Making COTS Rad-Hard for Space

The rapid expansion of satellite constellations has created a trillion-dollar space data challenge driven by unprecedented data volumes, limited downlink capacity, and increasing demands for real-time decision-making. Modern satellites routinely generate terabytes of data per day, yet traditional architectures rely on transmitting raw data to Earth for processing. This approach is constrained by expensive and slow ground station access, high latency, and reduced responsiveness for time-critical applications. As a result, much of the value embedded in space-derived data is delayed or lost entirely.

Edge computing in space addresses these challenges by shifting data processing from the Earth edge to the space edge. However, deploying high-performance computing in orbit introduces reliability concerns due to radiation effects, particularly when leveraging commercial off-the-shelf (COTS) and modified off-the-shelf (MOTS) hardware. Despite these risks, COTS-based FPGAs and GPUs offer a compelling advantage, operating three to five technology generations ahead of traditional radiation-hardened processors, delivering 10×–100× performance improvements and 20×–100× cost savings.

Spirit’s NOVI-powered solution enables reliable and resilient edge processing in space by combining high-performance on-board computers (OBCs) with radiation-tolerant system architectures. NOVI moves the analysis layer into orbit, allowing data to be processed in real time directly on the satellite. Raw imagery and sensor data are transformed into smaller, mission-specific outputs, including processed data products, actionable insights, and operational intelligence. Data is continuously collected, analyzed, and refined into user-requested outputs, supporting applications such as collision avoidance, Earth observation analytics, autonomous spacecraft operations, and on-board anomaly detection.

By transmitting only processed data and intelligence to Earth, NOVI significantly reduces bandwidth requirements, lowers latency, and decreases operational costs. This approach reframes radiation tolerance as a risk-management decision—effectively an insurance model—allowing satellite operators to balance performance, cost, and mission assurance while unlocking the full value of space-based data through edge computing.