From a $100 billion infrastructure commitment to launching a slew of AI offerings, here are the 10 biggest news stories from AWS this year that you need to know about.
Amazon Web Services captured headlines throughout the first half of 2025 thanks to AWS’ innovation engine, channel partner enhancements and billions of investments to build out its AI and cloud infrastructure on a global basis.
“Our AI revenue is growing at triple-digit year-over-year percentages and represents a multibillion-dollar annual revenue run rate,” said Amazon CEO Andy Jassy in his annual letter to shareholders in April.
[Related: AWS Kiro: Five Key Features In Amazon’s New AI Coding Tool]
“If your mission is to make customers’ lives better and easier every day, and you believe every customer experience will be reinvented by AI, you’re going to invest deeply and broadly in AI,” Jassy said. “That’s why there are more than 1,000 GenAI applications being built across Amazon, aiming to meaningfully change customer experiences.”
From revamping Amazon Bedrock and the AWS Marketplace to committing $100 billion in new infrastructure for the AI era and launching slew of agentic AI technologies, here are the 10 biggest AWS stories in 2025 that you need to know about.
No. 10: AWS’ $117 Billion ARR And 29 Percent Global Market Share
AWS remains the No. 1 cloud market-share leader in the world in 2025 by owning 29 percent share of the global cloud infrastructure services market as of first-quarter 2025.
However, AWS’ 29 percent market share is down 2 percentage points year over year compared with owning 31 percent share in first-quarter 2024, according to data from Synergy Research Group. In addition, in first-quarter 2022, AWS owned 33 percent global market share.
It is key to point out that AWS’ two cloud rivals, Google Cloud and Microsoft, have yet to gain any significant market-share gains in 2025.
On the revenue front, AWS continues to break sales records quarter after quarter thanks to growth rates in the high teens over the past several quarters.
AWS generated $29.3 billion in total sales during first-quarter 2025, representing a 17 percent increase year over year.
The cloud giant’s annual run rate now stands at a record $117 billion with no big slowdown in sight.
Click through to read the other nine biggest AWS news stories of 2025.
No. 9: Amazon Bedrock Revamped
Amazon Bedrock is the company’s flagship generative AI platform that enables customers to build and scale GenAI applications.
In 2025, AWS revamped Bedrock with new features and more foundation model integrations, such as its new family of Amazon Nova models and new AI video model Ray2 from Luma AI. With Luma Ray2 now inside Amazon Bedrock, users can add production-ready videos generated from text in their generative AI application through a single API.
AWS also launched multi-agent collaboration capabilities on Amazon Bedrock this year that enable developers to build, deploy and manage networks of AI agents that work together seamlessly to efficiently execute complex workflows.
In addition, AWS enabled customers to deploy DeepSeek’s R1 Distill Llama models on Amazon Bedrock. Bedrock customers can also utilize a new Custom Model Import feature that allows users to integrate their fine-tuned models—such as DeepSeek R1—seamlessly into the Bedrock environment.
Other highlights include the new API Keys feature, which provides quick access to Bedrock APIs by streamlining the authentication process as well as Anthropic’s newest Claude 4 model now available in Bedrock.
No. 8: New IRHX Liquid Cooling Technology For AI Era
In a move to boost its data center, Nvidia partnership and customers’ ability to deploy AI at scale, AWS unveiled its new in-house-engineered liquid-cooling system designed for its most powerful AI servers.
AWS’ new IRHX is a rack-level liquid-cooling platform engineered to support AWS’ highest-density AI training and inference instances built around Nvidia Blackwell GPUs. IRHX stands for In-Row Heat Exchanger technology.
This is big news for AWS as its own liquid-cooling technology helps customers scale quickly by standardizing on AWS hardware, is built for rack dimensions within AWS existing data centers and ensures optimal performance and AI deployment speeds.
IRHX includes three key features: a water‑distribution cabinet, an integrated pumping unit and in‑row fan‑coil modules. Direct liquid-cooling cold‑plates draw heat directly from the chips, then the warmed coolant flows through the coils of heat exchangers where fans discharge the heat into the hot‑aisle containment before the loop recirculates.
No. 7: AWS Launches SMB Partner Initiative, New Government Competency
On the channel front, AWS launched its Small Business Accelerator Initiative as well as a new AWS Government Competency for partners this year.
The AWS Small Business Accelerator Initiative is a partner-led channel sales motion that offers dedicated support to partners serving SMB customers. AWS Partner Territory Managers aim to drive customer awareness through AWS demand generation and will then pass leads to SMB partners and distributors.
Another big launch on the channel front this year is AWS’ new Government Competency, which helps government customers identify partners with deep expertise across citizen services, public safety, and defense and national security.
Partners who achieve the Government Competency are validated for their deep technical expertise and proven customer success in key areas of government procurement such as compliance, architecture and domain expertise critical to government customers.
In addition, AWS in 2025 enhanced its Think Big for Small Business partner program with new benefits and resources. This includes in-person workshops around technical enablement and funding opportunities, as well as access through a new public sector small-business solutions area on the AWS Marketplace.
No. 6: AWS Marketplace Makeover
The AWS Marketplace has been revamped in 2025 thanks to heavy investments to drive easier procurement and selling on the online marketplace, as well as new tools for partners.
AWS has released dozens of new features for the AWS Marketplace this year, including a new seller experience for machine learning products; customer payment schedules for channel partner private offers; a new fulfillment experience for container products; and expanding the AWS Financing Program for all Marketplace purchases for U.S.-based customers.
The company also unveiled support for SaaS products deployed on AWS, on other cloud infrastructure and on-premises, and added self-service seller on-boarding support for demo and private offer requests.
In 2024, over 99 percent of AWS’ top 1,000 customers had at least one active AWS Marketplace subscription.
No. 5: AWS Innovation Engine Roars: 5 Big Launches
AWS’ innovation engine around AI in 2025 has been groundbreaking. CRN reviewed the majority of all of AWS’ new launches so far this year and found five that highlight the wide variety of innovation coming from the company.
First, AWS said its new Amazon Aurora DSQL is the fastest serverless distributed SQL database in the world. Unlike most traditional databases, Aurora DSQL is disaggregated into multiple independent components such as a query processor, adjudicator, journal and crossbar. Customers can remove the operational burdens of patching, upgrading and maintenance downtime to create a new database in just a few steps.
Another monumental launch in 2025 was the second generation of its AWS Outposts racks. The new Outposts generation includes support for the latest x86-powered Amazon EC2 instances, new simplified network scaling and configuration, and accelerated networking instances designed for ultra-low latency and high-throughput workloads.
In addition, AWS recently unveiled specialized Model Context Protocol (MCP) servers for Amazon EC2, Amazon EKS and AWS Serverless. Whether customers are writing code for their environment or debugging production issues, the new MCP servers support AI code assistants with deep understanding of Amazon ECS, Amazon EKS and AWS Serverless capabilities.
On the chip front, AWS said its new quantum computing chip, Ocelot, reduces error correction resources by 90 percent. Ocelot’s building error correction technology was built from the ground up. AWS researchers combined technology and additional quantum error correction components onto a microchip that can be manufactured in a scalable fashion using processes borrowed from the microelectronics industry.
Lastly, Amazon’s new SageMaker Unified Studio lets customers leverage it as their single data and AI development environment, where users can access all of their organization’s data and work using the best tools for specific needs. SageMaker Unified Studio is a single data and AI development environment that brings together functionality and tools from stand-alone studios, query editors and visual tools such as Amazon Athena, AWS Glue and AWS Redshift.
No. 4: AWS Doubles Down On Nvidia Partnership, Integrations
AWS doubling down on its longstanding partnership with Nvidia has been key in 2025 so far.
AWS recently unveiled its new P6e-GB200 UltraServers based on Nvidia Grace Blackwell Superchips, targeting the training and deployment of the largest, most sophisticated AI models.
The availability of AWS’ new P6e-GB200 UltraServers followed the company’s launch of Amazon EC2 P6-B200 instances powered by Nvidia B200 GPUs. Those instances are primarily targeted at large-scale, distributed AI training and inferencing for foundation models.
Nvidia’s Blackwell-based servers are the first to leverage AWS’ new cooling In-Row Heat Exchanger (IRHX) technology to support the compute density of GB200 NVL72 racks.
Also in 2025, Nvidia’s made its DGX Cloud available on AWS as well as Nvidia’s AI Enterprise cloud-native software suite.
The two AI giants also formed new alliances this year around high-performance computing, machine learning, health care and financial services.
No. 3: New Chips And Inference Push For Better AI Costs
AWS made serious strides this year to make AI more affordable by building its own chips and inference technology to optimize performance and cost.
AWS launched its Trainium2 AI chip, saying that it typically offers 30 percent to 40 percent better price performance than other GPU-powered instances available.
AWS’ Trainium2 chips also powered the company’s new Amazon EC2 Trn2 UltraServers, which feature 64 such chips connected using its NeuronLink interconnect, allowing them to scale up to 83 peak petaflops.
“We’ll continue to advance our custom chip designs like AWS Graviton, AWS Trainium and AWS Inferentia to deliver even greater performance and energy efficiency,” Garman told CRN earlier this year.
In addition, AWS unveiled its new quantum computing chip, Ocelot, this year.
Amazon CEO Jassy said while AI model training still accounts for a large amount of the total AI spend, “inference will represent the overwhelming majority of future AI cost.”
“Because customers train their models periodically but produce inferences constantly in large-scale AI applications, inference will become another building-block service, along with compute, storage, database and others. We feel strong urgency to make inference less expensive for customers,” said Jassy in his annual letter to shareholders in April. “For AI to be as successful as we believe it can be, the price of inference needs to come down significantly. We consider this part of our mission and responsibility to help make it so.”
No. 2: Massive Global Data Center Expansion Of $100 Billion In 2025
AWS’ largest investment in 2025 has been around data centers as the company strives to expand its AI offerings and cloud infrastructure availability across the globe.
Amazon expects to spend $100 billion in 2025 alone ramping up its infrastructure for AI by expanding into new geographies or increasing the capacity of existing data centers to serve more customers and deliver AI solutions.
In 2025, AWS committed to investing in data center expansion efforts including $13 billion in Australia and over $8 billion in its Asia-Pacific Mumbai Region in India.
In the U.S., AWS has committed a $11 billion in new data center infrastructure in Georgia, $10 billion in North Carolina, as well as $20 billion in Pennsylvania to boost the company’s cloud computing and AI technologies.
“When AWS is expanding its Capex, particularly in what we think is one of these once-in-a-lifetime type of business opportunities like AI represents, I think it’s actually quite a good sign, medium to long term, for the AWS business,” said Jassy during the company’s quarterly financial earnings call in February. “Spending this capital to pursue this opportunity, which from our perspective we think virtually every application that we know of today is going to be reinvented with AI inside of it and with inference being a core building block, just like compute and storage and database.”
No. 1: AWS AI Agents And GenAI Push; AI Business Hits ‘Multibillion-Dollar’ Mark
This year, AWS quickly developed key building blocks for GenAI development and scale, including custom silicon Trainium AI chips; flexible model-building and inference services in Amazon SageMaker and Amazon Bedrock; new Amazon Nova models to provide lower cost and latency for customers’ applications; and AI agent innovation.
AWS launched new Strands Agents, an open-source SDK that takes a model-driven approach to building and running AI agents in just a few lines of code. Strands scales from simple to complex agent use cases, and from local development to deployment in production.
Dubbed AWS’ most capable AI model for complex tasks, the company also unveiled Amazon Nova Premier. “Nova Premier shines as the best teacher model, with its ability to train other Nova models to be just as capable while being faster and more cost-effective,” said AWS’ CEO. “That means customers can create customized AI that fits their exact needs.”
Amazon also released Amazon Nova Act, a new AI model trained to perform actions within a web browser, as well as Amazon Nova Sonic, its new speech-to-speech foundation model that enables developers to build voice-based AI applications that are expressive and humanlike.
AWS said thousands of customers are already using AWS’ Nova models.
As demand for agentic AI and GenAI grows, Amazon has invested billions in 2025 around AI software, services and hardware.
“Our AI business has a multibillion-dollar annual revenue run rate and continues to grow triple- digit year-over-year percentages,” Jassy said in May. “We continue to believe AI is a once-in-a-lifetime reinvention of everything we know, the demand is unlike anything we’ve seen before, and our customers, shareholders and business will be well-served by our investing aggressively now.”
