NVIDIA and Microsoft are extending their full-stack collaboration to help organizations build, customize, and deploy artificial intelligence applications across their entire computing infrastructure1. The partnership spans from next-generation accelerated computing systems in Azure data centers to Windows PCs and industrial systems1. This comprehensive approach addresses a critical challenge in AI commercialization: the need to seamlessly transition from large-scale model training and inference in cloud data centers to running applications at the edge, where businesses generate and use their data1.

The centerpiece of the cloud infrastructure is the Vera Rubin NVL72 system, now available in Azure data centers and designed for demanding reasoning tasks, agentic applications, and large-scale model training and inference1. Microsoft is already leveraging NVIDIA's acceleration technologies for its own services, deploying NVIDIA RTX PRO 6000 Blackwell GPUs to accelerate Bing search models and using NVIDIA Dynamo inference software for post-training optimization1. The partnership also brings customers access to NVIDIA's open-source models and customization capabilities through Microsoft Foundry, enabling organizations to fine-tune models for their specific applications1.

This expansion signals how enterprise AI strategy is evolving. Rather than choosing between cloud and local computing, organizations increasingly need both—cloud systems for large-scale training and inference, and edge systems for running applications where data originates. The NVIDIA-Microsoft partnership positions both companies to serve this emerging need by creating a coherent infrastructure path from Azure data centers to Windows PCs and industrial systems. This full-stack approach could become a competitive advantage for enterprises seeking to build AI applications that seamlessly operate across their entire technology landscape.