From the source
NVIDIA and
Microsoft are widening their full-stack collaboration, running NVIDIA Vera Rubin NVL72 systems in Azure data centers and adding NVIDIA open models and customization through Microsoft Foundry.
- The companies say the Azure infrastructure is built for demanding reasoning, agentic applications, and large-scale model training and inference.
- Microsoft is also using NVIDIA Dynamo inference software for post-training and RTX PRO 6000 Blackwell GPUs for Bing search models.
- Through Foundry, Microsoft's serverless Programmed Training offering gives developers control over post-training workflows while Foundry manages the underlying infrastructure, and Foundry Managed Compute now supports deploying custom model weights.
- EY used Programmed Training to adapt NVIDIA Nemotron open models for classifying corporate purchasing records, improving accuracy on its evaluation dataset by 55% and lowering cost by 15x compared with proprietary frontier models, according to the companies.
- The expanded Nemotron 3.5 lineup on Foundry Managed Compute includes Super for long-running agents and Lightning for fast responses and tool use, while Bolt and Transcribe are listed as coming soon.
- Beyond the cloud, Nemotron 3.5 support in Foundry Local reaches Azure Local environments, and Azure Local is expanding to NVIDIA Jetson Thor and Jetson Orin edge platforms for machine vision, robotics and autonomous operations.






