Data Partitioning is one of the key approaches followed in parallelism. It has proved its efficiency on many algorithms and still lot of work is going on this area. In this paper we propose an optimization technique for string matching using data partitioning with multi-core architecture. The paper primarily focuses on caching and re-utilization of processes. The experiments showed that concept of caching increased the speed drastically for frequently asked patterns. The MPI proposed implementation highlighted the increase in efficiency using multi-core and decrease in performance when the cores were reutilized.