The invention provides a parallel point cloud DEM generation method based on machine learning prediction calculation intensity, and the method is characterized in that the method comprises the steps:carrying out the training of a CART model including feature selection, generation of sample data, model training, and storage. Using a quadtree to recursively divide to-be-processed point cloud data,using a stored CART model to predict the computational strength of each tile, and based on the Z-curve, carrying out dimension reduction processing on the tiles of the two-dimensional space, and thenmapping the tiles to each process based on the predicted tile calculation intensity and spatial coding for parallel interpolation of DEM. The technical scheme of the invention has high efficiency andfeasibility. Compared with a traditional serial point cloud generation DEM, the parallel point cloud generation DEM method based on machine learning prediction calculation intensity is adopted, performance improvement can achieve breakthrough of the magnitude, execution time is effectively saved, and the DEM method is particularly suitable for data processing of mass dense point clouds and supports geology application real-time performance.