High performance computing (HPC) is based on advanced fundamental research in both applied mathematics and computer science. Motivated by both performance and power objectives, today's computing infrastructures evolve toward architectures with an increasing complexity, including sophisticated memory organizations. This is the case even for the mass market with the advent of multicores and GPGPUs, which has exposed the need for parallel programming and related code optimizations to a larger public. Because of this, supercomputing today requires a more global approach, from the design of numerical methods to extensive hardware considerations, in interaction with languages and compilers, to take into account both the complexity of architectures and the needs of their non-expert users.

Research communities in computer science (architecture, compilation) and applied mathematcs (numerical simulation) are not always aware of this need; at least their work do not always spread enough across the other discipline to lead to mutual influence. Automatic code optimizations and tools also require a better evaluation of their applicability. The goal of this research school – or meeting place of two communities – is to make the link between some of the most recent advances in computer science (program optimizations, in particular polyhedral techniques and tools) and applied mathematics (schemes for numerical simulation), in relation with application needs.

This school is therefore interdisciplinary, with a strong will to bring communities together on the common theme of supercomputing.

It is supported by the GDR (research group) Calcul, which is positioned at the interface of these two communities. It is also part of a thematic program on HPC from Labex Milyon proposed by LIP (Laboratoire de l'Informatique de Parallélisme, ENS Lyon) and ICJ (Institut Camille Jordan, Lyon).

This school can be seen as complementary to (but without requiring them as prerequisites) the first school on polyhedral code analysis and optimizations (http://labexcompilation.ens-lyon.fr/polyhedral-school) and the three days of keynotes on parallel languages for HPC (http://labexcompilation.ens-lyon.fr/hpc-languages), two events organized in 2013, also in the context of Labex Milyon.

• Current and future architectures
• Discretization, in particular discontinuous Galerkin
• Code analysis and optimizations, polyhedral techniques
• Cost models, performance analysis
• Optimizations for GPU, tiling, automatic tools
• Languages, programming paradigms, runtimes
• Numerical precision

The courses will be given in English.

• Matthieu Haefele (Maison de la Simulation)
• Markus Püschel (ETH Zürich)
• Alain Darte (CNRS, ENS Lyon)
• Philippe Helluy (Univ. Strasbourg)
• Cindy Rubio-Gonzalez (Univ. California, Davis)
• Uday Bondhugula (Indian Institute of Science, Bangalore)
• Sven Verdoolaege (Polly Labs)
• Ramanujam (Louisiana State Univ.)
• Thierry Dumont (Univ. Lyon)

Scientific committee :

• Alain Darte, DR CNRS, LIP, ENS Lyon
• Philippe Helluy, Pr, IRMA, Strasbourg
• François Irigoin, Mines ParisTech
• Violaine Louvet, IR CNRS, ICJ, Lyon

Organizing committee :

• Institut Camille Jordan : Thierry Dumont, Violaine Louvet
• Laboratoire de l'Informatique du Parallélisme : Alain Darte, Tomofumi Yuki
• Département de Mécanique de l'Université Lyon 1 :  Bastien Di Pierro
• MSSMat, Ecole Centrale Paris : Sophie Mouronval

This conference is part of the thematic program: HPC from the "laboratoire d'excellence" MILYON, an initiative from the French Ministry of Research. 

and supported by the Groupe Calcul iniitative

the CNRS

and école normale supérieure de Lyon