Sites Inria

Version française

HIEPACS Research team

High-End Parallel Algorithms for Challenging Numerical Simulations

Team presentation

An important force which has continued to drive HPC has been to focus on frontier milestones which consist in technical goals that symbolize the next stage of progress in the field. After the step of the teraflop machine in the 1990s, the HPC community envision the use of generalist petaflop supercomputers and soon coming exaflop machines in the 2020s. For application codes to sustain petaflop and more using a few millions of cores or more will be needed, regardless of processor technology. Currently, a few HPC simulation codes easily scale to this regime and major code development efforts are critical to achieve the potential of these new systems. Scaling to at this rate will involve improving physical models, mathematical modelling, super scalable algorithms that will require paying particular attention to acquisition, management and vizualization of huge amounts of scientific data.

In this context, the purpose of the HiePACS project is to perform efficiently frontier simulations arising from challenging research and industrial multiscale applications. The solution of these challenging problems require a multidisciplinary approach involving applied mathematics, computational and computer sciences. In applied mathematics, it essentially involves advanced numerical schemes. In computational science, it involves massively parallel computing and the design of highly scalable algorithms and codes to be executed on future petaflop (and beyond) platforms. Through this approach, HiePACS intends to contribute to all steps that go from the design of new high-performance more scalable, robust and more accurate numerical schemes to the optimized implementations of the associated algorithms and codes on very high performance supercomputers.

Research themes

In order to tackle these large scale scientific challenges the HiePACS activities are organized as a continuum of research based on a multidisciplinary approach involving applied mathematics, computational and computer sciences. The research topics of HiePACS are the following.

  • Methodological and algorithmic research
    • High performance computing on next generation architectures
    • High performance solvers for linear algebra problems
      • Hybrid direct/iterative solvers based on algebraic decomposition domain
      • Hybrid solvers based on a combination of multigrid methods and of direct solvers
      • Linear Krylov solvers
      • Eigensolvers
    • High performance Fast Multipole Method for N-body problems
    • Algorithmics for code coupling in complex simulations
  • Frontier simulations arising from challenging research and industrial multiscale applications
    • Material Physics
    • Application customers of high performance linear algebra solvers

International and industrial relations

  • International collaborations
    • Lawrence Berkeley National Laboratory
    • Antwerp University
    • University of Colorado at Denver
    • Stanford University
    • University of Tennessee at Knoxville
    • KAUST
  • Industrial collaborations
    • Airbus
    • CEA (CESTA, Ile-de-France, SACLAY)
    • EDF
    • TOTAL

Keywords: High-End Computing for Frontier Simulations Algorithmics Linear Algebra Fast Multipole Method Code Coupling Materials Physics