NVIDIA Partners With Microsoft on Unified Stack for Agentic AI Deployment, From Windows Devices to Cloud to Local
NVIDIA Blog · Dave Salvator · 2026-06-02
NVIDIA and Microsoft announce an expanded partnership at Microsoft Build 2026 covering RTX Spark personal AI PCs, DGX Station for Windows, GPU-accelerated Azure Fabric with up to 7x data warehouse speedups, and NVIDIA OpenShell sandboxing integrated into GitHub Copilot.
Appears in
Extraction
Topics: agentic-ainvidia-microsoft-partnershipai-hardwaremicrosoft-foundrygithub-copilot
Claims
- NVIDIA RTX Spark laptops deliver 1 petaflop of AI performance with up to 128GB unified memory and are positioned as the first Windows PCs purpose-built for personal AI agents.
- DGX Station for Windows provides up to 748GB coherent memory and 20 petaflops of FP4 performance, capable of running frontier models up to 1 trillion parameters.
- Microsoft Fabric with NVIDIA GPU acceleration achieves SQL execution up to 6x faster than CPU baselines and up to 7x faster than three leading cloud data warehouses for high-concurrency workloads.
- NVIDIA OpenShell, now integrated into GitHub Copilot, sandboxes each agent in an isolated container with policy-based access controls governing files, networks, and credentials.
- NVIDIA Vera Rubin delivers up to 10x inference throughput per megawatt compared to Blackwell, reducing cost per agentic token by an order of magnitude.
Key quotes
The agentic AI moment has arrived, but delivering on its promise requires more than good models. It also takes fast hardware, secure runtimes, a responsive data layer and models tuned for long-running reasoning.
RTX Spark is a new beginning, powering the world's first Windows PCs purpose-built for personal agents, with 1 petaflop of AI performance, up to 128GB of unified memory, all-day battery life, and full AI and graphics performance unplugged.
Each agent runs isolated in its own sandboxed container, and every outbound call is evaluated against policy before it can reach files, networks or credentials.