Low SWaP Flight Board for High-Speed On-Board Data Handling, Processing and AI Inference

The continuous increase in payload data volume and mission autonomy requirements is driving the need for increasingly advanced on-board data handling and processing technologies in satellite systems. In particular, next-generation Earth observation, telecommunications, and distributed space architectures require high-performance processing capabilities to enable real-time data reduction, intelligent decision-making, and efficient use of downlink resources.
Within an ESA-funded project, a high-performance flight board for on-board data handling and processing has been developed to address these challenges. The proposed architecture is based on a dual-FPGA design, enabling parallel high-throughput data processing and robust system management within a single flight-qualified unit. The board supports advanced on-board data processing functions and Artificial Intelligence (AI) algorithms through a GPU-like accelerator implemented on FPGA: GPU@SAT. It executes Machine Learning and Computer Vision workloads directly in orbit without the need of dedicated CPU or external controllers.
The modular architecture has been designed to maximize flexibility and scalability, supporting different mission profiles and processing demands. In parallel, a miniaturized, PC/104-compliant version of the flight board is currently under development, specifically targeting small satellite platforms with stringent constraints in terms of size, weight, and power. This evolution aims to extend high-performance on-board processing capabilities to small satellite missions without compromising performance or reliability.
As next step, the development of an additional flight board based on a high-performance processor is going on, targeting high-performance computing (HPC) and satellite communications (SATCOM) applications. The combination of FPGA-based processing, AI acceleration, and high-performance processors will enable the creation of a complete and versatile on-board computing unit, capable of addressing the demanding requirements of future space missions.
The presented work highlights a scalable and forward-looking approach to on-board processing, enabling intelligent, high-throughput satellite systems for both current and future space applications.