Performance enhancement, cost reduction, data security, and improved energy efficiency are the end goals for optimizing AI workloads at the edge.