We are developing spiking neural networks (SNNs) using advanced silicon neurons that consist on CMOS transistors operated in punch-through impact ionization regime [Nature 640, 69-76, 2025]. SNNs built using this approach consume very low energy and are ideal to deploy AI in mobile objects (edge AI). Our approach is 100% compatible with the industry.

We are integrating novel nanomaterials at the back-end-of-line (BEOL) of silicon microchips to achieve superior electronic performance and reliability for memristive applications [Nature 618, 57-62, 2023]. We use nanomaterials with advanced physical, chemical, thermal, and electronic properties, such as 2D materials and molecular crystals.

Our group is expert in nanomaterials characterization using atomic force microscopy and electron microscopy [Nature Electronics 2, 221–229, 2019]. We apply stress using in operando approaches to identify which are the atomic rearrangements responsible of device operation, and accelerated stresses to explore failure mechanisms.