FRIEL Liam

Abstract: The rise of artificial intelligence and embedded systems requires new hardware approaches capable of combining high performance, energy efficiency, and robustness. Neuromorphic architectures, inspired by the functioning of the human brain, represent a promising path to overcome the limitations of conventional (von Neumann) computing architectures.By reproducing the principles of parallel, asynchronous, and distributed processing found in biological neurons, these circuits enable systems that can perform local data processing, adaptive learning, and resilience to hardware faults. The goal of this PhD project is to design, model, and evaluate neuromorphic circuits with low energy consumption and high fault tolerance, suitable for applications in embedded AI and sensory signal processing. The candidate will explore approaches combining hardware neural modeling, network topology optimization, and implementation using emerging technologies.