SpaceX spent roughly $16 billion on AI infrastructure in the second quarter and reported $2.6 billion in AI revenue, up 213% quarter-over-quarter, while contracting $14.1 billion in cloud services agreements.

Musk avoided naming rival firms when drawing the ‘Little League’ comparison, instead spotlighting SpaceX’s own capabilities and partnerships.

Musk framed rockets and satellites’ engineering discipline as directly relevant to scaling terrestrial data centers and preventing failures in flight.

The remarks came on SpaceX’s first earnings call as a public company, amid competition from cloud giants—Amazon, Microsoft, Google—and a growing field of neocloud startups.

SpaceX plans to build its AI platform primarily on Nvidia GPUs, leveraging Nvidia’s Vera Rubin architecture and signaling a close partnership for substantial GPU access next year.

The company is expanding data-center capacity, ending the quarter with 1.4 GW nameplate compute and targeting over 2 GW by year-end, with a goal of about 15–20 GW by the end of 2027.

Musk argued SpaceX’s aerospace engineering edge gives it a competitive advantage in AI compute, likening rockets to a Yankees team facing a Little League squad to illustrate superiority.

Since acquiring xAI in February, SpaceX has maintained heavy capex, totaling $18.4 billion in the quarter, with expectations to stay around current levels for the next two quarters.