Framework for the rendering of large point clouds using GPU compute shaders instead of the traditional rendering pipeline. The size of the point clouds is limited by the memory capacity of the GPU, as they are fully loaded into the GPU, either as a single buffer or several chunks. The benefits of this framework are presented in a Conference paper entitled "Comparison of GPU-based methods for handling point cloud occlusion", where sorting algorithms are omitted, though they allow reducing the response time up to the half of the reported results. For that purpose, points are sorted along a Z-curve to place close points in similar buffer indices. This methodology is also used for providing a reduction tool, in order to decrease the number of points (mainly for large point clouds which cannot be fully wrapped in the GPU). Some screenshots from the rendering tool are following attached: