Amazon SageMaker now supports geospatial Processing jobs, making it easier for data scientists and ML engineers to run planetary-scale ML workloads. To run such large-scale workloads, customers need a flexible compute cluster that can scale from tens of instances to process a city block, to thousands of instances for planetary-scale processing. Manually managing a DIY compute cluster is slow and expensive. Additionally, building and maintaining a standardized environment to access, process, and visualize geospatial data is complex, time consuming, and expensive.
Source:: Amazon AWS