Required Education : Phd
Start date : 1 January 2020
Mission duration : CDI
Cerfacs is a research centre specialized in numerical modelling and simulation. Through its resources and expertise in high-performance computing, Cerfacs addresses the major scientific and technical problems of public and industrial research with the support of its partners (Airbus, Cnes, EDF, Météo-France, Onera, Safran, Total).
The Strategic Research Plan established with our partners to cover the period 2018 – 2022 highlighted 5 transversal strategic research axes: Linear algebra, Numerical methods for solving partial differential equations, Coupling, Exascale, Data oriented modelling.
Cerfacs wishes to strengthen its research capacity and is opening a new permanent research position for this purpose.
The proposed profile consists in the development of tools and methods concerning two main axes:
– The Data Science (DS) axis is growing rapidly at CERFACS, it is of interest to all our partners and requires strengthening the potential of the staff who run it. The fields of application are very broad and range from traditional statistical analyses in this field to the introduction of DS-type methods in large calculation codes.
– The HPC (High Performance Computing) theme is a permanent thread at CERFACS. Continuing to develop CERFACS’ expertise in this area is one of the actions identified in the strategic plan. This is particularly necessary for the development of the previous axis: developing the use of Data Science tools must be done in an HPC context, which links the two activities.
Title: Senior Researcher
Type of position: Full-time position with an open-ended contract (CDI)
Responsible: Reports to the team leader of COOP (Scientific Software Operational Performance)
Location: Toulouse (South of France)
Required level: Doctor of Physics or Doctor in Computer Science/Applied Mathematics.
Start date: Position to be filled. Last quarter 2019 – first quarter 2020.
– At least two years of experience after obtaining a thesis in HPC/HPDA and the development of scientific codes.
– References required.
Missions of the position.
– Development of new tools and methods in data science as part of the strategic axis “Data-driven modelling”
– Development of expertise to make the best use of the next generations of HPC and HPDA architectures within the framework of the “Exascale” strategic axis
– Contribution to the improvement of the models developed at Cerfacs with a view to their optimization on future architectures, participation in border simulations performed on international super-calculators
– Preparation of responses to national or European calls for tenders in coordination with Cerfacs’ partners
– Scientific project management
– Publication of papers in international peer-reviewed journals, oral presentations and posters at international symposia,
– Contributions to Cerfacs’ training activities, to Cerfacs’ seminar cycles, and to the animation of the strategic research plan.
– Proven development skills: Fortran, C++
– Proven HPC development skills: MPI, OpenMP, OpenACC or CUDA
– Competence in performance optimization, use of profilers and performance analysis tools in an HPC environment
– Knowledge of parallel HPC architectures, accelerators (GPUs) and processors (x86, Power…)
– Experience in source code management systems: Git, SVN…
– First experiences in Artificial Intelligence (Machine learning, deep learning) applied to the world of digital simulation is a plus
– Experience in handling code coupled to HPC architectures is a plus
– Ability to communicate easily (writing articles, oral presentation, teaching, transmission of know-how),
– Ability to set up and manage research projects,
– Ability to work in a team, reactivity, autonomy and rigour
– Ability to supervise (trainees, thesis students, young researchers)
– Fluency in English language, written and spoken, fluent in French
Contact for application:
– CV, cover letter and references to be sent to Nicolas Monnier (Head of IT and COOP activity at Cerfacs): Nicolas.Monnier@cerfacs.fr