Anthropic has partnered with Elon Musk's SpaceX to access the Colossus 1 artificial intelligence data center in Memphis, Tennessee [1, 2].
This agreement provides the AI startup with the massive compute capacity necessary to meet the surging demand for its Claude AI services. The deal marks a significant shift in the operational landscape for Anthropic, which requires immense processing power to train and maintain its large language models.
An Anthropic spokesperson said the company will utilize the full computing power of the Colossus 1 facility [3]. The center houses more than 220,000 Nvidia processors and is expected to provide 300 megawatts of new capacity within a month [3].
The partnership was announced May 6, 2024 [1, 4]. While some reports describe the arrangement as a rental agreement, others characterize it as a lease of one of SpaceX's most valuable assets [1, 2].
Financial details of the deal vary across reports. One source said Musk is receiving $4 billion to serve as Anthropic's data landlord [4]. Other reports on the partnership did not mention a specific monetary amount [1].
The collaboration comes despite previous public friction. Elon Musk once said the AI lab was "evil, misanthropic, and that the AI lab hated Western civilization" [4]. Despite these past comments, the two entities have reached an agreement to monetize SpaceX's super-computer assets while scaling Anthropic's infrastructure [2, 3].
“SpaceX will give Anthropic access to its massive Colossus 1 artificial intelligence data center”
This deal highlights the critical bottleneck in the AI race: physical compute hardware. By securing a massive cluster of Nvidia processors and significant power capacity, Anthropic can accelerate its development cycle. Furthermore, the partnership demonstrates that the immense demand for AI infrastructure can override personal or ideological conflicts between industry leaders.
