NSF/ANR grant on errors in analog neural networks
August 2022
Prof. Makris received a $624K NSF grant (CHAMELEON) to study the impact and mitigation of errors in analog implementations of on-die neural networks.