AWS sales growth hits 37% as AI run rate reaches $25 billion

The acceleration in cloud revenue suggests that enterprise investment in generative AI infrastructure is translating into direct top-line gains for hyperscalers. By reaching a $25 billion run rate for its proprietary chips and AI services, Amazon is successfully reducing its reliance on external silicon providers while capturing high-margin workloads. This performance sets a high bar for competitors like Microsoft Azure and Google Cloud, indicating that the cloud market is entering a new phase of expansion driven by specialized hardware. Watch for whether AWS can maintain this 30%-plus growth trajectory as it faces tougher year-over-year comparisons in upcoming quarters.

Amazon has been aggressively expanding its custom silicon portfolio to differentiate AWS from rivals. In August 2025, Amazon announced that its Trainium2 chips were being deployed at scale for Anthropic's Claude model training workloads, marking one of the largest known commitments to proprietary AI accelerators in the cloud market. The company's Graviton processors, now in their fourth generation, have also gained traction for general-purpose compute, with AWS claiming price-performance advantages of up to 40% over comparable x86 instances. This vertical integration strategy directly supports the margin expansion that underpins the 37% revenue growth figure investors are watching.

On the competitive front, Microsoft Azure and Google Cloud are both posting strong AI-driven growth that pressures AWS to maintain its acceleration. Microsoft reported Azure revenue growth of 39% year-over-year in its fiscal fourth quarter ending June 2025, driven largely by AI services contributing more than 16 percentage points of that growth. Google Cloud posted 32% year-over-year revenue growth in Q2 2025, with CEO Sundar Pichai noting that the cloud backlog exceeded $100 billion for the first time. These numbers establish that the hyperscaler AI infrastructure race is intensifying, and AWS's 37% growth, while impressive, represents a narrowing lead over Azure in percentage terms.

The broader capital expenditure environment provides additional context for evaluating AWS's trajectory. Amazon disclosed plans to spend more than $100 billion in capital expenditures during 2025, the majority directed at data center buildouts for AI workloads. That figure exceeds the combined capex guidance of most other hyperscalers and signals Amazon's confidence that demand for AI compute will sustain current growth rates. Meanwhile, analysts at Morgan Stanley estimated that the total addressable market for AI infrastructure spending would reach $1.5 trillion by 2028, providing a long runway for AWS to convert its chip investments into recurring revenue streams.