NVIDIA Partners With Microsoft on Unified Stack for Agentic AI Deployment, From Windows Devices to Cloud to Local

2026-06-02 · Source: NVIDIA Blog · Field: Technology & Digital — Artificial Intelligence & Machine Learning, Cloud Computing & IT Infrastructure, Software Development & Engineering · Depth: Expert, extended

Summary

NVIDIA and Microsoft have expanded their partnership to deliver a comprehensive stack for agentic AI deployment across Windows devices, Azure cloud, and local environments. Key announcements include RTX Spark PCs, offering 1 petaflop AI performance and up to 128GB unified memory, and DGX Station for Windows, a deskside AI supercomputer with 20 petaflops FP4 performance for 1 trillion parameter models, both running NVIDIA OpenShell for secure agent runtime. The collaboration also accelerates Microsoft Fabric Data Warehouse with NVIDIA GPUs, achieving up to 7x faster SQL execution. NVIDIA's open models, like Nemotron 3 Ultra and Cosmos 3, are now available on Microsoft Foundry, supporting enterprise-scale agentic workflows. Furthermore, the partnership extends to physical AI, secure agent development in GitHub Copilot, and the deployment of NVIDIA Grace Blackwell and Vera Rubin systems in Azure's Fairwater AI factories, significantly boosting inference throughput and reducing costs. Microsoft also introduced new MAI models and Frontier Tuning for custom enterprise AI.

Key takeaway

For AI Architects evaluating agentic AI deployments, this partnership provides a robust, full-stack solution spanning edge to cloud. You should consider integrating NVIDIA's RTX Spark or DGX Station for Windows for local agent development and leveraging GPU-accelerated Microsoft Fabric for data-intensive workflows. Utilize NVIDIA OpenShell for secure agent runtimes and explore Frontier Tuning to build continuously improving, custom enterprise AI models, ensuring your solutions are both performant and secure.

Key insights

NVIDIA and Microsoft are building a full-stack, secure, and scalable ecosystem for agentic AI, from edge devices to cloud.

Principles

• Agentic AI requires a full stack, not just models.
• Secure runtimes are critical for autonomous agents.
• Co-designing silicon and models optimizes performance.

Method

Frontier Tuning enables organizations to build "hill-climbing machines" by defining private evals, contexts, tools, and training models on their own data for continuous, task-specific agent improvement.

In practice

• Deploy agentic AI on Windows PCs with RTX Spark or DGX Station.
• Utilize NVIDIA OpenShell for sandboxed agent execution in GitHub Copilot.
• Accelerate data warehouses with NVIDIA GPUs in Microsoft Fabric.

Topics

• Agentic AI
• NVIDIA RTX Spark
• DGX Station
• Microsoft Azure
• GPU Acceleration
• AI Security
• Microsoft Foundry

Code references

• microsoft/physical-ai-toolchain

Best for: CTO, VP of Engineering/Data, MLOps Engineer, AI Engineer, AI Architect, Director of AI/ML

Related on AIssential

• Nvidia stretches compute power to Windows PCs in support of agents — Nvidia is expanding its full-stack AI compute strategy by bringing powerful agentic AI capabilities directly to Windows PCs and deskside supercomputers.
• NVIDIA Levels Up Local AI Agents Across RTX PCs and DGX Spark — NVIDIA is driving secure, high-performance local AI agent adoption across its RTX and DGX hardware and software platforms.
• 621: Going Deep on Nvidia GTC 2026, Agentic Era, Groq Strategy, OpenClaw Hype, The Android of Self-Driving, 1990 Video Store Simulator, Shingles Vaccine vs Cognitive Decline, and Rick Beato Interview — The AI industry is rapidly transitioning to an agentic era, driven by Nvidia's interconnected hardware and software innovations.
• The foundation for a governed agent workforce: DataRobot and NVIDIA RTX PRO 4500 — The NVIDIA RTX PRO 4500 provides a balanced hardware foundation for enterprise AI agent inference.
• AWS and NVIDIA deepen strategic collaboration to accelerate AI from pilot to production — AWS and NVIDIA are deepening their partnership to provide integrated, scalable, and secure AI infrastructure and services.
• Nvidia's DGX Station is a desktop supercomputer that runs trillion-parameter AI models without the cloud — Nvidia's DGX Station brings data center-scale AI compute to the desktop, enabling local execution of trillion-parameter models.

Open in AIssential →

Editorial summary, takeaway, and curation by AIssential. Original article published by NVIDIA Blog.