Expert in scientific computing and AI

The structure you will join

ENTROPIE is a UMR with 5 supervisory bodies (IRD, CNRS, IFREMER and 2 universities, Réunion and N-Calédonie) whose objective is to better understand the functioning of marine and island ecosystems in the tropical Indo-Pacific in the context of global warming. in order to propose strategies for conservation, enhancement and sustainable management. You will join a multidisciplinary team based at the IRD of New Caledonia (Nouméa) made up of ecologists, marine geneticists, specialists in remote sensing and marine geosciences and climate.

An attractive mission

Within the UMR ENTROPIE, you will be in charge of the administration of the HPC system, the modernization of its software ecosystem and the support of researchers in their intensive computing projects. You will organize yourself to push for the development of research activities in the fields of Big-Data and AI. The tasks are varied:

  • Participate in the establishment of a modern software ecosystem, in synergy with the University, including the implementation of software solutions facilitating the appropriation of HPC by Researchers, PhD students and young engineers in training,
  • Maintain in operational condition, ensure old technology on hardware and software, and anticipate the evolution of the backbone, infrastructures and operational requirements in order to effectively support IRD’s research programs,
  • Develop the software stack and propose other solutions (ie containerization, etc.) for the reproducibility of scientific applications through the servers of the IRD, the University, the Cloud and the National Centers,
  • Mobilize mathematical and computer methods to solve the problems encountered in Big-Data and Deep-Learning,
  • Participate in multidisciplinary research projects by mobilizing your know-how.

Your future team

Within the UMR ENTROPIE, you will find support from a second IT at the origin of the development of HPC in Noumea and you will act in concert with an HPC specialist from the University of New Caledonia, very invested the promotion of supercomputing within the local scope of academic research. Researchers and doctoral students already specialized in AI or experienced in the use of massively parallelized codes for climate and ocean sciences will be your main contacts.

The profile we are looking for

A Doctorate or Bac +5 level in computer science / engineering sciences or equivalent is required for this position.

With a solid background in Linux administration, familiar with DELL/HP servers and comfortable porting Fortran, C and Python codes, you have developed the following technical skills:

  • Mastery of diagnostic, monitoring, profiling and job/resource management techniques (SLURM, etc.),
  • Knowledge of the main HPC concepts and technologies in this field (InfiniBand, MPI/SMP, BeeGFS/Lustre, CPU/GPU hybrid calculations, etc.),
  • Knowledge of parallelization paradigms and techniques, performance optimization and underlying concepts (acceleration, efficiency, scalability limit, etc.),
  • Experience in deploying AI-specific software layers (TensorFlow, PyTorch, Caffe),
  • You are familiar with the use of environments, methodologies and algorithms for AI and/or large-scale data processing.
  • On the other hand, you have demonstrated great autonomy and know-how in project management methods, combined with a very good ability to listen to the needs expressed by users.
  • You have a natural taste for organizing training, workshops and know how to provoke and seize new opportunities to co-construct structuring projects.

The IRD, an Institute that gives meaning to your career

Your mission at the service of committed science for a sustainable future: The IRD in 230 seconds

Similar Posts

Leave a Reply

Your email address will not be published. Required fields are marked *