SpaceX Plans to Power AI Centers with Orbiting Solar Arrays, Sidelines Earth-Based Solar

SpaceX's IPO filing reveals a strategic pivot by Elon Musk’s enterprises, emphasizing space-based solar power over terrestrial solar energy to meet the growing demands of AI data centers. This marks a notable departure from Tesla’s long-standing mission to accelerate a transition to a solar electric economy on Earth.

Tesla’s series of four Master Plans have consistently promoted electrification and a shift away from fossil fuels. Musk’s initial plan explicitly stated the goal to move from a hydrocarbon economy to one powered by solar electricity. However, recent developments indicate a more complex energy approach within Musk’s business ventures.

One example is xAI, a Musk-affiliated company that operates AI data centers powered predominantly by dozens of unregulated natural gas turbines. The company plans to acquire an additional $2.8 billion worth of such turbines, reinforcing the role of fossil fuels in its operations despite Musk’s clean energy background.

Inter-company transactions within Musk’s ecosystem show SpaceX purchasing 1,279 Cybertrucks for $131 million, while xAI has invested $697 million in Tesla Megapacks, large-scale battery storage systems designed to manage peak energy loads. Nevertheless, xAI has not procured a significant quantity of Tesla solar panels for its data centers.

Solar power is present in the SpaceX filing, but it is focused exclusively on space-based solar technology rather than terrestrial applications. The company highlights space solar arrays as the future of powering data centers, citing their ability to generate over five times the energy of Earth-bound solar panels due to continuous illumination in orbit.

Interest in space-based solar power has grown among Silicon Valley leaders, including Musk, as AI data centers face resistance on Earth. Deploying servers in orbit powered by uninterrupted solar energy is being considered as a solution to terrestrial constraints.

Despite this vision, the economics of operating data centers in space remain challenging. The cost of powering Starlink satellites is substantially higher than typical terrestrial data center expenses. Additionally, safeguarding computing chips against the harsh space environment involves significant technical and financial hurdles. It is also uncertain whether AI training workloads can be effectively distributed across multiple satellites, meaning a substantial portion of AI processing would still occur on Earth.

Musk likely views xAI’s current terrestrial data centers as temporary, anticipating that SpaceX will soon be capable of launching gigawatts of servers into orbit. This transition, expected within a few years, would eliminate the need to address local opposition to data centers and reduce reliance on natural gas turbines. However, this plan carries risks if the technological and economic challenges prove insurmountable.

The filing also expresses Musk’s concern that AI’s computing requirements will rapidly exceed Earth’s current power supply capabilities. References to “terawatt-scale annual AI compute growth” appear throughout the document, a figure that dwarfs the approximately 40 gigawatts consumed by all global data centers today.

This approach reflects Musk’s “first principles” reasoning, projecting a future where an additional terawatt of computing power will be needed annually. The company argues that existing third-party estimates underestimate data center demand due to terrestrial supply limitations and that the actual power shortage could be much greater than research suggests.

Humanity’s current annual energy consumption is about 35,000 terawatt-hours, equating to roughly 4 terawatts continuously. While energy demand has been rising, especially with AI’s growth, it remains uncertain whether this trend will persist exponentially or stabilize. Musk’s track record includes identifying such inflection points and extrapolating accordingly.

Practical considerations raise questions about the efficiency of Musk’s space-based solar ambitions. Transporting solar panels via terrestrial methods likely consumes less energy than launching them into orbit. Furthermore, manufacturing space-grade solar panels at the necessary scale presents formidable challenges. These issues suggest that focusing solely on space-based solutions might divert attention from untapped potential in Earth-based solar power.

Three years ago, Tesla’s “Master Plan Part 3” outlined a comprehensive strategy to eliminate fossil fuels. Revisiting this goal might involve addressing the fossil fuel dependency of xAI’s data centers as a starting point, even as Musk pursues ambitious projects in space-based energy.