Be taught concerning the new AI options and merchandise In Azure introduced at Microsoft Ignite 2025 to assist modernize your cloud infrastructure.
The cloud is greater than only a platform—it’s the engine of transformation for each group. This 12 months at Microsoft Ignite 2025, we’re displaying how Microsoft Azure modernizes cloud infrastructure at international scale—constructed for reliability, safety, and efficiency within the AI period.
From scalable compute to resilient networks and AI-powered operations, Azure gives the muse that helps clients innovate quicker and function with confidence. Our technique is anchored in three key areas, every designed to assist clients thrive in a quickly altering panorama:
- We’re strengthening Azure’s international basis. We’re increasing capability and resilience throughout areas whereas optimizing datacenter design, energy effectivity, and community topology for AI-scale workloads. Our providers are zone-redundant by default, our edge footprint is rising to satisfy low-latency wants, and safety and compliance controls are embedded at each layer. From confidential computing and sovereign cloud architectures to our safety capabilities, Azure is engineered for belief by design.
- We’re modernizing each workload. We’re advancing compute, community, storage, software, and knowledge providers with Microsoft Azure Cobalt and Azure Enhance techniques, Azure Kubernetes Service (AKS) Computerized, and Azure HorizonDB for PostgreSQL. We embrace and combine with Linux, Kubernetes, and open-source ecosystems clients depend on.
- We’re reworking how groups work. We’re embedding AI brokers immediately into the platform by means of Azure Copilot and GitHub Copilot, bringing agent-based capabilities for migration, app modernization, troubleshooting, and optimization. These options take away repetitive duties so groups can deal with structure as an alternative of administration, making an integral a part of how Azure runs end-to-end.
Agentic cloud operations: Introducing Azure Copilot
Azure is getting into a brand new period the place AI turns into the muse for working your cloud. As environments develop extra complicated, conventional instruments and handbook workflows can’t sustain. This brings us to a frontier second, the place AI and cloud converge to redefine operations. We name this agentic cloud ops: a brand new mannequin for the AI period.
What’s Azure Copilot?
Azure Copilot is a brand new agentic interface that orchestrates specialised brokers throughout the cloud administration lifecycle, automating migration, optimization, troubleshooting, and extra, releasing up groups to deal with innovation.
Azure Copilot aligns actions—human or agent—together with your insurance policies and requirements, providing a unified framework for compliance, auditing, and enforcement that respects role-based entry management (RBAC) and Azure Coverage. It gives robust governance and knowledge residency controls, full visibility throughout brokers and workloads, and allows you to convey your individual storage for full management of chat and artifact knowledge. To make the working mannequin actually agentic, we’re introducing six Azure Copilot brokers—migration, deployment, optimization, observability, resiliency, and troubleshooting—in gated preview.
Be taught extra about Azure Copilot in our detailed weblog, and learn to join the preview.
Azure’s AI infrastructure: The spine of modernization
Azure is constructed for dependable, world-class efficiency, delivering at international scale and pace.
With greater than 70 areas and a whole lot of datacenters, Azure gives the most important cloud footprint within the trade. This unified infrastructure helps constant efficiency, capability, and compliance for purchasers in all places.
AI infrastructure that delivers efficiency and scale
We’ve reimagined how our datacenters are constructed and operated to assist the essential wants of the most important AI challenges. In September 2025, we launched Fairwater, our largest and most subtle AI datacenter thus far, and our latest website in Atlanta now joins Wisconsin to kind a planet-scale “AI superfactory.” By utilizing high-density liquid cooling, a flat community structure linking a whole lot of hundreds of GPUs, and a devoted AI WAN spine, we’re giving clients unmatched capability, flexibility, and utilization throughout each AI workload.
Azure is the primary cloud supplier to deploy NVIDIA’s GB300 GPUs at scale, extending our management from GB200 and persevering with to outline the infrastructure basis for the AI period. Every Fairwater website connects a whole lot of hundreds of those best-in-class GPUs, hundreds of thousands of CPU cores, and big storage—sufficient to carry 80 billion 4K films.
A key a part of this evolution is our AI WAN—a high-speed community linking Fairwater and different Azure datacenters to maneuver knowledge shortly and coordinate huge AI jobs throughout websites. It’s engineered to maintain GPUs busy, scale back bottlenecks, and scale workloads past the boundaries of a single location, so clients can deal with larger tasks and get outcomes quicker. Driving down prices by means of innovation, we’ve set a brand new benchmark for safe, high-performance AI: Azure processed greater than 1.1 million tokens per second for language fashions—the equal of writing seven books per second from a single rack.
Azure’s AI infrastructure places supercomputing-level energy in each buyer’s fingers—enabling bigger mannequin coaching, quicker deployment, and broader person attain inside a trusted, compliant atmosphere.
Extending AI infrastructure innovation to your workloads
An thrilling a part of our work on AI datacenters is that the identical architectural breakthroughs that permit us to coach frontier fashions additionally strengthen Azure’s core providers, immediately benefiting all workloads.
Considered one of these examples is Azure Enhance, which offloads virtualization processes historically carried out by the hypervisor and host working system onto purpose-built {hardware} and software program. Mixed with our new AMD “Turin” and Intel “Granite Rapids” digital machines—plus the most recent network-optimized and storage-optimized households—clients are seeing greater than 20 GB per second of managed-disk throughput and greater than 1,000,000 enter/output operations per second (IOPS). Greater than 1 / 4 of our international fleet is now Enhance-enabled, and community throughput has doubled to 400 gigabits per second for our general-purpose and AI SKUs. The infrastructure investments we’ve talked about are already being utilized by modern firms to convey providers to billions of customers.
Cloud-native apps and knowledge
Azure Kubernetes Service (AKS) delivers safe, managed Kubernetes with automated upgrades and scaling. Paired with cloud-native databases like PostgreSQL and Cosmos DB, groups construct quicker and recuperate immediately.
We’re doing the work to bridge the ability of the AI infrastructure into AKS by enabling cutting-edge GPUs to operate as AKS nodes out of the field or actively monitoring their well being.
We haven’t stopped on the infrastructure layer. We’re additionally reinventing how simple it’s to make the most of Kubernetes itself. That’s why we launched AKS Computerized. It embeds greatest practices, automates infrastructure provisioning, and operates essential Kubernetes parts to cut back complexity and enhance reliability. It handles the exhausting elements—patching, upgrades, observability, and safety—so groups can deal with innovation as an alternative of infrastructure.
With AKS Computerized-managed system node swimming pools, we’re making AKS Computerized even lower-touch as a result of now you don’t must run essential Kubernetes parts your self. It’s fully managed by the service. It strikes key providers like CoreDNS and metrics server to Microsoft-managed infrastructure, making it even simpler to focus fully in your apps.
You not solely want easy-to-use software infrastructure; you additionally want easy-to-use knowledge instruments on your functions. We’re introducing Azure HorizonDB for PostgreSQL, which brings breakthrough scalability and AI integration for next-generation functions. Azure DocumentDB is now usually accessible—the primary managed database constructed on the open-source engine we contributed to the Linux Basis.
We’re additionally excited to develop our longstanding partnership with SAP and announce the launch of SAP Enterprise Information Cloud Join for Microsoft Cloth, simplifying entry to knowledge sharing throughout each platforms. Learn the announcement weblog to be taught extra.
Constructing for belief: Resiliency, operational excellence, and safety
The world runs on Azure’s cloud infrastructure. Each enterprise, authorities, and developer depends upon it to be dependable, safe, and at all times accessible. That accountability drives all the pieces we do in Azure. Our mission is to construct probably the most environment friendly, dependable, and cost-effective infrastructure platform of the AI period, one which clients can depend upon on daily basis.
Resiliency isn’t just a characteristic. It’s a design precept, a tradition, and a shared dedication between Microsoft and our clients. Each area, service, and operation is constructed with that accountability in thoughts.
At Microsoft Ignite, we’re taking this dedication additional with new capabilities that strengthen reliability, simplify operations, and assist clients construct with larger confidence.
- Elevating the bar on operational excellence: Operational excellence means reliability is designed from the beginning. Each Azure area is constructed with availability zones, redundant networking, and automatic fault detection. We’re extending that basis with providers like NAT Gateway, now zone-redundant by default for improved community reliability with none configuration required.
- Empowering clients: With Azure Resiliency (public preview), we’re co-engineering resiliency with clients. This new expertise helps groups set restoration aims, check failover drills, and validate software well being, strengthening readiness collectively earlier than points come up.
Evolving safety for contemporary threats
We proceed to develop Azure’s safety basis with new capabilities that make safety less complicated, smarter, and extra built-in throughout the platform. These updates strengthen boundaries, automate protection, and produce AI-powered perception immediately into how clients defend and function their environments.
- Simplifying safety: Azure Bastion Safe by Default is constructed into the platform. It routinely hardens distant entry to digital machines by means of RDP and SSH, decreasing setup time and threat
- Strengthening boundaries: Community Safety Perimeter, now usually accessible, gives safe and centralized firewall to manage entry to PaaS sources.
- Higher protection: And we’re making developments within the Net Software Firewall with Captcha for human verification.
All of this builds on Azure’s broader stack of confidential digital machines, containers, hardware-based attestation, and encryption, supporting safety from {hardware} by means of to software.
What does modernizing workloads seem like in the present day?
Organizations are on a journey to modernize. Most run a mixture of techniques that span many years—from mission-critical databases to new cloud-native providers. Azure meets clients the place they’re, serving to modernize functions and knowledge with flexibility, openness, and built-in intelligence.
Fairly than a one-off effort, modernization is basically about re-architecting agility, scaling effectively, and utilizing AI and open applied sciences with out sacrificing reliability or management.
To assist simplify and speed up the modernization journey, we’re investing that can assist you discover and get to the very best vacation spot on your workloads, whether or not it’s infrastructure as a service (IaaS), platform as a service (PaaS), or software program as a service (SaaS).
Azure’s agentic migration and modernization instruments make it simpler than ever to modernize your apps, knowledge, and infrastructure with pace and precision. For instance, you may transfer current .NET functions immediately into a completely managed atmosphere—no refactoring or containers required—into the brand new App Service Managed Occasion (now in preview).
On the info aspect, the next-generation Azure SQL Managed Occasion (now usually accessible) delivers as much as five-times quicker efficiency and double the storage capability. And Azure Copilot and GitHub Copilot simplify SQL Server, Oracle, and PostgreSQL modernization.
Plus, throughout infrastructure—from VMware to Linux and IT operations—AI brokers streamline migrations, scale back licensing overhead, and automate patching, governance, and compliance, so modernization turns into a repeatable, clever movement.
Clients migrating and modernizing to Azure utilizing our agentic instruments have shared the spectacular outcomes they’ve skilled. Listed below are only a few examples:
- .NET: Greater than 500,000 strains of code upgraded and migrated in weeks.
- Java: 4 occasions quicker modernized functions than earlier than utilizing the brokers.
Learn extra about GitHub Copilot app modernization.
Trying forward
The promise of the cloud was at all times about scale, flexibility, and innovation. With our improvements throughout our infrastructure, datacenters, Azure Copilot, providers, and open-source contributions, that promise expands to drive your enterprise ahead on daily basis.
The subsequent period of the cloud is inevitable. It’s agentic, clever, and human-centered—and Microsoft helps paved the way.
Be part of us at Microsoft Ignite the place you may tune in to our periods to be taught extra:
- Innovation Session
- Breakouts
