SpaceX partners with Anthropic, expanding AI computing at Memphis data center
Anthropic won access to a Memphis supercluster with more than 220,000 Nvidia chips, while SpaceX landed a marquee AI tenant and a future investor story.

SpaceX handed Anthropic access to Colossus 1 in Memphis, Tennessee, one of the world’s largest AI data centers, giving the Claude maker a new base of computing muscle built around more than 220,000 Nvidia processors, including H100, H200 and GB200 accelerators.
Anthropic said the arrangement would substantially increase its computing resources and let it raise usage limits for Claude Pro and Claude Max subscribers. Bloomberg reported that Anthropic planned to tap more than 300 megawatts of capacity from the site, a scale that shows how much AI competition now depends on electricity, chips and data-center footprint as much as it does on model quality.
The timing matters for Anthropic’s product push. The company held its Code with Claude developer conference in San Francisco on May 6 and introduced dreaming, a research preview for Claude Managed Agents that reviews prior work between sessions, identifies patterns and updates memory files. Anthropic also removed peak-hours limit reductions for Claude Code Pro and Max accounts and raised API limits for Claude Opus models, signaling that its coding and enterprise products were running into demand constraints. Claude Code is Anthropic’s agentic coding system that reads codebases, makes changes across files, runs tests and delivers committed code.
For SpaceX, the deal offered more than revenue. A marquee AI tenant gives Elon Musk’s company a commercial proof point as it prepares for a planned public offering, and it ties a private-space infrastructure business to one of the fastest-growing software companies in the market. Anthropic has also expressed interest in partnering with SpaceX to develop multiple gigawatts of orbital AI compute capacity, a concept that fits Musk’s long-running push to move heavy compute into orbit or other off-world infrastructure.
The partnership also underscored how quickly AI infrastructure has become a strategic national issue. A single Memphis facility drawing more than 300 megawatts illustrates the pressure on local power grids, land use and transmission planning, while the Nvidia count shows how tightly the race remains tied to advanced semiconductor supply. Anthropic has moved quickly around that bottleneck, announcing agents for financial services on May 5 and a new enterprise AI services company on May 4, after earlier engineering and research posts on managed agents and long-running Claude with persistent memory.
The rapprochement was striking given Musk’s public criticism earlier this year, when he wrote on X that Anthropic “hates Western civilization.” Even so, the Memphis deal suggests the AI race is now being fought at industrial scale, where compute access can shape product speed, market share and the balance of power among the companies trying to define the next phase of the industry.
Know something we missed? Have a correction or additional information?
Submit a Tip

