Here are the key takeaways for chip stocks from SpaceX's S-1 Filing By Investing.com

Investing.com -- SpaceX's S-1 filing contains several significant data points for semiconductor investors, with Nvidia emerging as the primary chip supplier to the company's AI clusters and Intel named as a key partner in a major new fabrication facility, according to Wolfe Research.

In a note from analyst Liz Pate, Wolfe said SpaceX operates two AI clusters, with Colossus 1, a roughly 130-megawatt cluster running approximately 100,000 H100 GPUs, and Colossus II, currently a 430-megawatt site running around 110,000 GB200 and 110,000 GB300 accelerators.

Total planned Colossus II capacity includes an additional 400 megawatts of compute and approximately 220,000 additional GB300 processors, bringing total planned capacity across both sites to 1 gigawatt.

On capital expenditure, Wolfe noted that SpaceX's first-quarter 2026 AI spend of $7.7 billion implies annualized spending of more than $30 billion, up sharply from $12.7 billion in fiscal 2025, with the filing suggesting further acceleration ahead.

Wolfe highlighted Terafab, a large-scale chip fabrication facility announced in the first half of 2026, as another key data point, with Intel named as a partner citing its expertise across design, fabrication and packaging.

The facility is said to be targeting an annual production of approximately 1 terawatt of compute hardware, though Wolfe noted there are currently no specific disclosures on timeline or capital expenditure.

Perhaps most notably for the longer-term outlook, SpaceX is aiming to launch AI compute satellites as early as 2028, targeting 100 gigawatts in satellite constellation-driven compute capacity annually, an ambition Wolfe said will mandate thousands of annual launches.