Because of big data on energy consumption, there is a lack of research on the discrete manufacturing system. The discrete manufacturing system has plenty of multi-source and heterogeneous data; it was challenging to collect real-time data. Recently, low carbon and green manufacturing is a hot field; especially, it can save electrical energy. This paper proposes a significant energy consumption data of a data-driven analysis framework, which promoting the energy efficiency of discrete manufacturing plant, equipment, and workshop production process. Firstly, put forward the evaluation standards of energy efficiency for discrete manufacturing shops. Then make energy-consumption data preprocessing. Efficiency optimization of big data mining method is put forward based on grid computing function. Design the discrete manufacturing system energy-consumption parameter values, then summarizes prediction algorithms and models in order to predict the results and the trends. Finally, introduce the application of a mobile phone shell manufacturing shop to verify the proposed framework. Further research will focus on energy-consumption data mining processing.