Anthropic Joins OpenAI, Google in AI Chip Race as Compute Crunch Deepens

Anthropic is building an internal chip-design team to develop custom processors for its Claude artificial intelligence models.

Anthropic is assembling a team of over 20 engineers to design custom processors for its Claude artificial intelligence models. The company aims to reduce its reliance on suppliers like NVIDIA and AMD, which have struggled to meet demand for AI-specific hardware. Anthropic's chip-design team will be led by a former Google engineer, and the company has already begun recruiting experts in digital design and verification. The custom processors are expected to be designed in-house within the next 12-18 months.

The development of custom AI processors by Anthropic and its competitors will directly impact the cost of cloud computing services used by businesses and individuals. As AI models become more complex, they require more computing power, which can drive up costs for companies like Amazon Web Services and Microsoft Azure. If Anthropic's custom processors are successful, they could help reduce the cost of running AI workloads, making it more affordable for businesses to adopt AI technologies. This could lead to a decrease in the prices of services that rely on AI, such as virtual assistants and image recognition software.

The move by Anthropic to develop custom AI processors is part of a larger trend in the tech industry, where companies are seeking to control their own hardware destiny. Google and OpenAI have already made significant investments in custom AI hardware, and other companies like Facebook and Microsoft are also exploring similar strategies. The development of custom AI processors is a response to the growing demand for AI computing power, which has led to a shortage of available hardware and driven up costs. Insiders know that the current hardware supply chain is not equipped to handle the rapid growth of AI adoption.

In the next quarter, Anthropic is expected to announce a partnership with a leading semiconductor manufacturer to produce its custom AI processors. The company will also release a whitepaper detailing the architecture and performance of its custom processors. A surprising detail is that Anthropic's custom processors will be designed to be compatible with existing AI software frameworks, making it easier for developers to transition to the new hardware. This compatibility will be a key factor in determining the success of Anthropic's custom processors in the highly competitive AI hardware market.