
SYNEART : Artificial synapses and neurons based on ferroelectric relaxors: towards lower-power AI.
Coordinator: Brahim DKHIL
MCF
Keywords: Relaxor ferroelectric, thin film, neuromorphic, high throughput synthesis, AI-assisted characterization and analysis

This project aims at revealing the potential of relaxor ferroelectric materials whose characteristics
are very promising for the design of nanoelectronic devices for neuromorphic computing. Unlike
conventional ferroelectric materials, the field of thin films based on relaxers is only just beginning.
Through a combinatorial synthesis approach, high-throughput structural, chemical and electrical
characterization tools coupled with analyzes and predictions based on artificial intelligence, we
wish to identify the most relevant compositions/ thicknesses/electrodes and use them as elementary building blocks to design artificial neurons and synapses. Beyond the neuromorphic relaxor materials of interest in this project, high-throughput synthesis, characterization and predictive protocols, data processing and analysis tools by artificial intelligence, will be developed or consolidated to be made available to the community in order to facilitate and accelerate the discovery of new materials in other fields and their bringing to market.