Breaking
OpenAI announces GPT-5 with breakthrough reasoning capabilities | OpenAI announces GPT-5 with breakthrough reasoning capabilities |

Home / Starcloud Bets on SpaceX Laser Mesh to Power Orbiting AI Data Centers

Science

Starcloud Bets on SpaceX Laser Mesh to Power Orbiting AI Data Centers

Saran K | May 27, 2026 | 4 min read

orbital data centers

Table of Contents

    Moving the Cloud to Low Earth Orbit

    The push to move heavy AI compute off the terrestrial grid and into vacuum is accelerating. Starcloud, the Redmond, Washington-based startup, recently announced a strategic contract with SpaceX to integrate optical terminals into its upcoming fleet of orbital data centers. By leveraging Starlink’s laser-based mesh network, Starcloud aims to solve one of the most persistent bottlenecks in space operations: the reliance on bandwidth-constrained ground stations.

    The agreement includes an order for more than 50 Starlink Mini Lasers. These terminals will be deployed in pairs across at least 25 satellites, serving as the critical networking backbone for a constellation that Starcloud envisions eventually scaling to 88,000 orbital nodes. Rather than waiting for a satellite to pass over a specific ground antenna to offload data, these laser links allow Starcloud to route information across the Starlink mesh, providing a more seamless, global relay for commercial cloud workloads.

    From Demonstrators to Commercial Scale

    This hardware shift comes as Starcloud prepares for the launch of Starcloud-2, slated for January on a SpaceX Falcon 9. While the company’s first effort, Starcloud-1, was a 60-kilogram demonstrator deployed in November, the second iteration is a significant leap in capability. The Starcloud-2 spacecraft weighs 450 kilograms and is engineered to generate roughly eight kilowatts of power—a hundredfold increase over its predecessor.

    This power jump is essential for the company’s early customers, including AI infrastructure provider Crusoe. With partners like Nvidia, AWS, and Google already in the mix, Starcloud is positioning itself not just as a satellite operator, but as an infrastructure-as-a-service (IaaS) provider for the vacuum of space. The goal is to provide a sanctuary for AI inference and training that bypasses the land, water, and power constraints currently choking terrestrial data centers.

    The Starship Factor and Technical Hurdles

    While the Falcon 9 launches provide immediate momentum, the long-term play depends on SpaceX’s Starship. Starcloud is designing the Starcloud-3—a three-ton, 200-kilowatt-class behemoth—specifically for the massive payload capacity of Starship. These larger nodes would handle the most intensive AI workloads, which require immense power and thermal management capabilities that current small-sat launchers cannot support.

    CEO Philip Johnston has estimated a mid-to-late 2028 window for Starship to begin deploying these larger payloads. In the interim, the company plans to “tread water” by deploying Starcloud-2-class satellites to serve edge-computing needs for defense and Earth observation users. These users require real-time analytics—such as wildfire detection and weather forecasting—where processing the data in orbit is far more efficient than sending raw sensor data back to Earth for analysis.

    The Connectivity Trade-off

    The Starlink Mini Lasers provide up to 25 gigabits per second (Gbps) of connectivity over distances up to 4,000 kilometers. To a terrestrial network engineer, 25 Gbps might seem modest compared to fiber-optic speeds, but in the context of a 200-kilowatt orbital node, it represents a highly efficient ratio of connectivity per watt. Furthermore, outsourcing the networking layer to SpaceX allows Starcloud to focus its engineering resources on the much harder problems of radiation-tolerant computing and heat dissipation in a vacuum.

    The market for orbital compute is becoming increasingly crowded. SpaceX itself has discussed plans for a million orbital data centers to support xAI and Tesla. However, Johnston argues that a distinction exists: while SpaceX will likely prioritize internal workloads, Starcloud is building an open infrastructure for third-party developers and enterprises.

    The ambition is matched by a need for capital. Sources indicate that Starcloud is currently seeking at least $200 million in new funding to scale its constellation, following a $170 million round in March that brought its total funding to $1.1 billion. As other players like Vast and Muon Space also integrate Starlink laser links, the orbit is quickly becoming a networked extension of the global cloud.

    Related News

    #space-tech #artificialIntelligence #cloudInfrastructure #satelliteNetworks #dataRelay #orbitalDataCenters #sn #spacex #starcloud

    Related Posts

    Leave a Reply

    Your email address will not be published. Required fields are marked *