Bump mapping is a technique that represents the detailed parts of an object surface, such as the skin of a peanut, using the geometry mapping without complex modeling. However, the hardware implementation for bump mapping is very expensive, because a large amount of per pixel computations, including the normal vector shading, is required. In this paper, we propose an effective bump mapping algorithm that utilizes the reference space with the polar coordinate system and also propose a new hardware architecture associated with the proposed bump mapping algorithm. The proposed architecture reduces the computations to transform the vectors from the object space into the reference space by using a new vector rotation method. It also reduces the computations for the illumination calculation by using the law of cosine. Compared with the previous approaches, the proposed architecture reduces multiplication operations up to 78%.