On the Eve of Its IPO, Musk Details "Space Data Center Plan": Not a Major Challenge for SpaceX

On the eve of SpaceX's IPO, Musk detailed his orbital AI data center plan for the first time, positioning it as the company's core growth engine. The initiative aims to break the constraints imposed by Earth's power supply on the development of the AI industry by transferring massive computing power to low Earth orbit.

In a video released on Monday, Musk publicly unveiled the design sketches and core technical parameters of the first-generation AI satellite, "AI1," for the first time.

Musk stated clearly that establishing data centers in space does not require undiscovered "magic." The technical challenges involved are even lower than those of the existing Starlink business. This is not an extremely difficult engineering problem for SpaceX, which will strive to achieve an annualized deployment rate of 1 gigawatt (GW) of space-based AI computing power by the end of 2027.

This latest statement sent a clear signal to the capital markets: SpaceX is attempting to transform its absolute scale advantages in satellite mass production and launch into a moat for next-generation AI computing infrastructure.

Although industry competitors remain cautious about the economic feasibility of space-based computing, SpaceX has applied to the Federal Communications Commission (FCC) to launch up to one million AI satellites. To address the high barriers associated with computing and launch costs, SpaceX is advancing self-developed chip factories in partnership with Tesla and Intel, and plans to use the Starship heavy-lift rocket to disrupt existing launch economics. This strategy directly widens the gap between SpaceX and its competitors in terms of strategic deployment.

Targeting Computing Bottlenecks and a Vast Potential Market

In its IPO filing, SpaceX pointed out that the total potential market for AI, estimated at up to $26.5 trillion, would be severely constrained by "Earth's inability to rapidly expand power generation capacity." Therefore, solar-powered orbital AI data centers are viewed by Musk and aerospace executives as a key technology to meet the growing energy demands of AI companies.

Regarding the construction timeline, Musk provided an ambitious forecast. He stated that SpaceX would strive to achieve an annualized deployment rate of 1 gigawatt (GW) of space-based AI computing power by the end of 2027, seeking to expand at an order-of-magnitude pace annually, ultimately reaching a computing scale of 1 terawatt (TW).

However, he also cautioned investors to "remain reserved" about this aggressive timeline, while the official expectations outlined in the IPO filing were more conservative, suggesting a gradual commercial rollout starting in 2028.

Unveiling the AI1 Satellite: An NVIDIA Rack in Orbit

Addressing the misconception that space data centers involve simply "launching ground server rooms into space," SpaceX clarified the actual hardware configuration. The core engineering challenge is not the physical relocation of buildings, but rather acquiring electrical power in a vacuum environment and efficiently radiating the waste heat generated by high-power computing.

The video showcased the quantitative metrics of the AI1 computing satellite for the first time. The satellite has a peak power consumption of 150 kW and a sustained average computing power consumption of 120 kW.

Musk stated that these figures precisely match the operating power envelope of the NVIDIA GB300 computer racks (containing 72 GPUs) used in ground-based data centers, equivalent to sending an entire NVIDIA AI computing module directly into space.

To meet the extreme energy consumption and cooling requirements, the AI1 satellite features a pair of massive wings with a wingspan of 70 meters. The solar array power generation density is set at 250W/m², and the two-sided cooling panels have a heat dissipation density of 1400 W/m². In orbit, the satellite will adopt a "blade-like" orientation facing the sun to maximize heat dissipation.

Streamlined Architecture and Technology Reuse to Build a Manufacturing Moat

In terms of hardware structure, the engineering design of the AI1 satellite is even more streamlined than that of traditional Starlink satellites.

Existing Starlink satellites require extremely complex large-scale phased-array antennas and parabolic antennas. In contrast, AI satellites are essentially more like pure large-scale hardware: they mainly consist of huge solar arrays, oversized cooling panels, and basic laser links, eliminating the need for complex ground-communication antennas.

Musk and the engineering team emphasized that the manufacturing of AI satellites largely reuses the Starlink V3 satellite platform technology already developed and verified by SpaceX.

This means SpaceX can directly transfer its existing experience in large-scale satellite mass production, launch, and operations without needing breakthroughs in fundamental science. On the verge of its IPO, this high rate of technology reuse and engineering scalability constitutes a unique competitive advantage that the company is showcasing to investors.

Ecosystem Synergy to Address Cost and Latency Challenges

Addressing concerns about network latency associated with space data centers, SpaceX provided a clear solution.

AI satellites will be deployed in low Earth orbit (LEO) at altitudes of 600 to 800 kilometers above the ground, resulting in a one-way network latency of only about 3 milliseconds. The satellites will integrate inter-satellite laser links with bandwidths of up to 1 Terabit per second, and will high-speed downlink data through Starlink's existing KA and KU band antenna networks, or directly via space-to-ground laser links.

However, there are still divisions within the industry regarding commercial feasibility.

Jeff Bezos, founder of Blue Origin and Amazon, along with researcher Andrew McCalip, among others, have pointed out that expensive AI chips and high launch costs currently constitute industry hurdles, making the current economic model unreasonable.

To this end, SpaceX is attempting to build a vertically integrated supply chain to break through cost barriers: On one hand, it relies on the Starship heavy-lift rocket to significantly reduce launch costs; on the other, through a planned factory named Terafab, it is collaborating with partners Tesla and Intel to research, develop, and manufacture its own AI chips. By controlling both launch capabilities and underlying computing hardware, SpaceX is accelerating the implementation of its space computing commercialization.