Julie Grollier, ERC laureate from the CNRS, presents her bioSPINspired project
The bioSPINspired project
The brain as a source of inspiration
The goal of the bioSPINspired project is to imitate the brain's neurons with tiny magnetic oscillators. Actually, we can see the brain as a huge orchestra that produces its own rhythms. The neurons are hundreds of thousands of musicians interacting with one another. All these musician neurons vibrate differently depending on the messages they get from the others, thus contributing to the complex final melody. that's why neurons are often depicted as oscillators. The magnetic oscillators that are studied here are nanometric, they vibrate very quickly and can be produced millions at a time on silicon. Julie Grollier's project aims at getting them to interact with one another through the electric signals and the magnetic fields they produce. bioSPINspired will study how those networks of oscillators can create microwave melodies. Also, the project gets its inspiration from the brain to show it is possible to process information and compute it by modulating the rhythms generated by these magnetic oscillators that emulate faster the neurons' tune.
Innovation and interdisciplinarity at the the heart of the bioSPINspired project
The idea to use the magnetic oscillators’ vibrations to produce computation inspired from the brain is totally new. Usually, those isolated oscillators are rather used as radio transmitters. The project is interdisciplinary, between physics, electronics and neurosciences.
An ambitious project with numerous expected applications
The brain undertakes extraordinary tasks, as for instance the recognition of a partially hidden person in a crowd. It completes this computation almost instantly and with a very low consumption of energy. A basic computer would consume a thousand times this energy and would be much slower. This work will lay the basis to produce electronic chips able to recognize patterns in flows of data in the real world, to order them almost instantly and with a very low consumption of energy. Applications may be found in robotics, autonomous vehicles and medicine.
Interview of Julie Grollier, coordinator of the bioSPINspired project
Could you present your research background as well as your research activity?
I have a double degree in engineering and physics. Ever since I completed my thesis in the CNRS/Thales Lab, I’ve been studying physics and nanometric magnetic oscillators. In 2009, researchers from the United States made electronic devices that imitate some functions of the biological synapses. This work has fascinated me, and pushed me to present innovating concepts to make faster and more performant artificial nanometric synapses. With the help of an ERC Starting Grant, I was able to follow through this project and make these nano-devices, thus giving new leads for brain-inspired electronics. As part of this first ERC project, I had the chance to collaborate with researchers in neurosciences, informatics, electronics and physics, which gave me the interdisciplinary virus. Together we built a national research network, the GDR BioComp, to exchange and get further towards the conception of brain-inspired electronic chips. Indeed, these last ten years have witnessed extraordinary advances in the comprehension of the brain and artificial intelligence which lay the basis for our research. It’s a flourishing field that remains to be cleared out and it’s full of fundamental results such as the comprehension of the basis of natural computation as well as applicative results such as the production of intelligent chips. I personally hope to take part in this momentum by inventing a research between physics and nano-electronics, which experimental developments will be the communicating and dynamic networks of nano-synapses and nano-neurons, and which theoretical advances will allow to optimize the computing methods with the complex vibrations of these nanometric orchestras.
You’ve just been granted an ERC Consolidator, what is the benefit of such a grant on your research for the next five years?
This grant will allow me to get my own team to go through my ideas and to buy the necessary equipment. I will be able to recruit during five years several young PhD and post-PhD researchers on this interdisciplinary and thrilling subject. We will combine our efforts to lay the basis for new brain-inspired electronics. As it is a long-term grant, I will be able to devote myself to my research during these five years, without having to answer to low-success-rate calls for projects all the time.
Where do you see yourself in 2021, at the end of your project?
I have the chance to work in an extraordinary laboratory, CNRS/THALES, with an excellent environment and with dynamic researchers. Five years from now, my only wish would be to be able to recruit a young researcher for a permanent job. It would allow to strengthen this theme of research in the laboratory, to stay ahead in a very promising field and to explore the multiple leads towards a bio-inspired computing between physics and nanotechnologies.
What were the supports you received to prepare you proposal?
I benefited from the advice of an agency which is specialized in the management of ERC projects through the University Paris-Saclay and also from the advice of several colleagues for the written part. Although in my case, it is really the help provided by the University Paris Saclay and the CNRS for the preparation to the orals that was decisive. Presenting your project several times in front of researchers from different fields gives the opportunity to sharpen you presentation until it’s understandable and exciting for a large public, and to be ready to answer any type of questions.
To go further:
Julie Grollier's mail : firstname.lastname@example.org
Julie Grollier's blog : julie.grollier.free.fr/
Interview by the Service Europe of the Université Paris-Saclay.