This research aims to design an Active Magnetic Bearing (AMB) after performing an optimization process via reducing the number of poles and by reducing air gap, Dia. Yoke, and Z-length (deep of model). To increase the performance of a radial Active Magnetic Bearing (AMB), all particular equations of design based on the Genetic Algorithm method by using ANSYS Maxwell (Version 17.1) program of electro-magnetic have been studied. Manufacturing an active magnetic bearing standing for two counts, each one containing 12 poles instead of 16, led to a significant improvement in the performance. Some conclusions were obtained, including the complications in the control system will be reduced when they are linked in AMB. The complexities of the control system are inversely proportional to the number of poles and the model covered in this study is made of a material with good engineering and magnetic characteristics steel 37-2.