In 1980s a processor had only one core, whose performance was improved by increasing the processor frequency. But this caused overheating and thus improvement in performance was halted. To get more performance multi core processor was introduced that gave better performance. For a chip multiprocessor with hundreds of cores, the cache coherence communication affected the memory access time. The goal of this work is to improve scalability in System on Chip(SoC) multi core processors by using the Hammer cache coherence protocol to maintain cache coherency and reduce the communication traffic by using 3D Mesh Network on Chip as the interconnection network. The GEM5 open source simulator is modified as a part of this thesis to simulate the 3D Mesh NoC. The 3D NoC is observed to produce 11\% reduction in network traffic than the 2D NoC for the Hammer cache coherence protocol and is predicted to provide better performance for higher number of cores and thus improving the scalability of the system. This research is organized into three sections. In the first section, the discussion of different types of cache coherence protocols are discussed and the selection of Hammer protocol for scalability is explained. In the second section, the discussion of various Network on Chips are discussed. In the third section, the implementation in GEM5 and the network algorithm is discussed.