Architecture Shift
Major
High
90% Confidence
Anthropic Locks in Multi-Gigawatt Next-Gen TPU Capacity with Google and Broadcom
Summary
Anthropic has signed a new agreement with Google and Broadcom to secure multiple gigawatts of next-generation TPU capacity, expected online starting 2027. This expansion aims to power frontier Claude models and meet surging global customer demand. The partnership significantly expands Anthropic's $50 billion U.S. compute infrastructure commitment.
Key Takeaways
Anthropic deepens its partnership with Google Cloud and chip designer Broadcom to secure multi-gigawatt-scale next-generation TPU capacity, with delivery starting in 2027. This represents a major expansion of its $50 billion U.S. compute infrastructure investment pledge announced in November 2025.
Anthropic's CFO stated this move continues its 'disciplined approach to scaling infrastructure' to serve exponential customer growth and maintain frontier AI development. The company's run-rate revenue surpassed $30 billion in 2026, and the number of business customers spending over $1 million annually doubled to over 1,000 in less than two months.
Anthropic highlights its multi-hardware strategy, training and running Claude on AWS Trainium, Google TPUs, and NVIDIA GPUs to match workloads optimally. Amazon remains its primary cloud provider and training partner, and Claude is the only frontier model available across AWS, Google Cloud, and Microsoft Azure.
Anthropic's CFO stated this move continues its 'disciplined approach to scaling infrastructure' to serve exponential customer growth and maintain frontier AI development. The company's run-rate revenue surpassed $30 billion in 2026, and the number of business customers spending over $1 million annually doubled to over 1,000 in less than two months.
Anthropic highlights its multi-hardware strategy, training and running Claude on AWS Trainium, Google TPUs, and NVIDIA GPUs to match workloads optimally. Amazon remains its primary cloud provider and training partner, and Claude is the only frontier model available across AWS, Google Cloud, and Microsoft Azure.
Why It Matters
This signals a new phase in the AI infrastructure race, where leading model vendors secure next-gen custom chip capacity via massive, long-term agreements to ensure growth and shape the supply chain. Compute is evolving from a general-purpose resource to a strategic asset dominated by a few players, potentially reshaping the ecosystem balance between cloud providers, chipmakers, and AI companies....