Session

Session 13: Next on the Pad 2

Abstract

The Department of Defense Space Test Program (STP) provides spaceflight opportunities for conducting on-orbit research and technology demonstrations to advance the future of spacecraft. STP-H6, the next mission of the program to the International Space Station (ISS), will include a prototype spacecraft supercomputing experiment and framework, called Spacecraft Supercomputing for Image and Video Processing (SSIVP), developed at the National Science Foundation (NSF) Center for High-Performance Reconfigurable Computing (CHREC) at the University of Pittsburgh. SSIVP introduces scalable, high-performance computing (HPC) principles to a CubeSat form-factor to advance the state of the art in space computing. SSIVP adopts the CHREC Space Processor (CSP) concept, a multifaceted design philosophy for a hybrid system of commercial and radiation-hardened (rad-hard) components supplemented with fault-tolerant computing, and a hybrid processor combining fixed-logic CPU and reconfigurable-logic FPGA. SSIVP features five flight-qualified CSPv1 computers as compute nodes, to facilitate this supercomputing concept, and one μCSP smart module, for running a Gallium Nitride (GaN)-based power converter sub-experiment. SSIVP is a versatile, heterogenous platform capable of processing application workloads in the processor or on runtime-reconfigurable FPGA accelerators. In this paper, we present the flight hardware and software, frameworks for parallel and dependable computing, and mission objectives for SSIVP.

sebastian_sabogal.pdf (3590 kB)
Presentation

Share

COinS
 
Aug 10th, 11:30 AM

SSIVP: Spacecraft Supercomputing Experiment for STP-H6

The Department of Defense Space Test Program (STP) provides spaceflight opportunities for conducting on-orbit research and technology demonstrations to advance the future of spacecraft. STP-H6, the next mission of the program to the International Space Station (ISS), will include a prototype spacecraft supercomputing experiment and framework, called Spacecraft Supercomputing for Image and Video Processing (SSIVP), developed at the National Science Foundation (NSF) Center for High-Performance Reconfigurable Computing (CHREC) at the University of Pittsburgh. SSIVP introduces scalable, high-performance computing (HPC) principles to a CubeSat form-factor to advance the state of the art in space computing. SSIVP adopts the CHREC Space Processor (CSP) concept, a multifaceted design philosophy for a hybrid system of commercial and radiation-hardened (rad-hard) components supplemented with fault-tolerant computing, and a hybrid processor combining fixed-logic CPU and reconfigurable-logic FPGA. SSIVP features five flight-qualified CSPv1 computers as compute nodes, to facilitate this supercomputing concept, and one μCSP smart module, for running a Gallium Nitride (GaN)-based power converter sub-experiment. SSIVP is a versatile, heterogenous platform capable of processing application workloads in the processor or on runtime-reconfigurable FPGA accelerators. In this paper, we present the flight hardware and software, frameworks for parallel and dependable computing, and mission objectives for SSIVP.