/978-3-031-61763-8_10⟩ |
Comparative Study of Mixed-Precision and Low-Rank Compression Techniques in Sparse Direct Solvers2024 SIAM Conference on Applied Linear Algebra, Society for Industrial and Applied Mathematics, May 2024, Paris, France |
|
|
|
Student Paper Competition, Copper Mountain Conference 2023 - Toward a multilevel method for the Helmholtz equation21st SIAM Copper Mountain Conference on Multigrid Method, Apr 2023, Copper Mountain, CO, United States |
|
|
New parallel features in the sparse solver PaStiXSparse Days 2023, Jun 2023, Toulouse, France |
Slides, SIAM CSE 2023 - Toward a multilevel method for the Helmholtz EquationCSE 2023 - SIAM Conference on Computational Science and Engineering, Feb 2023, Amsterdam, Netherlands |
|
|
|
Deciding Non-Compressible Blocks in Sparse Direct Solvers using Incomplete FactorizationHiPC 2021 - 28th IEEE International Conference on High Performance Computing, Data, and Analytics, Dec 2021, Bangalore, India. pp.1-10, ⟨10.1109/HiPC53243.2021.00024⟩ |
|
|
Improving mapping for sparse direct solvers: A trade-off between data locality and load balancingEuroPar 2020 - 26th International European Conference on Parallel and Distributed Computing, Aug 2020, Warsaw / Virtual, Poland. pp.1-16 |
Recent Developments Around the Block Low-Rank PaStiX SolverPP 2020 - SIAM Conference on Parallel Processing for Scientific Computing, Feb 2020, Seattle, United States |
|
Rank Revealing QR Methods for Sparse Block Low Rank SolversSparse Days 2019, Jul 2019, Toulouse, France |
|
Block Low-rank Algebraic Clustering for Sparse Direct SolversSIAM Conference on Computational Science and Engineering (CSE19), Feb 2019, Spokane, United States |
|
Rank Revealing QR Methods for Sparse Block Low Rank SolversCOMPAS 2019 - Conférence d'informatique en Parallélisme, Architecture et Système, Jun 2019, Anglet, France |
|
|
|
Block Low Rank Algebraic Clustering for Sparse Direct SolversJournées Ondes Sud-Ouest (JOSO), Mar 2019, Le Barp, France |
Exploiting Parameterized Task-graph in Sparse Direct SolversSIAM Conference on Computational Science and Engineering (CSE19), Feb 2019, Spokane, United States |
|
|
|
Utilisation de la compression Block Low-Rank pour accélérer un solveur direct creux supernodalCOMPAS 2018 - Conférence d'informatique en Parallélisme, Architecture et Système, Jul 2018, Toulouse, France |
|
|
Supernodes ordering to enhance Block Low-Rank compression in sparse direct solversPMAA 2018 - 10th International Workshop on Parallel Matrix Algorithms and Applications, Jun 2018, Zurich, Switzerland |
Sparse Supernodal Solver Using Hierarchical Compression over Runtime SystemSIAM Conference on Computation Science and Engineering (CSE'17), Feb 2017, Atlanta, United States |
|
|
|
Sparse Supernodal Solver Using Block Low-Rank Compression18th IEEE International Workshop on Parallel and Distributed Scientific and Engineering Computing (PDSEC 2017), Jun 2017, Orlando, United States |
|
|
Sparse Supernodal Solver exploiting Low-Rankness PropertySparse Days 2017, Sep 2017, Toulouse, France |
Impact of Blocking Strategies for Sparse Direct Solvers on Top of Generic RuntimesSIAM Conference on Computation Science and Engineering (CSE'17), Feb 2017, Atlanta, United States |
|
Exploiting Modern Manycore Architecture in Sparse Direct Solver with Runtime SystemsSIAM Conference on Computation Science and Engineering (CSE'17), Feb 2017, Atlanta, United States |
|
Sparse Supernodal Solver Using Hierarchical CompressionWorkshop on Fast Direct Solvers, Nov 2016, Purdue, United States |
|
Impact of Blocking Strategies for Sparse Direct Solvers on Top of Generic RuntimesSIAM Conference on Parallel Processing for Scientific Computing (SIAM PP 2016), Apr 2016, Paris, France |
|
Exploiting H-Matrices in Sparse Direct SolversSIAM Conference on Parallel Processing for Scientific Computing (SIAM PP 2016), Apr 2016, Paris, France |
|
On the use of low rank approximations for sparse direct solversSIAM Annual Meeting (AN'16), Jul 2016, Boston, United States |
|
Exploiting Kepler architecture in sparse direct solver with runtime systems9th International Workshop on Parallel Matrix Algorithms and Applications (PMAA'2016), Jul 2016, Bordeaux, France |
|
Towards a recursive graph bipartitioning algorithm for well balanced domain decompositionSIAM Conference on Computational Science and Engineering (SIAM CSE 2015), Mar 2015, Salt Lake City, United States |
|
|
|
3D Cartesian Transport Sweep for Massively Parallel Architectures with PARSECIEEE International Parallel & Distributed Processing Symposium (IPDPS 2015), May 2015, Hyderabad, India. pp.581-590, ⟨10.1109/IPDPS.2015.75⟩ |
Blocking strategy optimizations for sparse direct linear solver on heterogeneous architecturesSparse Days, Jun 2015, Saint Girons, France |
|
Towards a recursive graph bipartitioning algorithm for well balanced domain decompositionMini-Symposium on "Combinatorial Issues in Sparse Matrix Computation" at ICIAM'15 conference, Aug 2015, Beijing, China |
|
On the use of H-Matrix Arithmetic in PaStiX: a Preliminary StudyWorkshop on Fast Direct Solvers, Jun 2015, Toulouse, France |
|
On the design of parallel linear solvers for large scale problemsMini-Symposium on Recent advances in matrix computations for extreme-scale computers at ICIAM'15 conference, Aug 2015, Beijing, China |
|
A task-based sparse direct solver suited for large scale hierarchical/heterogeneous architecturesSIAM Conference on Computational Science and Engineering (SIAM CSE 2015), Mar 2015, Salt Lake City, United States |
|
|
|
Nested dissection with balanced haloSixth SIAM Workshop on Combinatorial Scientific Computing, Jul 2014, Lyon, France |
|
|
Parallel 3D Sweep Kernel with PARSECHPCC Workshop on HPC-CFD in Energy/Transport Domains, Aug 2014, Paris, France |
|
|
Taking advantage of hybrid systems for sparse direct solvers via task-based runtimesHCW'2014 workshop of IPDPS, May 2014, Phoenix, United States. pp.29-38, ⟨10.1109/IPDPSW.2014.9⟩ |
Toward parallel scalable linear solvers suited for large scale hierarchical parallel platformsWorld Congress on Computational Mechanics, Jul 2014, Barcelona, Spain |
|
|
|
An improved recursive graph bipartitioning algorithm for well balanced domain decompositionIEEE International Conference on High Performance Computing (HiPC 2014), Dec 2014, Goa, India. pp.1-10, ⟨10.1109/HiPC.2014.7116878⟩ |
|
|
Shared Memory Parallelism for 3D Cartesian Discrete Ordinates SolverJoint International Conference on Supercomputing in Nuclear Applications and Monte Carlo 2013, Oct 2013, Paris, France |
|
|
From hybrid architectures to hybrid solversWorkshop Celebrating 40 Years of Nested Dissection, Jul 2013, Waterloo, Canada |
|
|
Towards Domain Decomposition with Balanced HaloWorkshop Celebrating 40 Years of Nested Dissection, Jul 2013, Waterloo, Canada |
Fine Grain Scheduling for Sparse Solver on Manycore ArchitecturesSIAM Conference on Parallel Processing for Scientific Computing (SIAM PP 2012), Feb 2012, Savannah, United States |
|
Sparse direct solver on top of large-scale multicore systems with GPU acceleratorsSIAM Conference on Applied Linear Algebra (SIAM LA 2012), Jun 2012, Valence, Spain |
|
Toward a supernodal sparse direct solver over DAG runtimesParallel Matrix Algorithms and Applications, Jun 2012, Londres, United Kingdom |
|
Memory Optimization to Build a Schur ComplementSIAM Conference on Applied Linear Algebra (SIAM LA 2012), Jun 2012, Valence, Spain |
|
Development of an International Matrix-Solver Prediction System on a French-Japanese International Grid Computing EnvironmentJoint International Conference on Supercomputing in Nuclear Applications and Monte Carlo 2010 (SNA+MC2010 2010), Oct 2010, Tokyo (JP), Japan |
|
A Non Overlapping Parallel Domain Decomposition Method Applied to The Simplified Transport EquationsInternational Conference on Mathematics, Computational Methods & Reactor Physics, 2009, New-York, United States |
|
Non-Linear MHD simulations of Edge Localised Modes36th EPS Plasma Physics Conference, Jun 2009, Sofia, Bulgaria |
|
A NUMA Aware Scheduler for a Parallel Sparse Direct SolverPMAA'08, 2008, Neuchâtel, Switzerland |
|
Toward an international sparse linear algebra expert system by interconnecting the ITBL computational Grid with the Grid-TLSE platformVECPAR 2008: 8th International Conference High Performance Computing for Computational Science, Jun 2008, Toulouse, France |
|
A domain decomposition method for the resolution of an eigenvalue problem in neutron physicsInternational Symposium on Iterative Methods in Scientific Computing (IMACS), 2008, Lille, France |
|
A Domain Decomposition Method Applied to Large Eigenvalue Problems in Neutron PhysicsPMAA'08, 2008, Neuchatel, Switzerland |
|
|
|
Dynamic Scheduling for sparse direct Solver on NUMA architecturesPARA'08, May 2008, Trondheim, Norway |
A domain decomposition method applied to the simplified transport equationsIEEE 11th International Conference on Computational Science and Engineering, 2008, Sao Paulo, Brazil. pp.91--97 |
|
Tunable parallel experiments in a GridRPC framework: application to linear solversVECPAR'08, 2008, Toulouse, France. pp.46--52 |
|
|
|
A NUMA Aware Scheduler for a Parallel Sparse Direct SolverWorkshop on Massively Multiprocessor and Multicore Computers, INRIA, Feb 2009, Rocquencourt, France. 5p |
Improvement of existing solvers for the simulation of MHD instabilitiesNumerical flow models for controlled fusion, 2007, Porquerolles, France |
|
Efficient solution technique for low Mach number compressible multiphase problemsPMAA'06, 2006, Rennes, France |
|
On finding approximate supernodes for an efficient ILU(k) factorizationPMAA'06, 2006, Rennes, France |
|
Partitioning and Blocking Issues for a Parallel Incomplete FactorizationPARA'06, 2006, Umea, Sweden. pp.929--937 |
|
On using an hybrid MPI-Thread programming for the implementation of a parallel sparse direct solver on a network of SMP nodesSixth International Conference on Parallel Processing and Applied Mathematics, Workshop HPC Linear Algebra, 2005, Poznan, Poland. pp.1050--1057 |
|
Blocking Issues for an Efficient Parallel Block ILU PreconditionerInternational SIAM Conference On Preconditioning Techniques For Large Sparse Matrix Problems In Scientific And Industrial Applications, 2005, Atlanta, United States |
|
Applying parallel direct solver skills to build robust and highly performant preconditionersPARA'04, 2004, Copenhagen, Denmark. pp.601--619 |
|
A Blockwise Algorithm for Parallel Incomplete Cholesky FactorizationPMAA'04, 2004, Marseille, France |
|
Using of the High Performance Sparse Solver PaStiX for the Complex Multiscale 3D Simulations performed by the FluidBox Fluid Mechanics SoftwarePMAA'04, 2004, Marseille, France |
|
High Performance Complete and Incomplete Factorizations for Very Large Sparse Systems by using Scotch and PaStiX softwaresEleventh SIAM Conference on Parallel Processing for Scientific Computing, 2004, San Francisco, United States |
|
Efficient algorithms for direct resolution of large sparse system on clusters of SMP nodesSIAM Conference on Applied Linear Algebra, 2003, Williamsburg, United States |
|
Towards High Performance Hybrid Direct-Iterative Solvers for Large Sparse SystemsInternational SIAM Conference On Preconditioning Techniques For Large Sparse Matrix Problems In Scientific And Industrial Applications, 2003, Napa Valley, United States |
|
Asymptotically optimal algorithm for Laplace task graphs on heterogeneous platformsFifth International Conference on Parallel Processing and Applied Mathematics, Workshop HeteroPar, 2003, Czestochowa, Poland. pp.880--887 |
|
Parallel factorization of very large sparse SPD systems on a network of SMP nodesPMAA'02, 2002, Neuchatel, Switzerland |
|
Optimisation de l'occupation mémoire pour un solveur parallèle creux direct hautes performances de type supernodalRenPar'2002, 2002, Hamamet, Tunisie |
|
Modélisation de pipelines hétérogènesGRID'2002, 2002, Aussois, France |
|
PaStiX: Un solveur parallèle direct pour des matrices creuses symétriques définies positives basé sur un ordonnancement statique performant et sur une gestion mémoire efficaceRenPar'2001, 2001, Paris, France |
|
PaStiX: A Parallel Direct Solver for Sparse SPD Matrices based on Efficient Static Scheduling and Memory ManagmentTenth SIAM Conference on Parallel Processing for Scientific Computing, 2001, Portsmouth, United States |
|
Algèbre Linéaire Creuse Hautes Performances : Application à la Mécanique des StructuresiHPerf'2000, 2000, Aussois, France |
|
PaStiX: A Parallel Sparse Direct Solver Based on a Static Scheduling for Mixed 1D/2D Block DistributionsIrregular'2000 workshop of IPDPS, 2000, Cancun, Mexico. pp.519-525 |
|
Description of the EMILIO Software Processing Chain and Application to Structural MechanicsPMAA'2K, 2000, Neuchatel, Switzerland |
|
PaStiX: A High-Performance Parallel Direct Solver for Sparse Symmetric Definite SystemsPMAA'2K, 2000, Neuchatel, Switzerland |
|
A Mapping and Scheduling Algorithm for Parallel Sparse Fan-In Numerical FactorizationEuroPar'99, 1999, Toulouse, France. pp.1059-1067 |
|
Optimisation des Communications et Régulation de Charge pour la Résolution par Méthode Directe de Grands Systèmes Linéaires CreuxICaRE'97, 1997, Aussois, France. pp.467-488 |
|
Calcul de la suite optimale de taille de paquets pour la factorisation de CholeskyRenPar'9, 1997, Lausanne, Suisse. pp.111-114 |
|
Optimal Grain Size Computation for Pipelined AlgorithmsEuro-Par'96, 1996, Lyon, France. pp.165-172 |
|
Calcul de la taille optimale des paquets pour les algorithmes macro-pipelinesRenPar'8, 1996, Bordeaux, France. pp.21-24 |
|
|
Partitionnement hiérarchique pour la simulation électromagnétique d'objets 3D complexesCOMPAS 2025 - Conférence francophone d'informatique en Parallélisme, Architecture et Système, Jun 2025, Bordeaux, France. |
|
|
Scalable and Portable LU Factorization with Partial Pivoting on top of Runtime SystemsJCAD 2024 - journée calculs et données, Nov 2024, Bordeaux, France |
|
|
Sparse supernodal solver using block low-rank compression: Design, performance and analysisJOREK development meeting, Nov 2019, Cadarache, France |
|
|
Utilisation de la compression low-rank pour réduire la complexité du solveur PaStiXJCAD 2018 - Journées Calcul et Données, Oct 2018, Lyon, France |
|
|
Sparse supernodal solver with low-rank compression for solving the frequency-domain Maxwell equations discretized by a high order HDG methodJournées jeunes chercheur-e-s - Résolution de problèmes d’ondes harmoniques de grande taille, Nov 2017, PARIS, France. pp.1-55, 2017 |
Deflation and augmentation techniques in Krylov linear solversTopping, B.H.V and Ivanyi, P. Developments in Parallel - Distributed - Grid and Cloud Computing for Engineering, Saxe-Coburg Publications, pp.249--275, 2013, 978-1-874672-62-3 |
|
|
Toward an algebraic multigrid method for the indefinite Helmholtz equation2024 |
|
|
Reaching the Quality of SVD for Low-Rank Compression Through QR Variants[Research Report] RR-9476, Inria Bordeaux - Sud Ouest. 2022, pp.43 |
|
|
Deciding Non-Compressible Blocks in Sparse Direct Solvers using Incomplete Factorization[Research Report] RR-9396, Inria Bordeaux - Sud Ouest. 2021, pp.16 |
|
|
Improving mapping for sparse direct solvers: A trade-off between data locality and load balancing[Research Report] RR-9328, Inria Rhône-Alpes. 2020, pp.21 |
|
|
A hierarchical fast direct solver for distributed memory machines with manycore nodes[Research Report] CEA/DAM; Total E&P; Université de Bordeaux. 2019 |
|
|
Supernodes ordering to enhance Block Low-Rank compression in sparse direct solvers[Research Report] RR-9238, Inria Bordeaux Sud-Ouest. 2018, pp.1-31 |
|
|
Sparse Supernodal Solver Using Block Low-Rank Compression[Research Report] RR-9022, Inria Bordeaux Sud-Ouest. 2017, pp.24 |
|
|
Efficient Parallel Solution of the 3D Stationary Boltzmann Transport Equation for Diffusive Problems[Research Report] RR-9116, Inria; EDF Lab. 2017, pp.22 |
|
|
Sparse Supernodal Solver Using Block Low-Rank Compression: design, performance and analysis[Research Report] RR-9130, Inria Bordeaux Sud-Ouest. 2017, pp.1-32 |
|
|
Reordering strategy for blocking optimization in sparse linear solvers[Research Report] RR-8860, Inria Bordeaux Sud-Ouest; LaBRI - Laboratoire Bordelais de Recherche en Informatique; Bordeaux INP; Université de Bordeaux. 2016, pp.26 |
|
|
Simul'Elec and PASTIX interface specifications[Rapport Technique] RT-0458, INRIA Bordeaux; AlgoTech; INRIA. 2015 |
|
|
Amélioration du comportement numérique des solveurs en prenant en compte les poids de la matrice lors de la décomposition de domaines[Rapport Technique] RT-0445, INRIA. 2014, pp.58 |
|
|
Taking advantage of hybrid systems for sparse direct solvers via task-based runtimes[Research Report] RR-8446, INRIA. 2014, pp.25 |
|
|
Adaptive mesh refinement for numerical simulation of MHD instabilities in tokamaks: JOREK code[Research Report] RR-8635, INRIA Bordeaux; INRIA. 2014, pp.18 |
|
|
An improved recursive graph bipartitioning algorithm for well balanced domain decomposition[Research Report] RR-8582, INRIA. 2014, pp.29 |
|
|
Deflation and augmentation techniques in Krylov linear solvers[Research Report] RR-8265, INRIA. 2013, 25 p |
|
|
Non regression testing for the JOREK code[Research Report] RR-8134, INRIA. 2012, pp.17 |
|
|
Sparse direct solvers with accelerators over DAG runtimes[Research Report] RR-7972, INRIA. 2012, pp.11 |
|
|
Memory Optimization to Build a Schur Complement in an Hybrid Solver[Research Report] RR-7971, INRIA. 2012, pp.11 |
|
|
A NUMA Aware Scheduler for a Parallel Sparse Direct Solver[Research Report] RR-7498, INRIA. 2010, pp.22 |
SCILAB to SCILAB// - The Ouragan Project[Research Report] RR-2001-24, LIP - ENS Lyon. 2001 |
|
|
|
SCILAB to SCILAB// - The Ouragan Project[Research Report] RR-4203, INRIA. 2001 |
|
|
Heterogeneous architectures, Hybrid methods, Hierarchical matrices for Sparse Linear SolversDistributed, Parallel, and Cluster Computing [cs.DC]. Université de Bordeaux, 2017 |