Anthropic in discussions over using Microsoft AI chips, says report

Anthropic is considering adopting Microsoft AI chips in a major update. The Artificial Intelligence startup is

in talks to rent servers powered by Microsoft-designed chips to meet rising demand for its AI services, The Information reported on Thursday, in what would be a major boost for ‌the tech giant's in-house chip efforts.

The information report suggests Claud's creators are still exploring the idea and did not lead to any agreement yet.

Shares of the Windows maker, which are down about 10% for the year, rose 1% after the news, while Microsoft and Anthropic did not immediately respond to the matter.

Landing Anthropic will be a major win for Microsoft's push to emulate the custom chip success of Alphabet and Amazon, which have built significant businesses renting out processors initially meant for internal use to AI startups.

Demand has grown for these custom chips as companies seek alternatives to Nvidia's costly and low-in-supply AI processors. Anthropic has emerged as a key customer for them by striking deals with companies including Amazon and ‌Google.

Microsoft has been deepening its ties with Anthropic in recent months by integrating its models into products, including the Copilot AI assistant, as its long-standing partnership with OpenAI loosens and it reduces its reliance on the ChatGPT maker.

The company in January unveiled the second generation of its in-house AI chip, "Maia 200", and software tools that targeted one of Nvidia's biggest competitive advantages with developers.

The Maia 200 is made by Taiwan Semiconductor Manufacturing Co. using 3-nanometer chipmaking technology and uses high-bandwidth memory chips, albeit an older and slower generation than Nvidia's forthcoming "Vera Rubin" chips.

Microsoft has loaded its chip with a significant amount of what is known as SRAM, a type of memory that can provide speed advantages for chatbots and other AI systems when they field requests from a large number of users.