Anthropic's Strategic Partnership with Google and Broadcom for Enhanced TPU Capacity

We've embarked on an exciting new chapter by partnering with Google and Broadcom to secure multiple gigawatts of next-generation TPU capacity, set to be operational by 2027. This strategic move is designed to bolster our infrastructure, ensuring we meet the soaring demand for our Claude models globally. As Krishna Rao, our CFO, highlights, this partnership is a testament to our commitment to scaling infrastructure in line with our rapid growth. Our revenue has skyrocketed to over $30 billion, with more than 1,000 business customers now spending over $1 million annually—a figure that has doubled in just two months.

Most of this new compute capacity will be based in the United States, aligning with our previous pledge to invest $50 billion in American computing infrastructure. This collaboration not only strengthens our existing relationship with Google Cloud but also enhances our work with Broadcom. Our diverse AI hardware ecosystem, including AWS Trainium, Google TPUs, and NVIDIA GPUs, allows us to optimize performance by matching workloads to the most suitable chips.

Amazon remains our primary cloud provider, and we continue to collaborate closely with AWS on Project Rainier. Claude is uniquely positioned as the only frontier AI model accessible on the world's three largest cloud platforms: AWS, Google Cloud, and Microsoft Azure. This ensures that our customers have the flexibility and resilience they need for their critical operations.