S U P E R C O M P U T I N G

An architecture diagram of the NASA Center for Climate Simulation (NCCS) Centralized Storage infrastructure, illustrating its connectivity to various NCCS systems. Aaron Knister, NASA/Goddard

Conceptual illustration of high-performance Centralized Storage that facilitates cross-system scientific discovery by enabling analytics and modeling that uses shared satellite, airborne, and surface observation data. For example, the arrows represent generation and utilization of a time series of Arctic-boreal forest canopy elevation measurements from NASA’s ICESat-2 lidars, which can provide validation data for simulations of forest succession and biomass change.Aaron Knister, Kate Ramsayer, NASA/Goddard

Aaron Knister, NASA Goddard Space Flight CenterEllen Salmon, NASA Goddard Space Flight Center

As science datasets expand and the NASA Center for Climate Simulation (NCCS) adds new computing platforms, there is an ever-growing need to share data between NCCS systems. To reduce duplication and time spent moving data, the NCCS designed and is implementing a Centralized Storage infrastructure providing high-speed access to commonly used shared datasets from a variety of systems. The Centralized Storage will not only offer native General Parallel File System (GPFS) access to data from the NCCS Discover supercomputer but also facilitate Server Message Block/Common Internet File System (SMB/CIFS) access to Windows servers and Network File System (NFS) access to data analytics workloads including the NCCS Advanced Data Analytics Platform (ADAPT).

SCIENCE MISSION DIRECTORATE

Multi-Protocol High-Performance Centralized Storage