Anthropic, an artificial intelligence company, has signed a multibillion-dollar agreement with Google to purchase additional processing capacity required for the startup’s chatbot, Claude.
Anthropic stated on Thursday that the agreement, which is “worth tens of billions of dollars and is expected to bring well over a gigawatt of capacity online in 2026,” will grant it access to up to one million of Google’s AI computer processors.
According to the U.S. Energy Information Administration, a gigawatt is equivalent to around 350,000 houses when referring to a power plant.
Tensor Processing Units, or TPUs, are the name Google gives to its proprietary AI hardware.
Additionally, Anthropic’s AI systems are powered by Nvidia chips and Amazon’s cloud computing division, which served as both its main cloud provider and its first significant investor.
Last month, the privately held Anthropic, established by former OpenAI executives in 2021, valued itself at $183 billion after securing an additional $13 billion in funding.
To attract business clients who use it to assist with coding and other duties, its AI assistant, Claude, competes with ChatGPT from OpenAI and others.
