Death of Janine Connes
This page was translated by an AI (LLM) with a cursory human check and is awaiting full review.
Janine Connes, founder of the CNRS Inter-Regional Electronic Computing Centre (CIRCÉ) in 1969 and its director until 1982, passed away on 28 November 2024 at the age of 98. Together with her husband Pierre Connes, she formed a high-flying scientific duo, with Pierre as the experimenter and Janine processing and analysing the results numerically. This pair established the Fourier Transform Infrared Spectroscopy method, which is of major importance in this field. The page on the University of California Los Angeles website dedicated to Mrs Connes attests to this. She also played a crucial role in the development of numerical simulation in France, enabling the CNRS to acquire the necessary resources to serve all scientific communities. She provides a detailed account of this saga in her book From the IBM 360/75 to the Jean Zay Supercomputer, written with the participation of Françoise Perriquet and published in 20221. We will briefly recall only the premises and the main stages here.
Janine Connes completed her thesis under the supervision of Pierre Jacquinot in 1954. She worked on determining the composition of Venus's atmosphere and attracted the attention of Frank Gray, then director of NASA, who was recruiting the research team for the first Viking mission to explore Mars. Mrs Connes thus gained extensive experience in computations during extended stays in the United States from 1963 onwards, notably at NASA's Jet Propulsion Laboratory (JPL). It was there that she made significant advances in Fourier Transform Infrared Spectroscopy, always hand in hand with Pierre Connes2. She was also able to discover the operation of the Stanford University computing centre and the activity of other major American centres, notably those at Argonne and the Lawrence Livermore National Laboratory. With all this experience, she became a specialist in numerical simulation at a time when this speciality was still in its infancy.
Pierre Jacquinot, himself a physicist and former director of the Aimé Cotton Laboratory then located in Meudon-Bellevue, was faced with the difficulty of having the necessary computing power for his own research. Having become Director General of the CNRS between 1962 and 1969, he imposed the idea of creating a computing centre dedicated to all researchers needing very high computing power. He wanted a generalist centre, detached from computer science research: "To give users of all disciplines the possibility to perform the calculations they need for their research and which they cannot do elsewhere."
In 1968, he turned to Janine Connes to design, set up and direct this new structure, which was then unique in Europe. Drawing on her experience across the Atlantic, she was able to propose to Pierre Jacquinot to build the new structure in question, drawing heavily on the successful American models.
Not without causing some turbulence, this led in 1969 to the creation of the CNRS Inter-Regional Electronic Computing Centre (CIRCÉ) on the Moulon plateau in Orsay, a national computing centre. Some of the staff came from the Institut Blaise Pascal and the rest were specially recruited. Its mission as a service unit, available to all scientific disciplines but detached from computer science research itself, made the centre a unique and essential place for all French scientific research.
Mrs Connes oversaw the development and direction of CIRCÉ for 13 years, until 1982, when she left to head the newly created CNRS Scientific Computing Directorate, created by the then Director General, Pierre Papon.
Later, in 1992, the incumbent Director General, François Kourilsky, launched the MIPS (High-Performance Computing Resources for Science) project, aimed at preparing the reorganisation of the CNRS computing centres, CIRCÉ and the CNRS share in the Centre de calcul vectoriel pour la recherche (CCVR) economic interest grouping, an organisation that played a primordial role in access to and promotion of vector computing in research between 1984 and 1993. The project resulted in the creation of the new CNRS high-performance computing centre at the end of 1993, the Institut du développement et des ressources en informatique scientifique (IDRIS), in the premises of the former CIRCÉ.
IDRIS, in turn, has seen constant and spectacular development. Since 2007, together with its partners at the CEA's Very Large Computing Centre (TGCC), the National University Computing Centre (CINES) and under the aegis of the GENCI (National High-Performance Computing Facility) civil society, it has been part of the national infrastructure serving the entire research world.
At the time of her passing, it is fitting to celebrate the pioneering, major and driving role that Janine Connes played in the development of scientific computing in France.