Cloud Computing attracts much attention in the community of computer science and information technology because of resource efficiency and cost-effectiveness. It is also evolved to Mobile Cloud Computing to serve nomadic people. However, any service in Cloud Computing System inevitably experiences a network delay to access the computing resource or the data from the system, and entrusting the Cloud server with the entire task makes mobile devices idle. In order to mitigate the deterioration of network performance and improve the overall system performance, we propose a collaborative framework that lets the mobile device participate in the computation of Cloud Computing system by dynamically partitioning the workload across the device and the system. The proposed framework is based on it that the computing capability of the current mobile device is significantly enhanced in recent years and its multi-core CPU can employ threads to process the data in parallel. The empirical experimentation presents that it can be a promising approach to use the computing resource of the mobile device for executing computation-intensive tasks in Cloud Computing system.