The JPL Snapdragon Co-Processor: A Compact High-Performance Computer for Spaceflight Applications

Following in the footsteps of the resounding success of the Ingenuity Mars Helicopter which was powered by a Qualcomm® Snapdragon<sup xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">TM</sup> 801 system-on-chip (SoC), the Jet Propulsion Laboratory has continued its investments into high-performance spaceflight computers based on Qualcomm's Snapdragon line of SoCs. One significant achievement of these efforts is the development of the JPL Snapdragon Co-Processor (SCP), a small form-factor computer for spaceflight applications featuring the automotive-grade Snapdragon SA8155P SoC. The technology readiness level (TRL) of the SCP was raised to TRL-6 in January 2024. The SCP is currently being included in two upcoming CubeSat-based on-orbit technology demonstration missions and is under consideration as a computer vision processor for missions up to and including class-B. The Snapdragon SA8155P features an octa-core Arm® CPU cluster with four Cortex® A-76 and four Cortex A-55 cores, a graphics processing unit (GPU) capable of 898 GFLOPs (32-bit floating-point), and a cluster of 4 Hexagon™ DSP cores. The SCP board is outfitted with 16 GB of RAM, 128 GB of non-volatile Flash memory, and 2 Mb FRAM. The external interfaces of the SCP are two USB 3.1 Gen2 ports, a 4×4 lane MIPI Camera Serial Interface connector, and a 200-pin space-grade mezzanine connector featuring three total PCI Express lanes, an RGMII interface to support Gigabit Ethernet, and low-speed UART, GPIO, JTAG, and SPI connections. To utilize the interfaces on the mezzanine connector, JPL has developed a variety of custom carrier cards for the SCP, ranging from the advanced and high-performance line of Swift Processor Modules to a dedicated SCP carrier card for CubeSat applications. In this paper, we present a detailed overview of the SCP design, its capabilities, and its interfaces to the spacecraft bus. We describe the available options to integrate the SCP into a spacecraft using the currently available and in-development carrier boards as examples. We comment on the software support and past and future JPL software benchmarking and porting efforts. Finally, we comment on the TRL-6 test campaign and give a brief outlook of the future of the SCP for JPL missions and the wider spaceflight community.

Authors

Institutions

Publication Details

Published
2025-03-01
DOI
https://doi.org/10.1109/aero63441.2025.11068431
Primary Topic
Distributed and Parallel Computing Systems
Type
article
Field-Weighted Citation Impact
0.00

Funders

Controls
|||
ALL TIME
JAN
FEB
MAR
APR
MAY
JUN
JUL
AUG
SEP
OCT
article

The JPL Snapdragon Co-Processor: A Compact High-Performance Computer for Spaceflight Applications

Joseph Sauvageau, Gregory Miles, Douglas Sheldon, Andre Jongeling et al.
Distributed and Parallel Computing Systems
article

The JPL Snapdragon Co-Processor: A Compact High-Performance Computer for Spaceflight Applications

Joseph Sauvageau, Gregory Miles, Douglas Sheldon, Andre Jongeling, Zaid J. Towfic, David Foor, Bell Irvin Wiley, Timothy Canham, Daniel Cho, Dennis Ogbe, Edwin Grigorian, Clayton Okino, Saba Janamian
article en

Abstract

Following in the footsteps of the resounding success of the Ingenuity Mars Helicopter which was powered by a Qualcomm® Snapdragon<sup xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">TM</sup> 801 system-on-chip (SoC), the Jet Propulsion Laboratory has continued its investments into high-performance spaceflight computers based on Qualcomm's Snapdragon line of SoCs. One significant achievement of these efforts is the development of the JPL Snapdragon Co-Processor (SCP), a small form-factor computer for spaceflight applications featuring the automotive-grade Snapdragon SA8155P SoC. The technology readiness level (TRL) of the SCP was raised to TRL-6 in January 2024. The SCP is currently being included in two upcoming CubeSat-based on-orbit technology demonstration missions and is under consideration as a computer vision processor for missions up to and including class-B. The Snapdragon SA8155P features an octa-core Arm® CPU cluster with four Cortex® A-76 and four Cortex A-55 cores, a graphics processing unit (GPU) capable of 898 GFLOPs (32-bit floating-point), and a cluster of 4 Hexagon™ DSP cores. The SCP board is outfitted with 16 GB of RAM, 128 GB of non-volatile Flash memory, and 2 Mb FRAM. The external interfaces of the SCP are two USB 3.1 Gen2 ports, a 4×4 lane MIPI Camera Serial Interface connector, and a 200-pin space-grade mezzanine connector featuring three total PCI Express lanes, an RGMII interface to support Gigabit Ethernet, and low-speed UART, GPIO, JTAG, and SPI connections. To utilize the interfaces on the mezzanine connector, JPL has developed a variety of custom carrier cards for the SCP, ranging from the advanced and high-performance line of Swift Processor Modules to a dedicated SCP carrier card for CubeSat applications. In this paper, we present a detailed overview of the SCP design, its capabilities, and its interfaces to the spacecraft bus. We describe the available options to integrate the SCP into a spacecraft using the currently available and in-development carrier boards as examples. We comment on the software support and past and future JPL software benchmarking and porting efforts. Finally, we comment on the TRL-6 test campaign and give a brief outlook of the future of the SCP for JPL missions and the wider spaceflight community.

Jet Propulsion Laboratory (US)
National Aeronautics and Space Administration, California Institute of Technology, Jet Propulsion Laboratory
Affordable and clean energy
Openalex Percentile: Top 81%
Distributed and Parallel Computing Systems
AI Navigator

Ask Laika to Summarize, Analyze, and Connect papers live on the map.

Summarize Papers & Methodologies

Extract key findings, datasets, and comparative methods across publications.

Benchmark Rankings & Visual Analytics

Rank top research institutions, authors, funders, topics, and journals by Field-Weighted Citation Impact (FWCI) and paper volume with instant charts.

Connect Distant Disciplines

Bridge topological clusters on the map to find hidden collaborative intersections.