How Amazon became one of the world’s top chip companies in a decade
Amazon’s custom silicon business exceeded a $25 billion annual revenue run rate—here’s what that means for customers and the future of AI.
Our chips are purpose-built for cloud and AI workloads, delivering the price-performance, scale, and efficiency that let more builders push AI forward.
Amazon has invested over a decade in custom silicon, designing chips in lockstep with the teams building frontier AI models. The result: hardware that arrives optimized for real customer workloads, manufactured and deployed at unprecedented speed and scale.
Learn about Amazon's custom chip innovations
In Q1, Amazon’s chips business saw nearly 40% quarter-over-quarter growth—and it has momentum.
Tour the micro metropolis of a Trainium chip, where calculations run 24/7 and data commutes at light speed.
From co-development with AI model teams to automation-ready manufacturing, Amazon's custom chips, Trainium and Graviton, are designed to perform at scale from day one.
Startups and enterprises are choosing Trainium for its industry-leading price-performance, training ambitious models and running inference more efficiently.
Agentic AI systems that reason, plan, and take action are driving massive central processing unit (CPU) demand. Graviton delivers up to 40% better price-performance, and 98% of top EC2 customers use it.
AWS powers the work of “inference” with custom chips, smart routing systems, and purpose-built infrastructure—making AI faster and more affordable.
Latest chips news
AWS revenue growth accelerated to 36.7% year over year—the fastest in 18 quarters—with AI and core services driving each other's growth.
At VivaTech 2026 in Paris, Peter DeSantis argued we're at the starting line of AI development with the most important breakthroughs still ahead.
The most powerful and efficient CPU Amazon has ever built is now available to all customers, with up to 25% better performance.
University researchers are using Amazon’s Trainium chip to push the boundaries of their work—and making the chip better for all developers in the process.