GUILLAUME MERCIER
Associate Professor at Bordeaux INP / ENSEIRB since 2006.
Argonne National Laboratory Alumnus (March 2005, August 2006).
CONTACT:
TADaaM INRIA (Satanas Team)
INRIA Bordeaux - Sud-Ouest Research Center
LaBRI (Laboratoire Bordelais de Recherche en Informatique, Unité Mixte de Recherche CNRS 5800)
Adresse postale :
INRIA Bordeaux - Sud Ouest
200, Avenue de la Vieille Tour
F-33405 Talence cedex
Tél: +33 (0)5 24 57 41 00 (Inria, office 425)
Tél: +33 (0)5 56 84 23 61 (ENSEIRB, office S331)
ENSEIGNEMENTS
Programmation Impérative -- Introduction au langage C (PG109, 1A, S5)
- Le support de cours est distribué lors de la première séance.
Cet incunable est néanmoins disponible sur mon compte ENSEIRB (/net/ens/mercier/PG019).
-
SHA1 sum de la version courante : 3fba1fa471b8468c3d397148c5ef96389c4e3675
- Des transparents sont disponibles sur le site de R. Giraud :
ici
- Tous les exemples sont également disponibles sur mon compte (/net/ens/mercier/PG019/Exemples).
- Les sujets de TP, ainsi que leurs corrections, y seront disponibles en temps utile.
- Une plate-forme avec des exercices et des auto-corrections sont mis à votre disposition
sur Thor.
Introduction à la Programmation Système (IF210 2A, S7)
- Le support de cours est distribué lors de la première séance.
Cet incunable est néanmoins disponible sur mon compte ENSEIRB (/net/ens/mercier/IF210).
-
SHA1 sum de la version courante : b7c39e8e636a0aff908c16fc77d06b14bfe9fbca
- Les sujets de TP y seront disponibles en temps utile.
Projet Réseau et Système (PR204, 2A, S7)
- Le sujet ainsi que les codes de départ seront disponibles en cours de semestre sur
Thor.
RESEARCH
Research Interests
I am working in Emmanuel Jeannot's
TADaaM INRIA Team.
My research interests include (but are not limited to):
-
Hardware topology management and support in parallel applications and
software (e.g. the Hsplit project).
Follow the discussion on Github: HWT-WG
-
Efficient process placement on multicore and NUMA hierarchical machines,
for instance with the TopoMatch
and hwloc projects.
-
Hybrid (eg MPI + OpenMP/threads) programming models for HPC:
Habilitation Thesis
I defended my habilitation thesis on December the 4th, 2019. It is available here
Publications
-
MPI collective communication through a single set of interfaces: A case for orthogonality,
Daniel J. Holmes, Sascha Hunold, Guillaume Mercier and Jesper Larsson Träff
In in Parallel Computing, volume 107, p 102826, 2021
Details.
-
Application-Driven Requirements for Node Resource Management in Next-Generation Systems,
Edgar León, Balazs Gerofi, Brice Goglin, Julien Jaeger, Guillaume Mercier, Rolf Riesen and Masamichi Takagi
In International Workshop on Runtime and Operating Systems for Supercomputers (ROSS 2020), Nov 2020, Atlanta, GA, United States
Details.
-
Collectives and Communicators: A Case for Orthogonality (Or: How to get rid of MPI neighbor and enhance Cartesian
collectives), Daniel J. Holmes, Sascha Hunold, Guillaume Mercier and Jesper Larsson Träff
In EuroMPI/USA 2020, Sept. 2020, Austin, USA, TX Details.
-
Exposition, Clarification, and Expansion of MPI Semantic Terms and Conventions: Is a nonblocking function permitted to block?,
Purushotham V. Bangalore, Claudia Blaas-Schenner, Daniel J. Holmes, Julien Jaeger, Guillaume Mercier, Rolf Rabenseifner and Anthony Skjellum
In EuroMPI 2019, Sept. 2019, Zürich, Switzerland Details.
-
Hardware topology management in MPI applications through hierarchical communicators.
Brice Goglin, Emmanuel Jeannot, Farouk Mansouri and Guillaume Mercier
In in Parallel Computing, volume 76, pp 70--90, 2018
Details.
-
Topology-Aware Job Mapping.
Yiannis Georgiou, Emmanuel Jeannot, Guillaume Mercier and Adèle Villiermet
In International Journal of high-Performance Computing applications, 2017
Details.
-
Large-scale experiment for topology-aware resource management.
Yiannis Georgiou, Guillaume Mercier and Adèle Villiermet
In Coloc Workshop, in conjunction with EuroPAR, Aug. 2017, Santiago de Compostella, Spain
Details.
-
A Hierarchical Model to Manage Hardware Topology in MPI Applications.
Emmanuel Jeannot, Farouk Mansouri and Guillaume Mercier
In EuroMPI/USA 2017, Sept. 2017, Chicago IL, USA
Details.
-
Online Dynamic Monitoring of MPI Communications.
George Bosilca, Clément Foyer, Emmanuel Jeannot, Guillaume Mercier and Guillaume Papauré
In EuroPAR 2017, Aug. 2017, Santiago de Compostella, Spain
Details.
-
Topology-aware resource management for HPC applications.
Yiannis Georgiou, Emmanuel Jeannot, Guillaume Mercier and Adèle Villiermet
In ICDCN 2017, Jan 2017, Hyderabad, India
Details.
-
Topology and Affinity Aware Hierarchical and Distributed Load-Balancing in Charm++.
Emmanuel Jeannot, Guillaume Mercier and François Tessier
In COMHPC@SC 2016, Nov 2016, Salt Lake City, United States
Details.
-
Matching communication pattern with underlying hardware architecture.
Emmanuel Jeannot, Guillaume Mercier and François Tessier
In 6th European Conference on Computational Fluid Dynamics, Jul 2014, Barcelona, Spain
Details.
-
Process Placement in Multicore Clusters: Algorithmic Issues and Practical Techniques.
Emmanuel Jeannot, Guillaume Mercier and François Tessier
In IEEE Transactions on Parallel and Distributed Systems, 2014, Volume 25, number 4, pp 993-1002,
Details.
-
Communication and Topology-aware Load Balancing in Charm++ with TreeMatch.
Emmanuel Jeannot, Esteban Meneses, Guillaume Mercier, François Tessier, Gengbin Zheng
In IEEE Cluster 2013, Sep 2013, Indianapolis, United States
Details.
-
Improving MPI Applications Performance on Multicore Clusters with Rank Reordering.
Guillaume Mercier and Emmanuel Jeannot.
In Proceedings of the 16th International EuroMPI Conference, Santorini, Greece, September 2011, LNCS 6960, pp 39--49.
Details.
-
Near-Optimal Placement of MPI processes on Hierarchical NUMA Architectures.
Emmanuel Jeannot and Guillaume Mercier.
In Proceedings of the 16th International Euro-Par Conference, Ischia, Italy, September 2010, LNCS 6272, Volume 2 pp 199--210.
Details.
-
hwloc: a Generic Framework for Managing Hardware Affinities in HPC Applications.
Francois Broquedis, Jerome Clet-Ortega, Stephanie Moreaud, Nathalie Furmento,
Brice Goglin, Guillaume Mercier, Samuel Thibault and Raymond Namyst
In Proceedings of the The 18th Euromicro International Conference on Parallel, Distributed and Network-Based Computing (PDP 2010),
Pisa, Italy, February 2010. pp 180--186.
Details
-
Cache-Efficient, Intranode Large-Message MPI Communication with MPICH2-Nemesis.
Darius Buntinas, Brice Goglin, Dave Goodell, Guillaume Mercier, and Stephanie Moreaud.
In Proceedings of the 38th International Conference on Parallel Processing (ICPP-2009),
Vienna, Austria, September 2009.
IEEE Computer Society Press.
Details.
-
Towards an Efficient Process Placement Policy for MPI Applications in Multicore Environments.
Guillaume Mercier and Jerome Clet-Ortega,
in Recent Advances in Parallel Virtual Machine and Message Passing Interface , Proceedings of the 16th
European PVM/MPI users' group Meeting, Volume 5749 of Lecture Notes in Computer Science, pages 104-155,
Espoo, Finland, September 2009. Springer.
Details.
-
NewMadeleine: An Efficient Support for High-Performance Networks in MPICH2.
Guillaume Mercier, Francois Trahay, Darius Buntinas, and Elisabeth Brunet,
in Proceedings of 23rd IEEE International Parallel and Distributed Processing Symposium (IPDPS'09)
Rome, Italy, May 2009. IEEE Computer Society Press.
Details.
-
Implementation and Evaluation of Shared-Memory Communication and Synchronization Operations in MPICH2 using
the Nemesis Communication Subsystem . Darius Buntinas, Guillaume Mercier, and William Gropp.
In Parallel Computing, Selected Papers from EuroPVM/MPI 2006, 33(9):634-644, September 2007
Details.
-
High-Performance Multi-Rail Support with the NewMadeleine Communication Library.
Olivier Aumage, Elisabeth Brunet, Guillaume Mercier, and Raymond Namyst.
in HCW 2007: the Sixteenth International Heterogeneity in Computing Workshop, held in conjunction with IPDPS 2007
Long Beach, California, USA, March 2007. IEEE Computer Society Press
Details.
-
Implementation and Shared-Memory Evaluation of MPICH2 over the Nemesis Communication Subsystem.
Darius Buntinas, Guillaume Mercier, and William Gropp.
In Recent Advances in Parallel Virtual Machine and Message Passing Interface: Proc. of 13th European PVM/MPI
Users' Group Meeting,
Bonn, Germany, September 2006
Details.
-
Data Transfer in a SMP System: Study and Application to MPI .
Darius Buntinas, Guillaume Mercier, and William Gropp.
In Proceedings of the 34th International Conference on Parallel Processing(ICPP 2006),
Colombus, Ohio, August 2006.
Details.
-
Design and Evaluation of Nemesis: a Scalable, Low-Latency, Message-Passing Communication Subsystem.
Darius Buntinas, Guillaume Mercier and William Gropp.
In Proceedings of 6th IEEE/ACM International Symposium on Cluster Computing and the Grid (CCGrid 2006),
Singapore, May 2006. Held in conjunction with IEEE Computer Society and ACM.
Details.
-
MPICH/MadIII: a Cluster of Clusters Enabled MPI Implementation.
Guillaume Mercier and Olivier Aumage.
In Proc. 3rd IEEE/ACM International Symposium on Cluster Computing and the Grid (CCGrid 2003),
Tokyo, May 2003.
Details.
-
A Portable and Efficient Communication Library for High-Performance Cluster Computing.
Olivier Aumage, Luc Bouge,Alexandre Denis, Lionel Eyraud, Jean-Francois Mehaut, Guillaume Mercier, Raymond Namyst, and Loic Prylli.
In Cluster Computing: Selected Papers from the IEEE Cluster 2000 Conference, 5(1):43-54, January 2002
Details.
-
MPICH-Madeleine: a True Multi-Protocol MPI for High-Performance Networks.
Guillaume Mercier, Olivier Aumage and Raymond Namyst.
In Proc. 15th International Parallel and Distributed Processing Symposium (IPDPS 2001)
San Francisco, April 2001
Details.
-
A Portable and Efficient Communication Library for High-Performance Cluster Computing.
Olivier Aumage, Luc Bouge,Alexandre Denis, Jean-Francois Mehaut, Guillaume Mercier, Raymond Namyst, and Loic Prylli.
In IEEE Intl Conf. on Cluster Computing (Cluster 2000)
Chemnitz, Germany, November 2000.
Details.