SpaceX acquires xAI in $1.25 trillion merger to build orbital AI data centers

The Largest Merger in History

On February 2, 2026, SpaceX announced the acquisition of xAI in an all-stock transaction. This deal created a combined entity valued at approximately $1.25 trillion, making it the largest corporate merger ever recorded by valuation. SpaceX accounted for roughly $1 trillion of that total, while xAI contributed $250 billion.

Elon Musk framed the acquisition as a strategic move to build computing infrastructure in space. He argued that these orbital data centers would eventually provide the funding needed for self-sustaining bases on the Moon and an industrial civilization on Mars. The merger serves as the structural foundation for Musk's vision of long-range expansion beyond our solar system.

This consolidation brings together the heavy-lift capabilities of SpaceX with the artificial intelligence expertise of xAI. By unifying these assets, Musk intends to address the accelerating demand for AI computing power. The goal is to move the core of AI processing away from Earth's surface to overcome physical and logistical limitations.

A Blueprint for Orbital Compute

Just two days before the merger, SpaceX submitted a filing to the Federal Communications Commission (FCC). The company requested permission to deploy up to one million satellites that function as orbital data centers. These satellites would be powered by solar energy harvested in space and would operate between 500 and 2,000 kilometers above the Earth.

The satellites are designed to use optical laser links to communicate with the existing Starlink network and various ground stations. SpaceX described this orbital approach as the most efficient way to meet the growing demand for AI power. They specifically pointed to the increasing strain that traditional data center construction places on terrestrial power grids as a primary driver.

The logistical backbone for this massive deployment is SpaceX's Starship rocket. Musk has projected a launch cadence capable of carrying payloads of up to 200 metric tons each. The company aims for a yearly recurring delivery of 100 to 200 gigawatts of orbital AI compute capacity rather than a cumulative total.

The Economics of Sunlight

The primary argument for moving data centers into space is based on energy efficiency. Terrestrial facilities currently compete with hospitals and factories for electricity, often facing interconnection queues of three to five years. In contrast, solar panels in space receive radiation with no atmospheric filtering or weather interference.

Musk stated in a February 2026 podcast that a single solar panel in orbit produces about five times the energy of an identical panel on Earth. Furthermore, satellites in sun-synchronous orbits face virtually no periods of shadow. This eliminates the need for heavy and expensive battery storage systems, which are currently a major cost factor in power architecture.

By removing the constraints of land use and local power grids, Musk contends that the unit economics of space-based compute become more favorable. However, the transition requires overcoming significant engineering barriers. These include the need for much larger solar arrays than currently exist and managing the latency introduced by satellite communication links.

Market Skepticism and Rivals

While Musk predicts cost parity within two to three years, independent analysts offer a more conservative timeline. A total cost of ownership model from SemiAnalysis suggests that orbital compute costs will remain more than four times higher than ground facilities in 2026. They estimate that parity might not be reached until around 2040, and only if launch prices drop significantly while ground costs continue to rise.

Other major tech players are already exploring similar orbital concepts. Google launched a prototype satellite in October 2026 as part of Project Suncatcher to test orbital solar power for AI workloads. Meanwhile, Blue Origin filed an FCC application in March 2026 for a network of satellites called Project Sunrise.

Despite the high costs and technical hurdles, the industry remains watchful of these developments. Some analysts suggest that while the timelines are optimistic, the move reflects a growing reality of terrestrial limitations. The race is now shifting toward who can successfully place gigawatts of compute above the clouds before the economics fully catch up with the ambition.