Brigitte Rozoy: 'At uni, I discovered with delight that maths were more than an assimilation - they were a thought process'

For the first annual Maths Week, March 12-18, the theme was "Women and Maths". It offered an opportunity to interview three researchers and team leaders at Saclay, with three different profiles and careers, but who all chose to discuss with us their choice to work in science and research. This week we met with Brigitte Rozoy, head of the Grand-Large team.

I was a good student in high school and my parents would never have understood if I had chosen to do anything other than science! I was rather gifted for maths. I found it amusing, but not very interesting. I preferred biology or physics. But since I didn't have good marks in biology, and I was not good at all in physics-chemistry lab, I chose maths, a bit as a fall-back solution. It was also complicated because there was this strong family imaginary, an uncle who was a brilliant mathematician and who had died young, and so it was difficult to summon the confidence to do maths with this legacy. One of my brothers, a year younger than me, chose to pursue maths too, but also in a roundabout way.

After high school, much to my surprise, maths became exciting, which I wasn't expecting at all! I found myself dealing with real problems, for example continuity or the existence of real irrational numbers. We weren't dealing with computing or control, but really the concepts and thought process. So I did a Bachelor, a Master and then a teaching degree. This would allow me to finance my studies if I went on to teach ten years after that. Then I was hired as a research assistant and I soon joined a research institute dedicated to teaching mathematics (IREM) where I served as director. At the time, with the introduction of the "new maths" in secondary schools, there was a real sociological problem in teacher training. I spent almost ten years working on these questions with teachers in secondary schools and higher education. It was a very rewarding experience

Computing as a science
is the intellectual adventure of our century

In the 80s I made a new discovery through logic, computing, and I really developed a passion for the field. I was very happy with this change of direction because for me computing as a science is really the intellectual adventure of our century. As Gilles Dowek once said, mankind has experienced several giant steps, several revolutions: speech, writing, printing and computing. These four upheavals have had an equal impact on the evolution of mankind and have fundamentally and irreversibly modified its future.

As a research professor in Avignon and then Caen, I experienced the rise of computing and the significant development of teaching and research in the field at universities. There were plenty of things to discover, plenty of room and enthusiasm. I worked hard and learned a lot. I found my research topic and did a doctoral thesis. Suddenly I found myself 10 years behind everyone else, but I also had many curious and passionate colleagues. I got lucky and landed a job as a professor at Paris Sud.

I know I took a winding, unconventional path, but I am very pleased with the choices I made, that I explored various new and intriguing fields, on both a personal and scientific level. This only reinforced my attachment to teaching, research and universities.
Moreover, I was lucky enough to have some very positive opportunities. In fact, I was highly involved in research administration, in universities at every level, but also within the Ministry of Research and Higher Education. This gave me another perspective on how universities and research work, a global, strategic vision on a national and international level.

Now that I am close to retiring, I am currently in charge of an Inria team on an interim basis. In fact Franck Cappello created the Grand-Large team in which I worked before he headed off to the United States to set up the Inria-Illinois Petascale Computing Joint Lab at Urbana-Champaign. As the project that will soon take over, I hope, for Grand-Large is not fully mature, I accepted to oversee the transition, until Laura Grigori's project is up and running. For me, my role now consists in passing the torch. I try to help the new generation benefit from opportunities that are as rich and inspiring as those I experienced.

Research theme: large distributed systems and high-performance computing

I first became interested in the theory of formal languages, then in the mathematical aspects of parallelism and distribution. Computers had been sequential for a long time, that is to say they performed one function at a time. But now that we know how to organise them in a network they can work "parallel" to each other. To do this we need to control traffic and communications, to avoid blocking the system. In the 90s there was an important shift because behind all this there was the advent of the Internet and communication on a global scale! Then in the years 2000 very, very large networks appeared and the Grand-Large team was set up to study these questions in depth. The idea is to create systems for large numbers of machines working together, like, for example, the Grid'5000 project, which aims to harness 5000 machines across France. This raises complex issues, in particular regarding resistance to failures and scientific computing. Finally, since 2010 we have been working with Laura Grigori, Joffroy Beauquier and the entire team on the fields of High Parallel Computing (HPC) and large distributed systems.

Keywords: Brigitte Rozoy HPC Saclay - Île-de-France Testimony Grand-Large research team