High Performance Computing
Geospatial data are nowadays widely accepted as the base for decision making in almost all aspects of life. Numerous satellite missions are started and especially data from low and medium resolution satellites are freely available. While in theory any citizen can access such data, in practice the processing data in the Giga and Tera byte range remains challenging and complex. In order to get most out of the newly available high temporal resolution time series expertise in the processing of huge amounts of geospatial data is crucial.
We at mundialis implement innovative analysis methods in cloud computing environments. We process geospatial data on High Performance Computing systems be they on Bare Metal servers or Cloud Computing infrastructures. For us the benchmark is: linear scaling with tasks and data in order to transform raw satellite data into aggregated information for decision making.
We employ a series of technologies:
- Gluster File System (GlusterFS): an open source distributed file system which can scale out to multiple petabytes of data
- GDAL: our library of choice for reading all the manifold formats
- GRASS GIS: a free and open source Geographic Information System (GIS) suitable for the analysis of massive geospatial data, image processing, spatial modeling, and visualization. A new HPC related REST API is in the making!
- sen2cor software: used for the atmospheric correction of Sentinel-2 satellite data
- HPC related job managers: distribution of data analysis jobs over multiple compute nodes in the cloud
- and so much more!
We are ready to process massive geodata and satellite data for you, including:
- vegetation index time series, color composites, high resolution image mosaiks
- geospatial data processing from local to continental scale
- time series analysis with true temporal database support
- identification of land use changes in the last 30 years (with Landsat-5, 7, 8, Sentinel-2A)
- identification of urban heat islands (with Sentinel-3A + MODIS + Landsat 8)
- point data interpolation
- regional analysis and the very high resolution analysis for precision farming
Our data centers are located in Germany / Europe. Optionally, we also collaborate with North American cloud computing providers.
Download brochure: mundialis flyer GRaaS HPC (PDF)
Contact us today!