Anthropic is expanding its collaboration with Google to leverage a substantial number of the tech giant”s artificial intelligence chips, valued at tens of billions of dollars. This initiative aims to enhance Anthropic”s AI capabilities as competition in the market intensifies.
Under the agreement revealed on Thursday, Anthropic will gain access to over one gigawatt of computing power, which is expected to become operational in 2026. This capacity will be utilized for training the next iterations of its Claude AI model, specifically on Google”s proprietary tensor processing units (TPUs), which have previously been designated for internal applications.
Anthropic selected TPUs for their advantageous price-performance ratio and efficiency, as well as the company”s prior experience in deploying its Claude models with these processors. This agreement underscores the ever-growing demand for chips within the AI sector, where firms are competing to create technologies that can either replicate or exceed human cognitive capabilities.
Google, owned by Alphabet, offers its TPUs for rental through Google Cloud, presenting an alternative to the currently limited supply of Nvidia chips. In addition to the chip access, Google will also provide various cloud computing services to Anthropic.
Meanwhile, rival OpenAI has recently secured multiple contracts that could cost over $1 trillion to acquire approximately 26 gigawatts of computing capacity, sufficient to power around 20 million U.S. households. Industry leaders indicate that acquiring one gigawatt of compute capacity can be roughly estimated at around $50 billion.
OpenAI, the developer behind ChatGPT, is heavily reliant on Nvidia graphics processing units and AMD AI chips to meet its increasing operational demands. Earlier this month, Reuters disclosed that Anthropic anticipates more than a twofold increase, and potentially a nearly threefold increase, in its annualized revenue run rate for the next year, driven by the swift adoption of its enterprise products.
The startup, which prioritizes AI safety, focuses on creating models tailored for business applications. Its innovations have contributed to a surge in coding startups, including the likes of Cursor.
