Companies·3 min read·Anthropic

Anthropic, Google and Broadcom Team on Gigawatts of TPUs

Anthropic announced a partnership with Google and Broadcom on June 23 for multiple gigawatts of next-generation TPU capacity, coming online from 2027 and sited mostly in the U.S. — backing a run-rate that has climbed from $9 billion to over $30 billion.

GIGAWATTS OF NEW TPU CAPACITY Anthropic × Google × Broadcom · primarily U.S.-sited Anthropic Google Broadcom MULTIPLE GIGAWATTS OF TPUs · ONLINE 2027 $9B → $30B run-rate 1,000+ customers over $1M/yr $50B U.S. infra pledge BITSMINDS.COM
Share:

Anthropic on June 23 unveiled a partnership with Google and chipmaker Broadcom to bring multiple gigawatts of next-generation tensor processing unit (TPU) capacity online, with the first sites expected to power up in 2027. The build-out will be sited primarily in the United States and is pitched as the infrastructure needed to keep Claude at the frontier while serving demand that has outrun the company's existing footprint.

The scale of that demand is the headline justification. Anthropic says its run-rate revenue has surpassed $30 billion, up from roughly $9 billion at the end of 2025, and that more than 1,000 business customers are now spending over $1 million a year each — a cohort that doubled in under two months. The new capacity builds directly on the company's November 2025 commitment to invest $50 billion in American AI infrastructure, and it follows Google's pledge to put up to $40 billion into Anthropic earlier this year.

Notably, the deal deepens Anthropic's multi-silicon strategy rather than betting on a single vendor. The company says it deliberately spreads workloads across AWS Trainium, Google TPUs, and NVIDIA GPUs for better performance and resilience, with Broadcom co-designing the custom accelerator systems behind the TPU expansion. Amazon remains Anthropic's primary cloud provider, and the new TPU gigawatts sit alongside — not in place of — arrangements such as its access to SpaceX's Colossus cluster.

"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," said Krishna Rao, Anthropic's chief financial officer.

The announcement lands as the largest AI labs race to lock down compute years in advance, treating power and silicon as the binding constraint on how fast frontier models can improve. By committing to gigawatt-scale TPU capacity for 2027, Anthropic is effectively pre-buying its way through the next stage of that race — and signaling to investors, ahead of a widely reported IPO, that it can fund the bill.

Want AI news before everyone else?

The morning's most important AI stories, straight to your inbox. No fluff.

Related Articles