1 Introduction
Since its original application in graph processing, general-purpose computing on graphics processing unit (GPGPU or GPU) has been applied in many other fields that require massive parallel computing. In fact, lots of applications have get benefits from the massive parallelism capability of GPU [13], [14], [25], [27], [33], [36], [38]. In addition, researchers have also utilized GPU to solve some specific artificial intelligence (AI) problems successfully [1]. These practice shows that GPU becomes a promising way to solve compute-intensive AI problems due to its SIMD architecture specialized for parallel computing. Furthermore, considering the important role of Game Tree Search (GTS for short) in AI, it is worth to study the possibility and feasibility of utilizing GPU to improve the performance of GTS related algorithms that are widely used in computer games.