Amazon is investing heavily in further expansion of its cloud and AI infrastructure and plans to allocate significant capital to this over the next few years. The company plans to invest about $200 billion by 2026, with Amazon Web Services accounting for the bulk of that budget.
According to The Register, the explosive growth of AI applications means that the demand for computing power structurally exceeds the current supply.
According to CEO Andy Jassy (pictured), virtually all new capacity added by AWS is immediately purchased by customers. Limitations are primarily on the supply side, such as available energy, hardware, and data center location. Amazon has added nearly 4 gigawatts of additional capacity in the past year, highlighting the scale of its data center network. The goal is to further double this production capacity by the end of 2027.
Growth will primarily be driven by organizations that leverage AI intensively. More and more companies are moving their data, applications, and training processes to the cloud. Amazon expects AI to play a fundamental role in virtually all digital services and that new AI-driven applications will soon become the norm. This development has led to a long-term shift to cloud platforms such as AWS.
Large investments affect financial ratios. AWS currently has operating margins of approximately 35%, but Amazon expects this to fluctuate due to temporary increases in capital expenditures. However, the company says these investments are not speculative. Amazon bases its expansion plans on specific demand patterns and years of experience in capacity planning, with the goal of avoiding overcapacity and ensuring profitability.
Significant revenue growth on AWS
This number shows that AWS is still growing strongly. Fourth-quarter revenue reached $35.6 billion, a 24% year-over-year growth. For the full year, sales increased to $128.7 billion. Annual revenue based on current occupancy rates now exceeds $140 billion. Nevertheless, investors reacted cautiously to the announced investment, and the stock price plunged in after-hours trading.
AWS’s strategic advantage is the development of its own chips. Amazon is increasingly focusing on Trainium accelerators and Graviton processors to reduce dependence on external suppliers. These chips are used in large-scale AI environments, including Project Rainier, one of the world’s largest operational AI clusters. AI company Anthropic uses this infrastructure to train its models.
Trainium’s roadmap shows Amazon continuing its chip strategy. The current generation is almost fully booked, but subsequent versions promise significant performance improvements. This highlights Amazon as one of the most important providers of scalable and affordable AI infrastructure.
