New options for AI-powered innovation, resiliency, and control with Microsoft Azure

Organizations running mission‑critical workloads operate under stricter standards because system failures can often affect people and business operations at scale. They must ensure control, resilience, and operational autonomy such that innovation does not compromise governance. They need agility that also maintains continuity and preserves standards compliance, so they can get the most out of AI, scalable compute, and advanced analytics on their terms.

For example, manufacturing plants need assembly lines to continue to operate during network outages, and healthcare providers need the ability to access patient data during natural disasters. Similarly, government agencies and critical infrastructure operators must comply with regulations to keep systems autonomous and data within national borders. Additionally, regulations sometimes mandate that sensitive data remains stored and processed locally under local jurisdiction and personnel control.

These are exactly the challenges Azure’s adaptive cloud approach is designed to solve. We are extending Azure public regions with options that adapt to our customers’ evolving business requirements without forcing trade-offs. Microsoft’s strategy spans both our public cloud, private cloud, and edge technology, giving customers a unified platform for operations, applications, and data with the right balance of flexibility and control. This approach empowers customers to use Azure services to innovate in environments under their full control, rather than maintaining separate, siloed, or legacy IT systems.

Meeting unique operational and data sovereignty needs

To address unique operational and data sovereignty needs, Microsoft introduced Azure Local—Azure infrastructure delivered in customers’ own datacenters or distributed locations. Azure Local comes with integrated compute, storage, and networking services and leverages Azure Arc to extend cloud services across the management, data, application, and security layers into hybrid and disconnected environments.

Learn more about what’s new in Azure Local

Over the past six months, our team has significantly expanded Azure Local’s capabilities to meet requirements across industries. We are seeing tremendous momentum from customers like GSK, a global biopharma leader extending cloud innovation and AI to the edge using Azure Local. GSK is enabling real-time data processing and AI inferencing across vaccine and medicine manufacturing and R&D labs worldwide. GSK joined our What’s new in Azure Local session at Ignite last month, offering insight into how they are using Azure Local.

We are also engaging deeply with public sector organizations to ensure essential services can run independent of internet connectivity when needed, from city administrations to defense and emergency response agencies.

To support these customers, we are enabling a growing set of Azure Local features and functionalities across Microsoft and partners, many of which have reached General Availability (GA) and preview.

Microsoft 365 Local (GA) delivers full productivity—email, collaboration, and communications—within a private cloud, ensuring sovereignty and security for sovereign scenarios.

Next-gen NVIDIA RTX PRO 6000 Blackwell Server Edition GPUs (GA) bring high-performance AI workloads on premises, enabling advanced analytics and generative AI without sacrificing compliance.

Azure Migrate support for Azure Local (GA) streamlines lift-and-shift migrations, reducing costs and accelerating time-to-value.

AD-less deployments, Rack-Aware Clustering, and external SAN storage integration (Preview) offer more options for identity, fault tolerance, and flexible storage strategies.

Rack aware clustering is now available in preview for Azure Local

Multi-rack deployments (Preview) dramatically increase options for high scale, supporting larger IT estates in a single integrated Azure Local instance.

Disconnected operations (Preview) delivers a fully disconnected Azure Local experience for mission-critical environments where internet connectivity is infeasible or unwanted.

In short, Azure Local has rapidly evolved into a robust platform for operational sovereignty. It delivers Azure consistency for all workloads from core business apps to AI, in customers’ locations—from a few nodes on a factory floor up to thousands of nodes. These advancements reflect our commitment to meet customers where they are. 

Intelligent and connected physical operations

Azure’s adaptive cloud approach helps bring AI to physical operations. Our Azure IoT platform enables asset-intensive organizations to harness data from devices and sensors in a secure, scalable, and resilient fashion. When combined with Microsoft Fabric, customers get real-time insights from their operational data. This integration allows industries such as manufacturing, energy, and industrial operations to bridge digital and physical systems and adopt AI and automation in ways that align with their specific needs.

Demonstrating how the cloud, edge AI, and simulation can help orchestrate human-robotic collaboration on manufacturing product lines at Microsoft Ignite

Our approach to enable AI in physical operations environments follows two basic patterns. Azure IoT Operations enables device and sensor data from larger sites to be aggregated and processed close to its source for near real-time decision-making and reduced latency, streaming only relevant data to Fabric for more advanced analytics. Azure IoT Hub, on the other hand, enables device data to securely flow directly to Fabric with cloud-based identity and security. The integration across Microsoft Fabric and Azure IoT helps bridge Operational Technology (OT) and Information Technology (IT), delivering cost-effective, secure, and repeatable outcomes.

In the last six months, we introduced several enhancements to Azure IoT tailored for connected operations use cases:

In Azure IoT Hub, a new Microsoft-backed X.509 certificate management capability provides enhanced secure identity lifecycle control. Integration with Azure Device Registry streamlines identity, security, and policy management across fleets.

Enhanced Azure Device Registry capabilities improve asset registration, classification, and monitoring for operational insight while allowing Azure connected assets and devices to be used with any Azure service.

Azure Device Registry (ADR) acts as the unified control plane for managing both physical assets from Azure IoT Operations and devices from Azure IoT Hub

Azure IoT Operations’ latest release includes a number of new features. WebAssembly-powered data graphs enable fast, modular analytics for near-instant decision-making. Expanded connectors for OPC UA, ONVIF, REST/HTTP, SSE, and MQTT simplify interoperability. OpenTelemetry endpoint support enables smooth telemetry pipelines and monitoring. Advanced health monitoring provides deep visibility and control over operational assets.

In Microsoft Fabric, Fabric IQ and Digital Twin Builder turn raw telemetry into actionable context for simulation and intelligent feedback loops thanks to the use of models and knowledge graphs that bring clarity to streaming data.

Customers like Chevron and Husqvarna are scaling Azure IoT Operations from single-site pilots to multi-site rollouts, unlocking new use cases such as predictive maintenance and worker safety. These deployments demonstrate measurable impact and adaptive cloud architectures delivering business value. Our partner ecosystem is also growing with Siemens, Litmus, Rockwell Automation, and Sight Machine building on the platform.

Managing a distributed estate with unified Azure management and security

Organizations often grapple with the complexity of highly distributed IT estates—spanning on-premises datacenters, hundreds or sometimes thousands of edge sites, multiple public clouds, and countless devices. Managing and securing this sprawling ecosystem is challenging with traditional tools. A core promise of Azure’s adaptive cloud approach is helping to simplify centralized operations through a single, unified control plane via Azure Arc.

Over the last six months, we have delivered a wave of improvements to help customers manage distributed resources at scale, across heterogenous environments, in a frictionless way. Key enhancements in our Azure Arc platform include:

Azure Arc site manager (Preview) organizes resources by physical site for easier monitoring and management of distributed operations.

New GCP connector (Preview) projects Google Cloud resources into Azure for a single pane of glass across Azure, AWS, and GCP.

The Multicloud connector enabled by Azure Arc is now in preview for GCP environments

Azure Machine Configuration (GA) enforces OS-level settings across Azure Arc-managed servers for compliance and security.

New Azure policies to audit and configure Windows Recovery environment to be ready for critical patch to recover from machine unbootable state such as faulty drivers.

New subscription level enrollment of essential machine management services with a simplified pricing model and a unified user experience from Azure for the hybrid environment, lowering adoption barrier for legacy environments.

Workload Identity (GA) lets Azure Arc-enabled Kubernetes clusters use Entra ID for secure resource access, eliminating local storage of secrets.

AKS Fleet Manager (Preview) integrates Azure Arc-connected clusters for centralized policy sync and deployments across hybrid environments.

Azure Key Vault Secret Store Extension (GA) allows Azure Arc-enabled Kubernetes clusters to cache secrets from Azure Key Vault, improving security and workload resiliency to intermittent network connectivity for hybrid workloads.

These enhancements underscore our belief that cloud management and cloud-native application development should not stop at the cloud. Whether an IT team is responsible for five datacenters or 5000 retail sites, Azure provides the tooling to manage that distributed environment and develop applications as one cohesive and adaptive cloud.

Azure’s adaptive cloud approach gives organizations the freedom to innovate on their terms while maintaining control. In an era defined by uncertainty, whether from cyber threats or geopolitical shifts, Azure empowers customers to modernize confidently without sacrificing resiliency or control.

Innovate on an adaptive cloud

The post New options for AI-powered innovation, resiliency, and control with Microsoft Azure appeared first on Microsoft Azure Blog.
Quelle: Azure

Announcing vLLM v0.12.0, Ministral 3 and DeepSeek-V3.2 for Docker Model Runner

At Docker, we are committed to making the AI development experience as seamless as possible. Today, we are thrilled to announce two major updates that bring state-of-the-art performance and frontier-class models directly to your fingertips: the immediate availability of Mistral AI’s Ministral 3 and DeepSeek-V3.2, alongside the release of vLLM v0.12.0 on Docker Model Runner.

Whether you are building high-throughput serving pipelines or experimenting with edge-optimized agents on your laptop, today’s updates are designed to accelerate your workflow.

Meet Ministral 3: Frontier Intelligence, Edge Optimized

While vLLM powers your production infrastructure, we know that development needs speed and efficiency right now. That’s why we are proud to add Mistral AI’s newest marvel, Ministral 3, to the Docker Model Runner library on Docker Hub.

Ministral 3 is Mistral AI’s premier edge model. It packs frontier-level reasoning and capabilities into a dense, efficient architecture designed specifically for local inference. It is perfect for:

Local RAG applications: Chat with your docs without data leaving your machine.

Agentic Workflows: Fast reasoning steps for complex function-calling agents.

Low-latency prototyping: Test ideas instantly without waiting for API calls.

DeepSeek-V3.2: The Open Reasoning Powerhouse

We are equally excited to introduce support for DeepSeek-V3.2. Known for pushing the boundaries of what open-weights models can achieve, the DeepSeek-V3 series has quickly become a favorite for developers requiring high-level reasoning and coding proficiency.

DeepSeek-V3.2 brings Mixture-of-Experts (MoE) architecture efficiency to your local environment, delivering performance that rivals top-tier closed models. It is the ideal choice for:

Complex Code Generation: Build and debug software with a model specialized in programming tasks.

Advanced Reasoning: Tackle complex logic puzzles, math problems, and multi-step instructions.

Data Analysis: Process and interpret structured data with high precision.

Run Them with One Command

With Docker Model Runner, you don’t need to worry about complex environment setups, python dependencies, or weight downloads. We’ve packaged both models so you can get started immediately.

To run Ministral 3:

docker model run ai/ministral3

To run DeepSeek-V3.2:

docker model run ai/deepseek-v3.2-vllm

These commands automatically pull the model, set up the runtime, and drop you into an interactive chat session. You can also point your applications to them using our OpenAI-compatible local endpoint, making them drop-in replacements for your cloud API calls during development.

vLLM v0.12.0: Faster, Leaner, and Ready for What’s Next

We are excited to highlight the release of vLLM v0.12.0. vLLM has quickly become the gold standard for high-throughput and memory-efficient LLM serving, and this latest version raises the bar again.

Version 0.12.0 brings critical enhancements to the engine, including:

Expanded Model Support: Day-0 support for the latest architecture innovations, ensuring you can run the newest open-weights models (like DeepSeek V3.2 and Ministral 3) the moment they drop.

Optimized Kernels: Significant latency reductions for inference on NVIDIA GPUs, making your containerized AI applications snappier than ever.

Enhanced PagedAttention: Further optimizations to memory management, allowing you to batch more requests and utilize your hardware to its full potential.

Why This Matters

The combination of Ministral 3, DeepSeek-V3.2, and vLLM v0.12.0 represents the maturity of the open AI ecosystem.

You now have access to a serving engine that maximizes data center performance, alongside a choice of models to fit your specific needs—whether you prioritize the edge-optimized speed of Ministral 3 or the deep reasoning power of DeepSeek-V3.2. All of this is easily accessible via Docker Model Runner.

How You Can Get Involved

The strength of Docker Model Runner lies in its community, and there’s always room to grow. We need your help to make this project the best it can be. To get involved, you can:

Star the repository: Show your support and help us gain visibility by starring the Docker Model Runner repo.

Contribute your ideas: Have an idea for a new feature or a bug fix? Create an issue to discuss it. Or fork the repository, make your changes, and submit a pull request. We’re excited to see what ideas you have!

Spread the word: Tell your friends, colleagues, and anyone else who might be interested in running AI models with Docker.

We’re incredibly excited about this new chapter for Docker Model Runner, and we can’t wait to see what we can build together. Let’s get to work!
Quelle: https://blog.docker.com/feed/

Docker, JetBrains, and Zed: Building a Common Language for Agents and IDEs

As agents become capable enough to write and refactor code, they should work natively inside the environments developers work in: editors. 

That’s why JetBrains and Zed are co-developing ACP, the Agent Client Protocol. ACP gives agents and editors a shared language, so any agent can read context, take actions, and respond intelligently without bespoke wiring for every tool.

Why it matters

Every protocol that’s reshaped development (LSP for language tools, MCP for AI context) works the same way: define the standard once, unlock the ecosystem. ACP does this for the editor itself. Write an agent that speaks ACP, and it works in JetBrains, Zed, or anywhere else that adopts the protocol. 

Docker’s contribution

Docker’s cagent, an open-source multi-agent runtime, already supports ACP, alongside Claude Code, Codex CLI, and Gemini CLI. Agents built with cagent can run in any ACP-compatible IDE, like JetBrains, immediately.

We’ve also shipped Dynamic MCPs, letting agents discover and compose tools at runtime, surfaced directly in the editor where developers work.

What’s next

ACP is early, but the direction is clear. As agents embed deeper into workflows, the winners will be tools that interoperate. Open standards let everyone build on shared foundations instead of custom glue.

Docker will continue investing in ACP and standards that make development faster, more open, and more secure. When code, context, and automation converge, shared protocols ensure we move forward together.
Quelle: https://blog.docker.com/feed/

SES Mail Manager is now available in 10 additional AWS Regions, 27 total

Amazon SES announces that the SES Mail Manager product is now available in 10 additional commercial AWS Regions. This expands coverage from the current 17 commercial AWS Regions where Mail Manager is launched, meaning that Mail Manager is now offered in all commercial Regions where SES offers its core Outbound service. SES Mail Manager allows customers to configure email routing and delivery mechanisms for their domains, and to have a single view of email governance, risk, and compliance solutions for all email workloads. Organizations commonly deploy Mail Manager to replace legacy hosted mail relays or simplify integration with third-party mailbox providers and email security solutions. Mail Manager also supports onward delivery to WorkMail mailboxes, built-in archiving with search and export capabilities, and integration with third-party security add-ons directly within the console. The 10 new Mail Manager Regions include Middle East (Bahrain), Asia Pacific (Jakarta), Africa (Cape Town), Middle East (UAE), Asia Pacific (Hyderabad), Asia Pacific (Malaysia), Europe (Milan), Israel (Tel Aviv), Canada West (Calgary), and Europe (Zurich). The full list of Mail Manager Region availability is here. To learn more, see the Amazon SES Mail Manager product page and the SES Mail Manager documentation. You can start using Mail Manager in these new Regions through the Amazon SES console.
Quelle: aws.amazon.com

Amazon Connect launches WhatsApp channel for Outbound Campaigns

Amazon Connect Outbound Campaigns now supports WhatsApp, expanding on the WhatsApp Business messaging capabilities that already allow customers to contact your agents. You can now engage customers through proactive, automated campaigns on their preferred messaging platform, delivering timely communications such as appointment reminders, payment notifications, order updates, and product recommendations directly through WhatsApp. Setting up WhatsApp campaigns uses the same familiar Amazon Connect interface, where you can define your target audience, choose personalized message templates, schedule delivery times, and apply compliance guardrails, just as you do for SMS, voice, and email campaigns. Previously, Outbound Campaigns supported SMS, email, and voice channels, while WhatsApp was available only for customers to initiate conversations with your agents. With WhatsApp support in Outbound Campaigns, you can now proactively reach customers through an additional messaging platform while maintaining a unified campaign management experience. You can personalize WhatsApp messages using real-time customer data, track delivery and engagement metrics, and manage communication frequency and timing to ensure compliance. This expansion provides greater flexibility to connect with customers on their preferred platforms while streamlining your omnichannel outreach strategy. This feature is available in all AWS Regions where Amazon Connect Outbound Campaigns is supported. To learn more, visit the Amazon Connect Outbound Campaigns documentation.
Quelle: aws.amazon.com

AWS Elastic Beanstalk now supports Node.js 24 on Amazon Linux 2023

AWS Elastic Beanstalk now enables customers to build and deploy Node.js 24 applications on Amazon Linux 2023 (AL2023) platform. This latest platform support allows developers to leverage the newest features and improvements in Node.js while taking advantage of the enhanced security and performance of AL2023.
AWS Elastic Beanstalk is a service that provides the ability to deploy and manage applications in AWS without worrying about the infrastructure that runs those applications. Node.js 24 on AL2023 delivers updates to the V8 JavaScript engine, npm 11, and security and performance improvements. Developers can create Elastic Beanstalk environments running Node.js 24 on AL2023 through the Elastic Beanstalk Console, CLI, or API.
This platform is available in all commercial AWS Regions where Elastic Beanstalk is available, including the AWS GovCloud (US) Regions. For a complete list of regions and service offerings, see AWS Regions.
To learn more about Node.js 24 on Amazon Linux 2023, see the AWS Elastic Beanstalk Developer guide. For additional information, visit the AWS Elastic Beanstalk product page.
Quelle: aws.amazon.com

Amazon SES adds VPC support for API endpoints

Today, Amazon Simple Email Service (SES) added support for accessing SES API endpoints through Virtual Private Cloud (VPC) endpoints. Customers use VPC endpoints to enable access to SES APIs for sending emails and managing their SES resource configuration. This release helps customers increase security in their VPCs. Previously, customers who ran their workloads in a VPC could access SES APIs by configuring an internet gateway resource in their VPC. This enabled traffic from the VPC to flow into the internet, and reach SES public API endpoints. Now, customers can use the VPC endpoints to access SES APIs without the need for an internet gateway, reducing the chances for activity in the VPC to be exposed to the internet.. SES supports VPC for SES API endpoints in all AWS Regions where SES is available. For more information, see the documentation for information about setting up VPC endpoints with Amazon SES.
Quelle: aws.amazon.com

Amazon SageMaker now supports self-service migration of Notebook instances to latest platform versions

Amazon SageMaker Notebook instance now supports self-service migration, allowing you to update your notebook instance platform identifier through the UpdateNotebookInstance API. This enables you to seamlessly transition from unsupported platform identifiers (notebook-al1-v1, notebook-al2-v1, notebook-al2-v2) to supported versions (notebook-al2-v3, notebook-al2023-v1). With the new PlatformIdentifier parameter in the UpdateNotebookInstance API, you can update to newer versions of the Notebook instance platform while preserving your existing data and configurations. The platform identifier determines which Operating System and JupyterLab version combination your notebook instance runs. This self-service capability simplifies the migration process and helps you keep your notebook instances current. This feature is supported through AWS CLI (version 2.31.27 or newer) and SDK, and is available in all AWS Regions where Amazon SageMaker Notebook instances are supported. To learn more, see Update a Notebook Instance in the Amazon SageMaker Developer Guide.
Quelle: aws.amazon.com