In this work, we have explored different task mapping algorithms for multi-core, packet processing systems. we also implemented these algorithms and compared the results of the algorithms. We first reviewed the previously designed algorithms which include UDFS algorithm and duplication process. We then applied the KL algorithm to our problem and were able to reduce the inter-processor communication by 20% while maintaining the similar utilization. We then modified the original KL algorithm by considering utilization during the mapping process. In this extended KL algorithm, we incorporated a tradeoff factor alpha to tradeoff between inter-processor communication and processor utilization. The best alpha is different for different system configurations in terms of communication bandwidth and computing power. Simulated annealing(SA) algorithm was then implemented. The parameters for SA algorithm were decided by following literature or by doing experiments. Results from SA algorithm shows that it can produce decent results that are comparable to KL algorithm. In order to further improve the utilization, merging operation was applied to the task graph before mapping algorithms were applied. The mapping results showed that merging is a good way to improve the utilization and at the same time keep the communication cost lower. Finally, we applied the mapping algorithms to different packet processing system architectures. The results show how inter-processor communication cost and processor utilization change as system architecture changes.