Elon Musk has announced that SpaceX will exclusively use NVIDIA technology for its computing infrastructure, with the company expecting to reach over 2 gigawatts of computing power by the end of the year.

The decision underscores SpaceX's deepening reliance on NVIDIA's hardware for its artificial intelligence compute buildout. Recent data indicates that SpaceX's infrastructure already includes over 220,000 NVIDIA GPUs, reflecting a significant investment in NVIDIA's technology for large-scale AI applications. The commitment deepens a pattern across Musk's companies—xAI, founded by Musk in 2023, has similarly built its Grok large-language-model development on NVIDIA GPU clusters, and Musk has publicly discussed plans for large-scale AI infrastructure under that venture as well.

NVIDIA, founded in 1993 and headquartered in Santa Clara, California, is the leading designer of graphics processing units (GPUs) and has become the dominant supplier of accelerators for AI training and inference workloads worldwide. Its data center GPUs, including the Hopper and Blackwell architectures, are widely used by major cloud providers, research institutions, and enterprises engaged in large-scale model development. Demand for these systems has outpaced supply for several product cycles, with major cloud providers—Amazon, Microsoft, Google, and Meta among them—collectively allocating tens of billions of dollars to NVIDIA-based infrastructure.

SpaceX, founded by Musk in 2002, is best known for its Falcon rockets, Dragon spacecraft, and the Starlink satellite internet constellation. The company has been expanding its involvement in AI and autonomous systems across multiple programs, including autonomous rocket landing operations and Starlink's satellite collision-avoidance maneuvering, both of which rely on machine-learning components.

The scale of SpaceX's GPU deployment—surpassing 220,000 units—places it among the largest privately operated AI compute clusters in the world, rivaling or exceeding the GPU footprints publicly disclosed by the largest publicly traded technology companies. For context, building out infrastructure of this magnitude requires substantial capital expenditure on chips, networking equipment, cooling systems, and power delivery, with a 2-gigawatt power footprint comparable to the energy consumption of a major metropolitan data center hub.

The announcement is perceived as a substantial endorsement of NVIDIA's hardware platform. Market participants have noted that prediction-market pricing reflects an increased likelihood of NVIDIA becoming the world's largest company by market capitalization by August 31.

Observers are expected to monitor NVIDIA's upcoming financial reports and forward guidance, as well as any further details regarding AI-infrastructure orders, for additional insight into demand trends. Broader developments in AI compute demand and potential regulatory changes may also influence NVIDIA's trajectory in the period ahead.