This paper describes a new parallel architectural system which we called Hybrid System. As the name implies, Hybrid System is a combination of both SIMD and MIMD systems working concurrently. This new parallel architecture has the capability of achieving speed-up rates more than MIMD can achieve alone and can also be more flexible than multiple SIMD on a single station. In this paper, we will introduce our new SIMD concept and also show the contribution of the SIMD on the Hybrid System. We have also developed a general formula for determining the Speedup of the Hybrid System so that accurate predictions can be made on its performance. This, we established by comparing the measured speedup with the computed speedup for a volume rendering application.