Amazon Web Services topped analyst expectations as enterprises kept ramping up spending on cloud infrastructure and AI workloads, reinforcing the view that the current investment cycle in compute is far from over.
Amazon’s latest earnings release showed AWS growing roughly 20% year over year, with $33.1 billion in revenue across the first three quarters and operating income rising to $11.4 billion for the period, up from $10.4 billion a year earlier. Management highlighted a clear reacceleration in usage, pointing to robust demand in both core infrastructure and AI services. And Jassy confirmed that AWS has been rapidly adding capacity to meet that demand, underscoring that enough capacity was needed to expand its infrastructure footprint by more than 3.8 gigawatts over the past 12 months and is monetizing new capacity as it comes online.
Growth is being driven by customers committing to bigger, longer contracts to secure compute and power; the catch-up is wholesale AI utilization. The catalyst remains AI, and training and inference clusters are driving heavier, steadier consumption patterns than typical lift-and-shift migrations. AWS cited strong momentum from customers building generative AI applications and modernizing their data stacks, trends that increase both compute hours and storage throughput. It said that services that span Amazon Bedrock to managed databases to networking were seeing spillover benefits as AI projects moved from pilots to production.
AWS continues to press its silicon advantage with Trainium and Inferentia chips used to cut the cost of AI workloads. While Nvidia GPUs are very much in vogue, AWS’s broader offering of instances and orchestration tools helps companies triangulate faster/cheaper/better. This is pertinent, as CFOs are now looking closely at AI unit economics.
Capacity buildout and new regions to meet strong demand
To meet demand, AWS inaugurated a new infrastructure region in New Zealand, with three more rapidly following suit. While geographic reach and data sovereignty are important, the more pressing constraint is power. Adding 3.8 gigawatts in a year is a lift—a whole lot of hyperscale campuses—and speaks to long-lead investments in land, grid interconnects, and energy contracts. Trade groups like the Uptime Institute continue to raise a flag that power availability increasingly determines the rate of data center expansion. AWS can’t build quickly enough.
The message to customers is clear: its buildout ahead of your capacity needs is well in hand, be that elastic general-purpose compute, or cluster compute you’ve reserved for 2 years to do all your model training. A big part of the AWS sales story is now long haul.
Customer wins underscore breadth across industries
The firm highlighted net new agreements across sectors, including with AI-native organizations. For example, Perplexity chose AWS to support its enterprise product, and code assistant Cursor similarly did a deal with AWS in the quarter. These businesses join the millions of individual and team AWS customers and legacy firms revitalizing their ERP, analytics, and collaboration stacks. The present-cycle player isn’t only tech companies or startups.
Big-ticket cloud deals reset market expectations
The demand signals are everywhere throughout the industry. Reports uncovered an extensive, multi-year OpenAI–Oracle compute agreement totaling hundreds of billions of dollars over its lifetime, including an annual planned spend estimated at $30 billion on data center services. Google and Anthropic have announced a multi-year agreement valued in the tens of billions. The precise amounts and durations vary, but the facts are evident: AI leaders are preoccupying with their demand at a scale never before seen to them.
Skeptics concern about overbuild risk if AI adoption decelerates, hyperscalers claim that the backlog and waitlists for high-end compute represent numerous years of runway. The truth is most likely in the middle: ahead of some normalization and up likewise measuring a structural pre-gen AI epoch tempo.
Outlook for AWS and the broader cloud infrastructure market
AWS is still substantially out front in the cloud infrastructure marketplace, even as Microsoft and Google articulate their AI-fed development. Gartner expects double-digit growth in the cloud infrastructure marketplace to endure robust as firms focus on speeding up app modernization, data platforms, and AI pipelines. At the same time, Amazon is reconstituting sectors of its workforce while boosting the upward pressure on capital savings in AI facilities, a trade-off many major tech companies are making.
The short-term strategy for AWS is exceedingly plain: secure more power and capacity, change commitments to value, and hold margins in step via effectiveness progress and cited silicon. AWS outpaced Wall Street since customers continued to acquire power, not just words. As long as AI workloads remain compute-fat and businesses continue to move essential systems into its ecosystem, the market for Amazon looks to have a brilliant prospect.
Research on cardiovascular risk factors and lifestyle habits
Research shows that either long-term dietary habits that result in a high content of cholesterol or factors such as hypertension, low level of physical activities, long-term sedentary work, and personality features may promote atherosclerosis. Indeed, other research argues that the lack of a digital platform for converting proteins is a predisposition to atherosclerosis, as shown by the high prevalence 2500 g/l diet in the blood of homocysteine.
