Sites Inria

Version française

ROMA Research team

Optimisation des ressources : modèles, algorithmes et ordonnancement

Team presentation

Resource Optimization: Models, Algorithms, and scheduling


The ROMA team aims at designing models, algorithms, and scheduling strategies to optimize the execution of scientific applications on High-Performance Computing platforms. More specifically, ROMA is interested in obtaining the “best” possible performance from the point of view of the user (e.g., application execution time) while using ressources as efficiently as possible (e.g., low energy consumption). The work performed by ROMA ranges from theoretical studies to the development of software used daily in the academic and industrial worlds.


Research themes

The ROMA team is devoted to the study of the fault tolerance, the energy consumption, and the memory usage of scientific computing applications executed on clusters and on supercomputers, with a special interest in direct solvers for sparse linear systems. The work of the Roma team will be organized along the three following research themes.
  1. Resilience. In this theme, we focus on the efficient
      execution of applications on failure-prone platforms. Here, we
      typically address questions such as: Given a platform and an
      application, which fault-tolerance protocols should be used, when,
      and with which parameters? Related to this problematic, is the
      optimization of the execution of applications whose behavior is
      described through some probability distributions: in both contexts,
      optimization problems must be solved in probabilistic settings.
  2. Multi-criteria scheduling strategies. In this theme, we focus on the design
      of scheduling strategies that finely take into account some platform
      characteristics beyond the most classical ones, namely the computing
      speed of processors and accelerators, and the communication
      bandwidth of network links. In the scope of this theme, when
      designing scheduling strategies, we focus either on the energy
      consumption of applications or on their memory behavior. All
      optimization problems under study are multi-criteria.
  3. Solvers for sparse linear algebra and related optimization problems. In this theme, we work on most aspects of
      direct multifrontal solvers for linear systems, usually in the scope
      of the MUMPS solver that we co-develop. We also work on
      combinatorial scientific computing, that is, on the design of
      combinatorial algorithms and tools to solve combinatorial problems,
      such as those encountered, for instance, in the preprocessing phases
      of solvers of sparse linear systems.  In addition, we also work on
      dense linear algebra: we focus on the adaptation of factorization
      kernels to emerging and future platforms.

International and industrial relations

International relations

  • Argonne National Laboratory, USA
  • Bilkent University, Turkey
  • Ohio State University, Columbus, USA
  • Rutherford Appleton Laboratory, Didcot, UK
  • University of Colorado, Denver, USA
  • University of Hawaii at M?noa, Honolulu, USA
  • University of Pittsburgh, USA
  • University of Tennessee, Knoxville, USA

Industrial relations

  • Bull
  • Altair, EDF, Michelin, etc. (consortium MUMPS

Keywords: Algorithms Scheduling Energy Memory Resilience Fault tolerance Multicriteria optimization Linear algebra Sparse linear systems Direct solvers Combinatorial Scientific Computing Hypergraph Partitioning