Anthropic said it will expand its use of Google Cloud's TPU chips t train its Claude models and meet demand for its services.
The expansion of the deal highlights how Anthropic is diversifying its infrastructure usage from Amazon Web Services. Anthropic and Google Cloud have been partners since 2023. That initial Google Cloud partnership gave Anthropic distribution through Vertex AI and Google Cloud Marketplace. Both AWS and Google are investors in Anthropic.
According to the companies, the expanded Google Cloud and Anthropic partnership will give the LLM provider access to more of a gigawatt of capacity coming online in 2026.
Anthropic said in a blog post:
"Anthropic’s unique compute strategy focuses on a diversified approach that efficiently uses three chip platforms–Google’s TPUs, Amazon’s Trainium, and NVIDIA’s GPUs. This multi-platform approach ensures we can continue advancing Claude's capabilities while maintaining strong partnerships across the industry. We remain committed to our partnership with Amazon, our primary training partner and cloud provider, and continue to work with the company on Project Rainier, a massive compute cluster with hundreds of thousands of AI chips across multiple U.S. data centers."
- Anthropic expands international presence, enterprise reach
- Anthropic lands $183 billion valuation, hits $5 billion annual revenue run rate
- Anthropic follows enterprise software industry playbook, hires Smith as commercial chief
In a statement, Anthropic said will have access to 1 million TPU chips and other Google Cloud services. Anthropic said the Google Cloud TPUs will give it efficiency gains with strong price performance.
For Google Cloud, the Anthropic win is significant. Google Cloud is also providing capacity to OpenAI. Google Cloud CEO Thomas Kurian said, "we are continuing to innovate and drive further efficiencies and increased capacity of our TPUs."
Constellation Research analyst Holger Mueller said:
"This is a big win for Anthropic and an even bigger one for Google Cloud. Anthropic continues its compute diversification, which is wise as it avoids Nvidia shortages and prices and gets access to the most efficient AI platform. Google can show that TPUs are not 'just' for Google workloads. And it's a win for customers since the deal allows Anthropic to be more cost competitive. Google can create further economies of scale for TPUs."
