The availability of reconfigurable radiation tolerant Field Programmable Gate Arrays (FPGAs) for space applications creates an opportunity to explore and exploit new processing paradigms for increased computing performance. Space-qualified device availability and promising initial performance studies suggest that development of a space-qualified reconfigurable computing platform is a viable computing solution for space applications. Additionally the availability of soft processors with operating system support provide the capability to combine traditional von-Neumann style computing with dataflow-based computing to implement an optimized computing solution for each application.

RSC provides a user-programmable reconfigurable fabric built from SRAM-based FPGAs coupled with all the resources needed to support development and implementation of space applications. Both software support (i.e. operating system and development tools) and hardware support (i.e. networking, command/control interfaces, state-of-health monitoring) are part of the RSC. Users can apply all or part of the available resources for their system’s implementation; thus allowing for customization to their mission’s operational concept without hardware redesign.Reconfigurability equates to flexibility and reusability, thus reducing overall system cost. Since the fabric of a reconfigurable system can change from mission to mission, the hardware can be reused over and over again with little or no redesign. This approach will eliminate qualification costs associated with one-of-a-kind custom-built systems. The less new hardware developed is for each mission, the less the risk to cost and schedule and therefore a better outlook for mission success.

RSC team members include:

• NASA Langley Research Center – Reconfigurable Systems, Computer Architecture, Project Management, Space Flight Systems
• NASA Goddard – Performance Analysis, Space Flight Systems, Reconfigurable Systems
• University of Queensland – Architectures and Operating Systems for Reconfigurable Computing, Soft cores
• Jefferson Laboratory (JLab) – High Performance Computing, Messaging Software, Multi-processor Computing