The Mapping Machine Learning to Physics (ML2P) program aims to increase the military's ability to adapt machine learning (ML) on the battlefield by providing energy-aware ML and enabling the strategic use of limited power resources. The program will map ML efficiency to physics using precise granular measurements, ensuring metrics are directly comparable across hardware architectures. The key objectives of the program include developing multi-objective optimization functions that balance power consumption with performance metrics and discovering how local optimizations interact through Energy Semantics of ML (ES-ML) to solve the energy-aware ML optimization problem. Some of the key aspects of this program include: • Developing energy-aware ML with optimized power and performance for a given task and candidate hardware • Exploring the power-performance interactions between local optimizations via capturing the energy semantics of ML (ES-ML) • Enabling accurate predictions of the power and performance of future ML models • Providing the foundation for simulation research in hardware design that understands power consumption throughout the ML life cycle