Is Amazon the Real Winner of the 2025 AI Cloud Race?

Amazon Web Services (AWS) is demonstrating significant momentum in the AI cloud market, reaccelerating its growth with a 29% market share in Q3 2025, $33 billion in revenue (up 20.2% year-over-year), and a substantial $200 billion backlog. The company is strategically investing in rapid data center capacity expansion, developing advanced custom AI silicon like Trainium (a multi-billion dollar business with 150% quarter-over-quarter growth), and building a comprehensive AI ecosystem that includes platform services such as SageMaker and Bedrock, alongside new agentic AI tools. This full-stack approach, highlighted by a $38 billion deal with OpenAI, positions AWS to capture surging AI demand, even as competitors like Azure and Google Cloud report faster revenue growth.

Analysis

AWS exited Q3 2025 with a 29% share of the global cloud infrastructure market, generating $33 billion in revenue, a 20.2% year-over-year increase and its fastest growth since 2022. This reacceleration is supported by a substantial $200 billion backlog, providing significant near-term revenue visibility, despite competitors Azure and Google Cloud reporting faster revenue growth rates of 39% and 34% respectively. AWS's operating income climbed 9.6% year-over-year to $11.4 billion, reflecting a strong operating margin over 34%.

Amazon is aggressively expanding its AI infrastructure, adding over 3.8 gigawatts of data center power capacity in the past 12 months and aiming to double its 2022 capacity by 2027. Its custom silicon strategy is gaining traction, with Trainium2 already a multi-billion dollar business, growing 150% quarter-over-quarter in Q3, and Trainium3 expected to deliver 40% better price-performance.

The company is building a comprehensive AI ecosystem, including platform services like SageMaker and Bedrock, alongside new agentic AI tools such as Strands and AgentCore. The $38 billion deal with OpenAI for AI training and inference workloads on AWS underscores its strategic positioning. This full-stack, model-agnostic approach, coupled with custom silicon, aims to expand its enterprise customer base and reduce vendor lock-in.

Key Decisions for Investors

• Monitor AWS's continued reacceleration in revenue growth and its ability to convert the substantial $200 billion backlog into realized revenue.
• Evaluate the competitive landscape, specifically AWS's custom silicon advantage (Trainium/Inferentia) and its impact on customer acquisition and retention against rival cloud providers.
• Consider Amazon's long-term positioning in the AI infrastructure market, given its aggressive data center capacity expansion and comprehensive AI ecosystem development.