Azure Arc is Microsoft’s central management layer for hybrid and multi‑cloud environments. It’s designed for organizations that run “thousands and thousands of apps, databases, and servers” across different locations—public clouds, private data centers, and edge sites—and need a consistent way to manage them. With Azure Arc, you can: - Bring servers, Kubernetes clusters, and data services from any environment under a single Azure control plane. - Apply consistent governance and security policies across on‑premises and multi‑cloud resources using tools like Azure Policy Guest Configuration. - Manage both virtual machines and containers from one place, regardless of whether they run in Azure, other clouds, or your own data centers. At Microsoft Ignite, Azure highlighted new Arc integrations that make this easier: - **Azure Stack HCI** is now Arc‑enabled by default, so hyperconverged infrastructure in your data center can be managed like native Azure resources. - **VMware vSphere** users can now get self‑service VM control from Azure, including managing VMs based on vSphere templates. For IT teams that are overwhelmed by regulatory change, security threats, and infrastructure sprawl, Arc provides a way to standardize operations without forcing every workload into the public cloud.

Azure Arc extends Azure’s AI and machine learning capabilities beyond the public cloud so you can run them where your data and applications already live. Key capabilities include: - **Model training in Arc‑enabled environments**: Earlier updates made it possible to build and train machine learning models in Arc‑connected infrastructure, not just in Azure’s core regions. - **Machine learning inferencing with Arc**: Azure Arc now supports running inference (serving trained models) across your hybrid and multi‑cloud footprint. This gives you “full consistency between cloud and edge,” so you don’t have to move data to the cloud if that’s not feasible or allowed. This is especially relevant for: - **Highly regulated industries** (for example, customers like Nokia serving regulated sectors) that must keep data in specific locations. - **Latency‑sensitive scenarios** where running AI close to the data source or end user is important. In terms of adoption, most customers today are using these AI features for **proofs of concept rather than full production**. However, Microsoft is already seeing some organizations move models into production in multi‑cloud environments, particularly where regulatory and data‑residency requirements are strict.

Microsoft is expanding how customers can run cloud‑style services in their own data centers, with a particular focus on regulated industries and scenarios where latency matters. A key update is **Azure Virtual Desktop on Azure Stack HCI**: - You can now run **multi‑session Windows 10 and 11 desktop instances** on Azure Stack HCI in your on‑premises data center. - This brings the Azure Virtual Desktop experience closer to your users and data, while still being managed through Azure. This matters for two main reasons: 1. **Regulatory requirements**: Organizations in highly regulated sectors have been asking for local desktop virtualization options that still integrate with Azure. Keeping desktops and data on‑premises can help meet compliance and data‑sovereignty needs. 2. **Latency‑sensitive workloads**: Running virtual desktops locally reduces round‑trip time to the cloud, which can improve user experience for applications that are sensitive to network delays. Combined with Azure Arc and Azure Stack HCI being Arc‑enabled by default, these capabilities let enterprises reimagine how they deliver secure, compliant, and performant desktop and application experiences across hybrid environments.