Anthropic and Elon Musk’s SpaceX said Wednesday that the two entities have signed an agreement for Anthropic to exploit computing resources from the xAI data center in Memphis, Tennessee. It’s the latest development in an industry struggling to find enough computers to run elaborate AI software. SpaceX and xAI were previously separate companies, but earlier this year they merged. The combined entity, also owned by Musk, is called SpaceXAI.
Anthropic executives made the announcement onstage at the company’s annual developer conference in San Francisco. SpaceXAI too publish a blog post shares more details about the deal under which Anthropic will draw power from xAI’s Colossus 1 supercomputer.
The collaboration comes at a key time for SpaceXAI, which plans to go public as early as next month. A relationship with a leading artificial intelligence lab could boost SpaceX’s credibility as it entices investors into the potential goldmine of creating more data centers, including in space.
Space XAI in particular he said in its blog post that Anthropic has “expressed interest” in collaborating to develop “orbital artificial intelligence computing capability” – essentially data centers in space. Having Anthropic as a potential customer could support SpaceXAI escalate investor confidence that there will be buyers for its super-expensive supercomputing project.
Earlier this year, Musk criticized Anthropic’s artificial intelligence models in the post on Xcalling the company’s policies “misanthropic” and “evil” and maintaining without evidence that AI models showed racial and sexual bias.
Now he has changed his tone significantly, writes with X that he “spent a lot of time last week with senior members of the Anthropic team to understand what they are doing to ensure that Claude is good for humanity, and he was impressed.”
Musk’s SpaceXAI first launched Colossus 1 at the former Electrolux facility in 2024. The company says it is one of the largest and fastest AI supercomputers in the world. It includes approximately 220,000 Nvidia GPUs, ranging from the Nvidia H100 and H200 to the newer GB200 chips. The company likes brag that he built a supercomputer in 122 days.
Emissions from gas turbines at the data center have been the source of numerous complaints from local residents. Last month, environmental protesters demonstrated at a nearby SpaceX investor meeting ahead of its initial public offering.
Anthropic plans to exploit its newfound computing power to “directly increase the performance of Claude Pro and Claude Max subscribers,” the two companies said. The deal will give Anthropic access to more than 300 megawatts of up-to-date power, or approximately 220,000 Nvidia GPUs. In recent months, software developers have increasingly complained about speed limits and disruptions to Claude Code services due to physical limitations of available computing resources.
As more and more people exploit services like OpenAI’s Codex and Anthropic’s Claude Code – sometimes running coding programs and agent tasks for hours on end – these services get bogged down. According to Anthropic, the average developer now spends at least 20 hours a week working with Claude Code.
Yesterday Information reported that Anthropic has committed to spending $200 billion on Google’s AI cloud services and TPU chips. As of 2023, Anthropic also uses Amazon cloud computing services and AI chips. Last month Anthropic involved “over $100 billion over the next ten years [Amazon] technologies.” According to the Information report, deals with Anthropic and OpenAI now account for “more than half of the $2 trillion in backlogs at major cloud providers” such as Amazon, Microsoft and Google.
Paresh Dave and Zoe Schiffer contributed to this report.