Via EdgeIR.comIn order to meet rising demand for its Claude AI models, Anthropic has teamed with Google and Broadcom for access to several gigawatts of next-generation TPU capacity likely available by 2027.

The partnership is part of Anthropic’s $50 billion investment in U.S. computing infrastructure, announced in November 2025.

In 2026, Anthropic’s revenues surged over $30 billion with over 1,000 business customers spending more than $1 million annually.

“This groundbreaking partnership with Google and Broadcom is a continuation of our disciplined approach to scaling infrastructure: we are building the capacity necessary to serve the exponential growth we have seen in our customer base while also enabling Claude to define the frontier of AI development,” says Krishna Rao, CFO of Anthropic. “We are making our most significant compute commitment to date to keep pace with our unprecedented growth.”

Claude AI lives on various hardware systems, such as AWS Trainium, Google TPUs or NVIDIA GPUs to maximize the performance and resiliency in each silo.

Amazon continues to be Anthropic’s main cloud provider, with Claude now on AWS, as well as Google Cloud and Microsoft Azure.

The expansion represents the company’s plan to increase infrastructure for accelerating customer growth and AI progress.

This suggests Claude is in a class of frontier AI models that will be available on the major cloud platforms around the world for use in various applications.

Anthropic’s multi-gigawatt TPU deal with Google and Broadcom highlights a shift to pre-secured, utility-scale AI infrastructure as the true competitive moat. It reinforces that access to power, silicon, and hyperscale capacity and not just models is now the defining factor in AI leadership, within an increasingly hybrid, multi-cloud ecosystem.
http://dlvr.it/TS898r