Kepler Launches the Largest Orbital Computing Cluster With 40 NVIDIA GPUs

Kepler Communications has launched the largest orbital computing cluster, integrating 40 NVIDIA GPUs across 10 operational satellites, with Sophia Space as its first customer.

Kepler Launches the Largest Orbital Computing Cluster With 40 NVIDIA GPUs
Kepler Communications launches the largest orbital computing cluster with 40 NVIDIA GPUs

In January 2026, Kepler Communications launched the largest orbital computing cluster, integrating 40 NVIDIA Orin GPUs across 10 operational satellites, all interconnected via laser communication links. This development marks a significant advance in processing data directly in space, reducing the need for transmissions to the ground and improving the efficiency of satellite operations.

An Advanced Computing Infrastructure in Space

Kepler's network combines optical connectivity, distributed in-orbit computing and secure payload hosting in a single space infrastructure. The optical links between satellites provide real-time data transport, while NVIDIA-based edge computing enables advanced applications such as AI analysis for Earth observation, multi-sensor data fusion and intelligent data optimization, run directly within the network. This approach makes it possible to generate insights at the edge, reducing downlink requirements and supporting scalable, cloud-native workloads across the constellation. (kepler.space)

Collaboration with Sophia Space

Kepler's first customer for this in-orbit computing capacity is Sophia Space, a startup developing passively cooled space computers. Sophia Space will test its proprietary operating system on one of Kepler's satellites, configured across six GPUs distributed over two spacecraft. This test is a key step for Sophia Space, which aims to launch its first satellite by the end of 2027.

Implications for the Space Sector

This initiative by Kepler Communications and Sophia Space highlights a growing trend in the space industry: processing data directly in orbit. By moving computation into space, companies can reduce latency, improve efficiency and optimize resource use. Moreover, integrating advanced hardware such as NVIDIA GPUs into space environments is a significant technical challenge, but it opens opportunities for more sophisticated applications, such as artificial intelligence and machine learning, directly in space. (datacenterdynamics.com)

Future Prospects

As in-orbit computing capabilities expand, more companies are expected to adopt this technology to improve their space operations. Collaborations between infrastructure providers such as Kepler Communications and software developers such as Sophia Space will be crucial to the evolution of this sector. Furthermore, the adoption of specialized hardware, such as GPUs designed for the space environment, will open new possibilities for advanced applications, from satellite data management to the automation of space operations. (datacenterdynamics.com)

Takeaways

  • Innovation in Orbital Computing: Kepler Communications has integrated 40 NVIDIA Orin GPUs across 10 satellites, creating the largest orbital computing cluster currently in operation.
  • Strategic Collaboration: Sophia Space is the first customer to test its operating system in orbit, using Kepler's infrastructure to validate its technology.
  • Benefits of In-Orbit Processing: Moving computation into space reduces latency and improves efficiency, supporting advanced applications such as AI directly in orbit.
  • Growth Prospects: The adoption of advanced hardware such as NVIDIA GPUs in space environments opens new opportunities for sophisticated applications and for automating space operations.
  • Future Collaborations: Partnerships between space infrastructure providers and software developers will be key to the evolution of orbital computing and advanced space applications.

For further details, see the original sources: , (datacenterdynamics.com).