Agility of physical manufacturing unit is the competitive advantage in the global manufacturing environment. It is believed that the agility can be realized by dynamically optimization deployment of networked manufacturing resources. To solve this problem, logical manufacturing unit (LMU) and logical manufacturing process (LMP) are proposed and defined to decompose and model networked manufacturing task according to the process of complex part. When selecting manufacturing resources for these manufacturing tasks, many factors should be taken into account. However, manufacturing cost, time to market and manufacturing quality are the most important factors. In this paper, networked manufacturing resources pre-deployment is carried out to find candidate manufacturing resources based on manufacturing resources abilities, such as part family, geometric feature, material type, rough type, dimension range, machining method, precision grade and production type. Then, taking transportation time and cost besides manufacturing time, cost and quality into consideration, the objectives and restrictions of manufacturing resources optimization deployment are analyzed, and manufacturing resources optimization deployment problem is considered as a multi-objectives optimization problem.