Microsoft Touts AI Advances with NVIDIA

Microsoft is doubling down on its AI ambitions, announcing a suite of new solutions at NVIDIA's GTC event designed to streamline the creation and deployment of sophisticated AI systems. The company is integrating its Microsoft Foundry platform with NVIDIA's latest hardware and models, aiming to accelerate the path from AI prototyping to production-ready deployments.

Central to these updates is an expansion of Microsoft Foundry's capabilities. The Foundry Agent Service and Control Plane are now generally available, allowing enterprises to build and manage AI agents that can reason, plan, and act across various tools and data sources. This move aims to boost developer productivity and build crucial enterprise trust in AI operations.

Microsoft is also simplifying the creation of voice-driven AI experiences through a public preview of Voice Live API integration with Foundry Agent Service. Coupled with expanded integrations for security tools like Palo Alto Networks’ Prisma AIRS, the platform seeks to offer a more robust and secure agent lifecycle.

Foundry Embraces NVIDIA's Nemotron

Further bolstering its model ecosystem, Microsoft Foundry now offers NVIDIA's Nemotron models. This integration, alongside recent additions like Fireworks AI, provides customers access to a wide array of frontier and open models, facilitating fine-tuning for low-latency edge deployments.

Next-Gen AI Infrastructure on Azure

Microsoft is also upgrading its Azure AI infrastructure to meet the demanding requirements of inference-heavy AI workloads. The company highlighted its rapid deployment of liquid-cooled Grace Blackwell GPUs and announced it will be the first hyperscale cloud to power NVIDIA's new Vera Rubin NVL72 systems.

This infrastructure push extends to sovereign and regulated environments. With initial support for NVIDIA Vera Rubin on Azure Local, Microsoft aims to bring accelerated AI capabilities to customer-controlled settings, maintaining Azure-consistent operations through Azure Arc and Foundry Local.

Bridging the Physical AI Divide

Microsoft and NVIDIA are also collaborating to advance what they term 'Physical AI.' This effort focuses on integrating NVIDIA's Physical AI Data Factory Blueprint with Microsoft Foundry on Azure. The goal is to enable developers to build, train, and operate robotics and physical AI workflows by connecting physical assets, simulation, and cloud training environments.

A public Azure Physical AI Toolchain GitHub repository is being introduced to support this. Furthermore, deeper integration between Microsoft Fabric and NVIDIA Omniverse libraries is intended to connect live operational data with digital twins for real-time insights and AI-driven actions in manufacturing and beyond.

Microsoft's strategy combines its extensive AI infrastructure, platforms, and systems with NVIDIA's latest innovations. This partnership aims to enable continuous intelligence operations, handling inference-heavy, reasoning-based, and physical AI workloads with enterprise-grade performance, security, and governance.