Super fast implementation of ICP in CUDA for compute capable devices 2.0 or higher - View it on GitHub