[New York = E-Daily Seong Joowon Correspondent] The growth engine of SpaceX, the world’s largest private space company, is shifting toward artificial intelligence (AI). At SpaceX, which grew through space launches and the Starlink satellite internet service, the AI business has become the company’s largest revenue source for the first time. CEO Elon Musk has set an ambitious target of up to $1 trillion (approximately 1,327 trillion won) in annualized revenue by the end of next year. His plan is to transform the company into a mega AI firm by pouring astronomical amounts of capital into AI computing infrastructure.

According to SpaceX’s third-quarter earnings report released on the 20th (local time), revenue from the Superintelligence (SI) division, which handles AI, surged 830% year-over-year to $7.56 billion (approximately 10.021 trillion won). This represents a more than ninefold increase in just one year. For the first time, it surpassed the $5.248 billion in revenue generated by the communications division, which includes Starlink. AI accounted for 51.9% of total revenue of $14.575 billion.
Changes were also evident in profitability. The AI division posted an operating profit of $431 million. The company explained that, excluding $1.2 billion in stock-based compensation expenses related to the acquisition of AI coding firm Cursor, operating profit would have reached approximately $1.6 billion. The company as a whole also recorded a net income of $4.404 billion, achieving its first quarterly profit since going public last June.
The rapid growth of the AI business is the result of the massive investments in computational infrastructure that Musk has been pursuing. SpaceX spent $19.593 billion on capital expenditures in the third quarter, of which $16.444 billion (approximately 84%) was allocated to the AI division. This amount is more than five times the combined investment in space launches and satellite communications.
The pace of investment is expected to accelerate further. Chief Financial Officer (CFO) Brett Johnson predicted during the earnings call that fourth-quarter capital expenditures would reach $35 billion to $40 billion—an increase of approximately 79% to 104% compared to the third quarter. The company also forecast first-quarter revenue for next year to be between $33 billion and $37 billion. This is more than double the revenue recorded in the third quarter of this year.

The long-term goal Musk presented is even more aggressive. He stated, “By the end of 2027, SpaceX’s annualized revenue could reach $500 billion to $1 trillion.” This is 5 to 10 times the company’s annualized revenue target of $100 billion for the end of this year. However, this figure does not represent actual revenue for the entire year next year, but rather the revenue growth rate projected as of the end of next year, annualized.
The key premise behind this forecast is the explosive expansion of AI computing capacity. As of the end of September, SpaceX’s nominal AI computing infrastructure capacity stood at 1.7 gigawatts (GW). Musk announced plans to increase this to 2.5–3 GW by the end of this year and to at least 13 GW by the end of next year. The plan calls for adding approximately 10 GW in the course of next year alone.
He identified power and cooling facilities, rather than semiconductors, as the biggest constraints on AI infrastructure expansion. He also outlined a goal to accelerate infrastructure development centered on the “Colossus 2” supercomputer cluster in Memphis, Tennessee, and to build a large-scale AI training cluster utilizing 440,000 of NVIDIA’s next-generation Vera Rubin-based computers by the end of the second quarter of next year.
SpaceX is supplying AI computing resources not only for its own model development but also to external companies. Musk explained that he is envisioning a plan where the ratio of in-house use to external leasing is roughly 50-50. CFO Johnson stated that, based on current prices, the payback period for investments in new computing facilities is less than one year. However, for this profitability outlook to be sustained, it must be supported by demand for AI services and leasing prices.

Musk’s official confirmation that day of merger discussions between Tesla and SpaceX is also linked to this business restructuring. He noted that special committees composed of independent directors from both companies have been established to discuss the possibility of a merger, emphasizing the synergies that would result from combining AI software, semiconductors, robotics, and space technology. However, he made it clear that no agreement had been reached on the terms of the deal.
SpaceX’s shift toward AI is already becoming evident in its revenue structure. However, there remains a significant gap to reach the goal of up to $1 trillion in annualized revenue by the end of next year. The key factor that will determine the company’s future value is whether it can secure power and semiconductors while managing massive capital expenditures, and whether it can translate its increased computing capacity into actual revenue and profits.