Amazon Launches $11 Billion Project Rainier AI Cluster

Amazon Web Services (AWS) has operationalized Project Rainier, an $11 billion AI supercomputing cluster distributed across multiple U.S. data centers. The project's primary hub in New Carlisle, Indiana, spans 1,200 acres and consists of thirty 200,000-square-foot facilities. The infrastructure currently utilizes nearly 500,000 proprietary Trainium 2 chips, with plans to exceed one million by the end of 2025.

The cluster is designed specifically to support the training and inference of Anthropic's Claude LLM models, backed by an $8 billion investment from Amazon in the AI firm. AWS claims the custom hardware provides 30% to 40% better price performance for AI tasks than Nvidia graphics cards. To manage energy and costs, the system employs vertical power delivery and a mix of air and direct-to-chip liquid cooling.

While analysts project the system could generate $6 billion in incremental revenue by 2026, the project has faced local criticism in Indiana over the loss of farmland and rising electricity costs. AWS CEO Matt Garman maintains that these proprietary chips outperform general-purpose alternatives, though some investors question if adoption will expand beyond Anthropic. Amazon is expected to provide further details on the rollout at its upcoming re:Invent conference.