About Me
I am a Ph.D. student at Inria Bordeaux in the STORM team and at CEA since October 2013.
My advisors are Raymond Namyst, David Goudin, Samuel Thibault and Olivier Aumage.
Research Overview
The ever-increasing supercomputer architectural complexity emphasizes the need for high-level parallel programming paradigms to design efficient, scalable and portable scientific applications. Among such paradigms, the task-based programming model abstracts away much of the architecture complexity by representing an application as a Directed Acyclic Graph (DAG) of tasks. Among them, the Sequential-Task-Flow (STF) model decouples the task submission step, which becomes sequential, from the parallel task execution step. While this model allows for further optimizations on the DAG of tasks at submission time, there is a key concern about the performance hindrance of sequential task submission when scaling. My thesis' work focuses on studying the scalability of the STF-based StarPU runtime system (developed at Inria Bordeaux in the STORM team) on dense linear algebra solvers for large scale industrial simulations. To that end, we collaborate with the HiePACS team of Inria Bordeaux on the Chameleon software, which is a collection of linear algebra solvers on top of task-based runtime systems.
Main Research Interests
- High performance scientific computing
- Parallel and distributed computing at large scale
- Task-based programming and runtime systems
Communications
Talk
Overview of Distributed Linear Algebra on Hybrid Nodes over the StarPU Runtime
Emmanuel Agullo, Olivier Aumage, Mathieu Faverge, Nathalie Furmento, Florent Pruvost, Marc Sergent, Samuel Thibault
SIAM Conference on Parallel Processing (SIAM-PP) 2014, Portland, USA
2014/02/19
Talk
Harnessing Clusters of Hybrid Nodes with a Sequential Task-Based Programming Model
Emmanuel Agullo, Olivier Aumage, Mathieu Faverge, Nathalie Furmento, Florent Pruvost, Marc Sergent, Samuel Thibault
Parallel Matrix Algorithms and Applications (PMAA) Workshop, Lugano, Switzerland
2014/07/03
Poster
Matrices Over Runtime Systems @ Exascale
Emmanuel Agullo et al.
SuperComputing 2015, Austin, USA
2015/11/17
Poster
Scalability of Distributed Task-Based Runtime Systems for Dense Linear Algebra Applications
Marc Sergent
EDMI 2016 Open Day, Talence, France
2016/03/11
Paper
Controlling the Memory Subscription of Distributed Applications with a Task-Based Runtime System
Marc Sergent, David Goudin, Samuel Thibault, Olivier Aumage
High-Level Parallel Programming Models and Supportive Environments (HIPS) Workshop, Chicago, USA
2016/05/23
Research Report
Achieving High Performance on Supercomputers with a Sequential Task-based Programming Model
Emmanuel Agullo, Olivier Aumage, Mathieu Faverge, Nathalie Furmento, Florent Pruvost, Marc Sergent, Samuel Thibault
2016/06/16
Travels
- Argonne National Laboratory (ANL)
- Lemont, Illinois, USA
- 31st of May - 1st of June, 2016
- Centre Européen de Recherche et de Formation Avancée en Calcul Scientifique (CERFACS)
- 4th of February, 2016
- Toulouse, France
- Innovative Computing Laboratory (ICL)
- University of Tenessee, Knoxville (UTK), USA
- 24-28th of February, 2014