Google Will Be Paying SpaceX $920 Million Monthly for AI Compute

SpaceX has signed a major infrastructure agreement to rent artificial intelligence compute capacity to Alphabet’s Google for $920 million a month, as the company seeks to highlight new revenue streams ahead of its planned initial public offering.

The 32-month contract will see Google use about 110,000 Nvidia graphics processing units (GPUs), along with central processors, memory and other components housed in SpaceX data centers. The agreement runs from October this year through June 2029, with capacity expected to ramp up through September at a reduced rate.

The deal follows SpaceX’s February merger with xAI, the artificial intelligence startup founded by Elon Musk, in a transaction that valued the combined entity at $1.25 trillion. SpaceX is now preparing to go public next week at a valuation above $1.75 trillion.

A Google Cloud spokesperson told Reuters via email that the agreement was struck “to ensure we have bridge capacity to meet surging customer demand for our agent platform, Gemini Enterprise, which has been even higher than we expected.”

For Alphabet, the deal underscores intensifying capital expenditure across major cloud providers as they race to expand AI infrastructure. The company recently raised its capital expenditure forecast to between $180 billion and $190 billion and said this week it plans to sell $85 billion in stock to fund expansion efforts. Alphabet was also an early investor in SpaceX, backing the company in 2015 when it was valued at $12 billion.

The agreement enables SpaceX to monetize data center infrastructure originally built to support xAI’s Grok chatbot. In its IPO prospectus, SpaceX said capital expenditures in the first quarter more than doubled year-over-year to $10.1 billion, with $7.7 billion directed toward AI. Its AI segment recorded an operating loss of $2.5 billion on revenue of $818 million.

“We believe our compute infrastructure and related strategy provides us with substantial flexibility in how we allocate and monetize capacity,” SpaceX said in its prospectus section on third-party compute service agreements.

The contract includes strict performance conditions. If SpaceX fails to “deliver access to the committed amount of GPUs by September 30, 2026,” Google can terminate the agreement immediately or accept reduced capacity at a lower fee after a one-month grace period. After 2026, either party may terminate the deal with 90 days’ notice.

The agreement marks SpaceX’s second major infrastructure deal since its merger with xAI, following a separate arrangement last month with Anthropic to use capacity at the Colossus 1 data center in Memphis, Tennessee.

The move places SpaceX in direct competition with so-called “neocloud” providers such as CoreWeave and Nebius. Shares of those companies rebounded after an initial selloff following the announcement.

The arrangement also reflects a reversal of roles between the two firms. Five years ago, Google agreed to supply computing and networking resources to SpaceX to support its Starlink satellite internet service.

At the time, Thomas Kurian, chief executive of Google Cloud, said: “They chose us because of the quality of our network and the distribution and reach of our network.”

Stay ahead of the stories shaping our world. Subscribe to Impact Newswire for timely, curated insights on global tech, business, and innovation all in one place.

Dive deeper into the future with the Cause Effect 4.0 Podcast, where we explore the ideas, trends, and technologies driving the global AI conversation.

Got a story to share? Pitch it to us at info@impactnews-wire.com and reach the right audience worldwide