TERATEC : le calcul haute performance

INRIA

Inria, the French national institute for computer science and applied mathematics, promotes scientific excellence for technology transfer and society. With graduates from the world’s top universities, Inria's 2,400 employees rise to the challenges of digital sciences. The institute’s open, agile model allows Inria to explore original approaches, in particular via joint laboratories (Inria Innovation Lab and Inria Project Lab) with its partners in industry and academia and provide an efficient response to the multidisciplinary and application challenges of the digital transformation.

A French research institute of international stature in the field of HPC

As a member of the European Technology Platform for High Performance Computing (ETP4HPC), Inria participates, through the presence of its researchers in various working groups, in the definition of the platform's Strategic Research Agenda (SRA).

Inria's project teams thus contribute to the European high-performance computing research ecosystem and maintain numerous collaborations in Europe, in particular through their participation in over ten Horizon 2020 collaborative projects in the HPC field. Inria is also a partner, through shareholding and collaboration with GENCI, of the PRACE (Partnership for Advanced Computing in Europe) program whose mission is to further high-impact scientific research and R&D development to improve European competitiveness for the benefit of society.

At the international level, Inria and the University of Urbana Champaign (United States) set up, in June 2009, a joint intensive computing research laboratory (JLPC) involving researchers from Inria, Illinois Center for Extreme-Scale Computation, and the National Center for Supercomputing Applications. In 2013, this laboratory took on other partners (the Argonne National Laboratory, the Barcelona Supercomputing Center, the Jülich Supercomputing Center, the Riken Advanced Institute for Computational Science and the University of Tennessee as an associate partner) to become the JLESC (Joint Laboratory for Extreme Scale Computing). Based at the University of Illinois, the JLESC focuses mainly on the software challenges encountered.

Inria also maintains long-standing HPC-related collaborations with its Brazilian partners, initially as part of a CNPq-Inria "HOSCAR" project on numerical simulation and massive data processing based on new high-performance computing architectures for environmental science applications (2012-2015), then as part of an EU-Brazil "HPC4E: HPC for Energy" project (2016-2017).

Industrial partnerships and technology transfer to businesses

Committed to assisting innovators, Inria provides the ideal conditions for fruitful relations between public research, private R&D and industry. Inria transfers its expertise and research results to startups, SMEs and major groups in fields as diverse as healthcare, transport, energy, communications, security and privacy protection, smart cities and industry of the future. Inria has also fostered an entrepreneurial culture that has led to the creation of 150 startups.

INRIA and TERATEC

High performance simulation has always been a recurrent topic at Inria since the institute was created in 1967. In recent years, Inria has had numerous collaborations with industrial partners in this field and has seen the birth of companies resulting from its research on high-performance simulation.

In the strategic plan for 2018-2022, the institute stresses the challenge of jointly leveraging the existing know-how and solutions from the HPC and Big Data domains to build next generation tools. Performing transversal research is expected to foster the emergence of original solutions to meet the upcoming needs in both domains. Implementing this transversal research approach, Inria has created the “HPC-Big Data” Inria Project Lab (2018-2022) bringing together several machine learning, Big Data and HPC teams from Inria as well as external partners such as Argonne National Lab, LBT/CNRS, ATOS/Bull, Esi-Group. The goal is to address issues related to the convergence between HPC, Big Data and machine learning. The three main research focuses are high performance analytics for scientific computing applications, high performance analytics for big data applications and infrastructure and resource management.

It was thus very natural for Inria to join TERATEC. The institute is keen to develop partnerships, transfer the results of its research teams working in this strategic field and generally bring all its HPC and numerical simulation expertise to TERATEC members and partners.