Routing update is a severe problem for Ternary Content-Addressable Memories (TCAMs) based systems. This study is focused on designing a power efficient and fast updating algorithm for the partitioned TCAMs in IP forwarding systems. TCAMs are usually partitioned into individual buckets for their high power consumption. Due to routing rules change quickly, keeping the latest routing table may cause the partitioned TCAM buckets to overflow and bring additional power consumption. In order to improve the performance of the routing update, an efficient updating algorithm for the partitioned TCAMs is proposed in this study. In the algorithm, free space added for update traces is provided according to the distribution of the routing rules. The problem of buckets overflow is solved by dynamically adjusting the free space. Furthermore, the update speed is accelerated through arranging the free space of each partitioned bucket. Compared with the algorithm which increase the partitioned buckets by k (k is a fixed constant) times, the experimental results show the proposed algorithm is faster and more power efficient.