SpaceX Targets 13 GW of AI Computing Capacity by 2027 as Demand Surges Beyond Satellite Business

The Quick Take

• SpaceX is expanding AI computing infrastructure plans to 13 GW by 2027, with UBS projecting substantial investment and revenue potential from leased compute.
• Strong demand for cloud computing power is driving SpaceX to scale data infrastructure beyond its core satellite and launch business lines.
• The pivot underscores how space-tech companies are competing for AI infrastructure spending alongside established hyperscalers at massive scale.

SpaceX decision to target 13 gigawatts of AI computing capacity by 2027 marks a strategic expansion beyond its core launch and satellite businesses. UBS projection of substantial investment and revenue potential reflects growing institutional recognition that AI workload demand is outpacing existing hyperscaler capacity and creating room for non-traditional providers. The move parallels aggressive datacenter buildouts by Microsoft, Google and Amazon but is notable because it comes from a company whose primary moat is satellite-based global connectivity infrastructure rather than traditional cloud services.

The market implication for established cloud and AI-chip players is mixed in important ways. SpaceX as a buyer of GPUs and high-bandwidth networking equipment is additive demand for Nvidia, Broadcom and Arista revenues. As a provider of leased compute capacity it adds supply-side competition to AWS, Azure and GCP over a multi-year horizon. Indian IT services firms with AI infrastructure deployment exposure including Infosys, Wipro and HCL Tech could benefit as new entrants like SpaceX require experienced systems integration partners to operationalize large-scale compute deployments.

Investors tracking the AI infrastructure buildout should monitor UBS detailed investment-case publication for SpaceX, expected to quantify capital expenditure, power sourcing strategy and target customer segments. The 13 GW figure if realized would rank SpaceX among the top five global AI compute providers by capacity, making this announcement a meaningful data point in assessing long-term AI supply-demand equilibrium and the durability of current GPU pricing. The announcement also validates continued private capital formation in AI infrastructure outside traditional hyperscaler structures.

Synthesized from 1 source.