SpaceX Nvidia Exclusivity Lifts Nvidia Outlook

SpaceX reported its first quarterly results as a public company on Aug. 3–4, 2026, and said it will build its AI infrastructure exclusively on Nvidia’s Vera Rubin architecture. This makes Nvidia the sole accelerator supplier for SpaceX’s ground and orbital compute plans.

Earnings and AI Capital Expenditure

SpaceX’s Q2 2026 report beat Wall Street expectations, highlighting an unusually large machine-learning investment program. The company recorded $18.7 billion in revenue in 2024 as context for its growth trajectory.

The filing showed AI compute capital expenditure (capex) of $15.8 billion in the quarter and $23.6 billion in the first half of 2026. Total Q2 capex reached $18.4 billion, mostly devoted to compute infrastructure.

SpaceX ended the quarter with 1.4 gigawatts (GW) of nameplate compute capacity and expects to exceed 2 GW by year-end 2026. Management outlined 2027 ambitions ranging from roughly 10 GW to as much as 20 GW of power-and-cooling capacity, while acknowledging projects could leave deployments closer to 15 GW.

Executives also presented long-range revenue goals, aiming for a $100 billion annualized run rate by year-end and an aspirational $1 trillion annual target by 2030.

Nvidia Exclusivity and Starmind AI1 Partnership

Elon Musk told investors SpaceX will build its AI infrastructure exclusively on Nvidia hardware, making Nvidia the sole AI accelerator supplier for both terrestrial data centers and planned orbital compute programs. Musk described the Vera Rubin architecture as the best AI computer and emphasized a close partnership with Nvidia.

SpaceX and Nvidia announced a joint program to design the Starmind AI1 satellite compute payload. Each Starmind satellite will house Rubin GPUs and Vera CPUs, with plans to deploy Nvidia’s NVL72 rack-scale systems on the ground and in orbit. Nvidia’s Space-1 Vera Rubin module can deliver up to 25 times the AI processing of an H100 GPU, with volume shipments expected to begin this fall.

Prototype testing for Starmind AI1 is scheduled for early 2027. SpaceX plans to begin launching Starmind satellites next year, with first full data-center satellites expected to launch late in 2027. The company has filed with the U.S. Federal Communications Commission (FCC) for authority to deploy up to one million orbital data-center satellites between 500 and 2,000 kilometers altitude, describing the planned constellation as a “petabit laser mesh.”

Large third-party compute agreements underpin demand. SpaceX’s contracts tied to Anthropic’s Colossus 1 involve roughly 300 megawatts of capacity, about 220,000 Nvidia GPUs, and payments near $1.25 billion per month through May 2029. A separate deal with Google grants access to about 110,000 Nvidia GPUs, with monthly payments near $920 million through June 2029. Together, these agreements are expected to add roughly $26 billion in annualized revenue, more than doubling the 2024 figure.

Analysts modeling demand scenarios suggest that even modest execution of SpaceX’s roadmap—2 to 3 GW of deployments in 2027—could translate into roughly $100 billion of incremental annual revenue for Nvidia and add about $2 to an EPS forecast of $13.36. More aggressive assumptions raise the possibility of substantially larger upside, including scenarios approaching $1 trillion in annual revenue.

Market commentary noted the competitive impact of SpaceX’s exclusivity pledge. Nvidia shares rose while rivals AMD and Intel declined, reflecting concerns about their AI data-center prospects.

If SpaceX follows its capex and deployment roadmap and maintains an exclusive supply relationship, it would create a sustained, large-scale demand stream that reallocates future GPU absorption toward Nvidia and reshapes hyperscale supply dynamics.