Accessibility to high resolution LiDAR datasets is essential in supporting earth science applications. LiDAR products, however, are provided at fine spatial resolutions; as such, the data volume of these products is very large, even for a small study area. Consequently, processing massive LiDAR data in a serial manner takes an extremely long time. The developed CUDA-based parallel library designed for processing LiDAR data. By exploiting the parallel computing capabilities provided by many computing cores of Graphics Processing Units (GPUs), this library can significantly accelerate the speed of geoprocessing operations for LiDAR datasets.