Andy Jassy says Amazon's Nvidia competitor chip is already a multi-billion-dollar business

Core Insights - Amazon's AI chip, Trainium, is positioned to compete with Nvidia's dominance in the AI chip market, with significant revenue potential for companies that can capture even a small share of this market [1][2] Company Developments - Amazon announced the next generation of its AI chip, Trainium3, which is four times faster and more power-efficient than Trainium2 [2] - The Trainium2 chip has achieved substantial traction, generating a multi-billion-dollar revenue run-rate, with over 1 million chips in production and more than 100,000 companies utilizing it [3][4] - Amazon's AI chip is gaining popularity among cloud customers due to its compelling price-performance advantages compared to other GPU options [4] Customer Insights - Anthropic is a key customer contributing significantly to Trainium's revenue, with over 500,000 Trainium2 chips deployed for their Project Rainier, aimed at developing advanced AI models [5][7] - Amazon is a major investor in Anthropic, which has made AWS its primary model training partner despite also using Nvidia's chips on Microsoft's cloud [8][9] Competitive Landscape - Only a few U.S. companies, including Google, Microsoft, Amazon, and Meta, possess the necessary engineering capabilities to compete effectively with Nvidia in the AI chip market [10] - Amazon plans to develop Trainium4 to interoperate with Nvidia's GPUs, which may either help capture more business from Nvidia or reinforce its existing market dominance [11]