Neuromorphic Engineering: Brain-Inspired Approaches for Enhanced Energy Efficiency and Resilience

At SMILIES, the central focus is to delve deeply into the intersection of resilient computer architectures and life sciences, addressing all dimensions of this convergence. This includes everything from designing robust, adaptive systems to leveraging biological principles for technological innovation. Neuromorphic engineering perfectly embodies this fusion, aligning seamlessly with the group’s multidisciplinary approach to innovation in both computing and biology.
 
The main goal of neuromorphic engineering is to design computational systems that mimic the structure and function of the brain. Drawing inspiration from neural networks and biological processes, this field seeks to develop more efficient, adaptive, and parallel computing models. Neuromorphic systems leverage insights from neuroscience to create hardware that can process information in a brain-like manner. This cross-disciplinary approach bridges cutting-edge technology and biology, unlocking new computing and cognitive science possibilities.
 
At SMILIES, there is a strong legacy in digital design and reconfigurable hardware, such as Field Programmable Gate Arrays (FPGA). Building on this expertise, the group aims to apply reconfigurable computing to create specialized hardware accelerators for neuromorphic systems, such as Spiking Neural Networks (SNN). The goal is to develop solutions that are both highly optimized for performance and easy to use, even for non-expert users, to ensure that advanced neuromorphic technologies remain accessible while maintaining cutting-edge efficiency.
 
Additionally, we aim to thoroughly assess the resilience of neuromorphic systems by stressing the architecture with potential malfunctions or errors, such as those caused by wear and tear or external agents (e.g., neutron strikes). By simulating these adverse conditions, we aim to pinpoint the most vulnerable components and develop effective strategies to safeguard them, thereby enhancing neuromorphic hardware’s reliability and fault tolerance.
 
By integrating insights from how biological systems efficiently handle energy and maintain robustness, we aim to develop cutting-edge, energy-efficient hardware. This hardware will support and complement traditional computing systems and remain accessible and user-friendly, even for those who are not experts in the field