Microsoft Build 2026: Satya Nadella unveils Windows 11 developer updates, Maia 200, and AI infrastructure advances

Microsoft continued its AI-focused push at Build 2026 with a series of announcements aimed at improving the Windows developer experience, expanding Azure infrastructure, and deepening its partnership with Nvidia to support the next generation of agentic AI applications.

More Improvements for the Developer-Optimized Windows 11 Experience

Microsoft is bringing familiar Linux command-line tools directly to the Windows environment, reducing friction for developers who regularly work across Linux, macOS, WSL, and Windows. The company also introduced WSL Containers, enabling Linux containers to run natively on Windows without the need for separate virtual machines.

Alongside this, Microsoft unveiled the new wslc.exe command-line tool, which allows developers to build, manage, and run Linux containers directly from Windows. WSL Containers will also enable native Windows applications to run Linux containers for AI workloads, testing pipelines, and other Linux-based tasks. For enterprises, Microsoft is adding management features such as policy-based controls, image source management, and enhanced IT administrator visibility for WSL Containers.

More from Tech

Expanding Azure AI Infrastructure

Microsoft also unveiled its second-generation AI accelerator, which is already operating in production environments in Iowa and is set to expand to Arizona, Italy, Australia, and South Korea. The infrastructure has been specifically designed to deliver the best performance per dollar and per watt across Microsoft's AI fleet, providing customers with additional inference capacity for AI workloads.

The company further announced that new virtual machines powered by its Cobalt 200 processor are now available in preview. The Cobalt 200 chip has already been deployed across more than 10 Azure regions, with additional expansion planned in the coming months.

Microsoft also highlighted Multipath Reliable Connection (MRC), an open networking protocol developed in collaboration with AMD, Broadcom, Intel, OpenAI, and Nvidia. The protocol shifts network intelligence to endpoint devices, allowing AI workloads to dynamically reroute around network issues and maintain performance without costly interruptions or restarts. According to Microsoft, this approach improves infrastructure utilization, delivers more predictable performance, and enhances resilience in large-scale AI computing environments.

To support adoption, the company is also releasing a suite of tools, including libMRC, NCCL integrations, and a verbs shim library, enabling existing RDMA applications to run on MRC with minimal modifications.

The new Azure Cobalt 200 virtual machines deliver up to 50% better CPU performance than the previous generation, support up to 128 vCPUs, and offer faster storage and networking through Azure Boost integration. Microsoft also highlighted an improved cache architecture for data-intensive workloads and built-in memory encryption to enhance security by default.

A New Union Between Nvidia and Windows PCs

Nvidia CEO Jensen Huang took the stage to outline a vision of bringing agentic AI to every developer through a unified, accelerated computing stack. The initiative includes new Windows PCs powered by Nvidia RTX Spark, as well as Nvidia DGX Station systems for Windows.

Through this full-stack partnership, Nvidia and Microsoft aim to address the growing demand for agentic AI by combining advanced hardware with a comprehensive software ecosystem designed for AI development, deployment, and large-scale inference workloads.