F. Danneville

Centre National de la Recherche Scientifique, Université de Lille

Papers

Total Citations

286

H-Index

About

François Danneville is a prominent researcher specializing in ultra-low power analog design, neuromorphic computing, and bio-inspired circuit technologies. Working at the intersection of microelectronics and artificial intelligence, his research has focused on developing highly energy-efficient hardware implementations of neural computing paradigms, particularly as conventional CMOS scaling approaches its physical limits. Danneville's most influential contribution is his groundbreaking work on artificial neuron circuits implemented in standard CMOS technology. His 2017 paper introducing a 4-fJ/spike artificial neuron in 65 nm CMOS — garnering an impressive 211 citations — demonstrated a remarkable leap in energy efficiency for neuromorphic hardware, establishing a benchmark for the field. This work addressed the critical challenge of energy dissipation in computing systems by leveraging spiking neural network architectures inspired by biological neurons. Building on this foundation, his 2019 work on a sub-35 pW Axon-Hillock artificial neuron circuit (62 citations) pushed power consumption boundaries even further. Danneville's research is particularly significant for its practical implications in cognitive computing and large-scale neuromorphic systems, offering viable pathways for hardware that mimics brain-like processing while maintaining compatibility with existing semiconductor manufacturing processes.