NEW YORK, United States — SpaceX is exploring plans to manufacture its own artificial intelligence chips, a move that could reshape the global semiconductor landscape as the company prepares for a projected $1.75 trillion initial public offering.
According to a Reuters review of excerpts from the firm’s S-1 filing with the U.S. Securities and Exchange Commission, the Elon Musk-led company has identified “manufacturing our own GPUs” among its major capital expenditure priorities.
The filing outlines risks and investment plans ahead of a public listing expected later this year.
The disclosure signals an aggressive expansion into one of the most complex segments of the technology sector—producing graphics processing units (GPUs), which are critical for training and running artificial intelligence systems.
The effort builds on collaboration between SpaceX, its artificial intelligence arm xAI, and Tesla to develop “Terafab,” an advanced chip manufacturing facility planned in Austin, Texas.
Chief Executive Elon Musk has said the plant will produce chips for electric vehicles, humanoid robots, and space-based data infrastructure, though specifics remain limited.
Industry analysts note that the initiative could position SpaceX against established semiconductor leaders such as Nvidia, which dominates GPU design, and Alphabet, whose Google division develops tensor processing units (TPUs) tailored for AI workloads.
AI models, including those powering chatbots like Claude, rely heavily on such specialized chips.
However, manufacturing advanced chips presents steep technical and financial challenges. Much of the global industry relies on TSMC for fabrication, leveraging years of investment, precision engineering, and experience producing billions of chips for clients, including Apple.
SpaceX’s filing underscores these hurdles, warning investors about potential supply constraints. “We do not have long-term contracts with many of our direct chip suppliers,” the company said, adding that it may continue to depend on third-party providers for critical computing hardware.
The company also indicated uncertainty around the timeline and feasibility of its Terafab project. It remains unclear whether partners such as Intel will lead fabrication processes within the facility.
Musk recently told Tesla analysts that Intel’s next-generation “14A” manufacturing process could be mature enough to support the initiative.
