Nvidia is pushing artificial intelligence (AI) computing beyond Earth with a new platform designed to process satellite data directly in orbit, which could speed up space research, disaster monitoring, and geospatial intelligence.
Nvidia said its latest accelerated computing platforms allow satellites and orbital systems to analyze data in space instead of sending everything back to ground stations for processing. This approach could reduce delays and enable real-time decision-making for satellites and autonomous spacecraft.
“Space computing, the final frontier, has arrived. As we deploy satellite constellations and explore deeper into space, intelligence must live wherever data is generated,” said Jensen Huang, founder and CEO of Nvidia. “AI processing across space and ground systems enables real-time sensing, decision-making, and autonomy, transforming orbital data centers into instruments of discovery and spacecraft into self-navigating systems.”
The Nvidia Space-1 Vera Rubin Module is a compact system designed to deliver powerful AI processing in the harsh environment of space. Nvidia said the Rubin GPU inside the module can provide up to 25 times more AI compute performance than the company’s earlier Nvidia H100 GPU when running space-based AI inference workloads.
The module is designed to support orbital data centers, geospatial intelligence analysis, and autonomous spacecraft operations. Its architecture combines CPU and GPU processing with high-bandwidth memory to handle massive streams of data from satellite sensors in real time.
For space-based edge computing, Nvidia is also deploying platforms such as Nvidia IGX Thor and Nvidia Jetson Orin. These systems are designed to run AI workloads in compact, power-efficient hardware that fits the size, weight, and power limits of satellites and spacecraft.
The Jetson Orin module supports real-time processing of navigation, vision, and sensor data onboard satellites. This capability helps spacecraft respond faster to changing conditions while reducing the amount of raw data that must be transmitted back to Earth.
Ground-based processing remains an important part of the ecosystem. Nvidia said data center systems powered by the Nvidia RTX PRO 6000 Blackwell Server Edition GPU can analyze massive satellite imagery archives up to 100 times faster than traditional CPU-based systems.
The demand for faster processing is rising as commercial space companies deploy thousands of satellites that collect data from Earth-observation sensors, radar systems, and radio-frequency instruments. These systems generate enormous datasets used for disaster response, environmental monitoring, weather forecasting, and infrastructure planning.