At the recent conference Re:Invent, Amazon announced partnership with AI company Anthropic to build the world’s largest artificial intelligence supercomputer.

This supercomputer will be five times the size of the cluster currently used by Anthropic’s most powerful model and, once completed, is expected to become the world’s largest AI training machine. Amazon said the project, called “ Rainer ”, will be equipped with hundreds of thousands of the latest AI training chips, Trainium2.

Matt Garman, CEO of Amazon Web Services (AWS), also revealed during the conference that Trainium2 will be widely available and will be specifically used to train edge AI in the Trn2UltraServer cluster. Compared to clusters using Nvidia graphics cards, the new AWS clusters will reduce costs by 30-40%. Although Amazon is the world’s largest cloud service provider, its competitors in the field of generative AI, such as Microsoft and Google, were previously ahead. However, this year, Amazon invested $8 billion in Anthropic and launched a series of tools through its AWS platform, Bedrock, to help businesses take advantage of generative AI.

Additionally, Amazon introduced its next-generation training chip, Trainium3, which is expected to be available to customers by the end of 2025, with performance expected to be four times higher than current chips. Industry experts noted that Trainium3 made significant optimizations to data transfer between chips, which is crucial for the development of large AI models. While Nvidia still dominates the AI ​​training space, Amazon’s innovations indicate competition is emerging in the market.

Amazon also plans to launch a series of tools to help its customers manage generative AI models, which are often expensive and unreliable. The new AWS service, Model Distillation, can generate smaller, cheaper models, while Bedrock Agents can create and manage AI agents to automate tasks. Garman mentioned that businesses will be particularly interested in Amazon’s recently launched tools, such as those that ensure the accuracy of chatbot results.

Amazon’s new validation tool, called “Auto Reasoning”, differs from similar products previously launched by OpenAI. It relies on logical reasoning to analyze the results of the models.

To achieve this, businesses must convert data and policies into a format suitable for logical analysis. This formalized reasoning approach draws on decades of application experience in areas such as chip design and cryptography. By combining various systems with automatic reasoning capabilities, businesses can create more complex applications and services.