Project Grant 2449506

This $400,000 Project Grant awarded by the National Science Foundation (NSF) Engineering program (CFDA 47.041) will develop a new hardware system to accelerate computational neuroscience modeling and simulation. The project aims to create a scalable accelerator capable of efficiently simulating complex, continuous-time models of neuronal learning and memory processes. The proposed solution leverages principles from neuromorphic computing to enable parallel processing of spatially discrete, non-locally coupled systems of differential equations. This collaborative effort between researchers in the U.S. and Switzerland will validate the approach through an FPGA implementation and silicon prototype of a cortical learning model, while also developing novel model variants for long-term processing and learning. The outcomes of this project are expected to advance fundamental neuroscience research and facilitate the translation of biological computing insights into novel commercial applications.