In September, Amazon said it would invest up to $4 billion in Anthropic, a San Francisco start-up specializing in artificial intelligence. Amazon aims to build its A.I. using specialized computer chips designed by the start-up, in order to create a viable competitor to chipmaker Nvidia.
The big tech companies depend on Nvidia to build chatbots and other A.I. systems. The company, however, has not been able to keep up with the demand, prompting Amazon, Google, Meta, and Microsoft to build their own A.I. chips. They hope to control costs, eliminate chip shortages, and eventually sell access to their chips to businesses using their cloud services.
The move to build their own chips comes as a result of the tension between these companies and Nvidia. Despite continuing partnerships, Amazon, Google, and Microsoft are moving into Nvidia’s business. Last year, Nvidia launched its own cloud service and is supplying chips to cloud providers competing with the big three tech giants.
The A.I. chip market is projected to double, reaching roughly $140 billion by 2027. Venerable chipmakers like AMD and Intel are also moving into the A.I. chip market.
Most software systems used to train A.I. technologies were tailored to work with Nvidia’s chips, which makes it difficult and time-consuming for companies to switch to new chips.
Google has laid a strong foundation with the introduction of its tensor processing unit, while Amazon and Meta have announced plans to work on A.I. chips tailored to their needs.
Nvidia’s position in the A.I. chip market remains strong due to the high performance of their chips, and it will continue to improve their speed in the future.