AWS Revenue Jumps 39% to $45.98 Billion, Fastest Growth Since 2021

Amazon Web Services generated $45.98 billion in third-quarter revenue, up 39% year over year and above the $45.29 billion consensus among analysts polled by StreetAccount, the company said in a statement. The growth rate was the highest for the cloud unit since the fourth quarter of 2021 and an acceleration from roughly 37% in the second quarter.

The AWS segment produced $17.77 billion in third-quarter operating income, up 55% and above the $16.97 billion StreetAccount consensus. The unit now accounts for almost 23% of Amazon's total revenue, a record share.

"Our core compute and storage revenue grew at their fastest rates in several years," Amazon CEO Andy Jassy said on a conference call with analysts.

Why the quarter matters for the cloud market

Amazon remains the top cloud provider by revenue, but the gap with its rivals is narrowing on paper. On Wednesday, Microsoft published its first earnings report that disclosed Azure cloud revenue in dollars. Azure is now 72% the size of AWS, up from 58% two years ago.

Alphabet reported the same day that Google Cloud, the segment containing cloud infrastructure and Google Workspace subscriptions, more than doubled in the third quarter.

All three major cloud vendors are racing to build data centers that can handle surging artificial intelligence demand. AWS used the quarter to announce plans to expand its portfolio of Nvidia graphics processing units capable of running AI workloads, deepening its reliance on the chipmaker's hardware alongside its own silicon.

The company also acquired data analysis startup DuckLabs and signed deals with Meta and OpenAI, according to the statement. Those agreements tie AWS more closely to two of the most prominent buyers of AI computing capacity.

What did Jassy promise investors?

Jassy gave analysts an unusually specific long-range forecast for the unit. He tied it to product velocity, enterprise demand for both conventional and AI workloads, and the volume of data center capacity Amazon adds each year.

"With our continued pace of innovation, new core and AI functionality, enterprises wanting to run their core and AI workloads in AWS, and the amount of capacity that we're adding each year, we believe that AWS can grow north of 40% in each of 2027 and in 2028, and be a trillion-dollar revenue business in the next five to 10 years," Jassy said.

That projection implies sustained acceleration from the current 39% rate for at least two more years, and it sets a concrete benchmark investors can test against quarterly results. A trillion-dollar annual revenue figure for a single cloud division would exceed the current revenue of most companies in the S&P 500.

How does the Trainium strategy change Amazon's chip business?

Amazon also said it will copy a page from Google's playbook. Google has been increasing its cloud revenue by making its tensor processing units for AI available outside its own data centers. Amazon now intends to do the same with Trainium, its chips for training AI models, starting in late 2027.

"With Trainium demand so strong, customers continue encouraging us to make it available outside of AWS. We intend to do so late in 2027," Jassy said.

The move would let customers run Trainium chips in their own facilities rather than only inside AWS data centers. It positions Amazon's custom silicon as a product line in its own right, competing for AI training workloads against Nvidia's GPUs and Google's TPUs on customer premises as well as in the cloud.

The late-2027 timing gives rivals a long runway to respond, and it leaves most of the near-term AI training market to Nvidia-based capacity such as the GPU portfolio AWS just expanded.

What is the competitive picture?

AWS predates its two main rivals. Its first services appeared in 2006, years before Microsoft launched Azure and Google built out the Google Cloud Platform. That head start helped Amazon hold the revenue lead it still commands.

Microsoft's new dollar disclosure of Azure revenue makes the standings easier to compare for the first time. Azure at 72% of AWS size, against 58% two years ago, quantifies a closing gap that previously required estimation from percentage growth figures.

The stakes extend beyond the three vendors' earnings reports. Cloud capacity has become the physical substrate of the AI industry: the amount of compute available from AWS, Azure and Google Cloud effectively sets how much model training and inference the sector can perform. Jassy's forecast of north-of-40% growth in 2027 and 2028, combined with the Trainium expansion, signals Amazon expects that demand to keep outrunning supply for years.

For Amazon shareholders, the quarter shows the cloud unit delivering both growth and profitability at once: 39% revenue expansion alongside a 55% jump in operating income. For the AI industry, the Meta and OpenAI agreements plus the expanded Nvidia GPU portfolio indicate where the largest AI workloads are being placed. The test for the next two years is whether AWS can hold acceleration while Microsoft and Google Cloud grow faster from smaller bases.