Power accelerated computing on Azure with NVIDIA | BRKSP630

Dave Salvator and Maraki Ketema present practical approaches and use cases for combining Azure infrastructure with NVIDIA GPUs, software, and services. Their Microsoft Ignite 2026 session includes a look at GPU-accelerated Microsoft Fabric Data Warehouse and its architecture and performance.

Accelerated computing for AI and enterprise workloads

The session examines how NVIDIA’s full-stack accelerated computing platform on Azure supports demanding AI, data science, and simulation workloads. The focus is on improving performance, scaling infrastructure efficiently, and moving workloads from experimentation into production.

The discussion follows the evolution of AI from chatbots to agentic AI and the growing data demands that accompany that shift. It also covers:

GPU-accelerated Microsoft Fabric Data Warehouse

The Fabric portion introduces Microsoft Fabric as a unified, AI-driven data platform, then examines enterprise analytics challenges and the need for scalable GPU-accelerated workloads.

An architecture and performance demonstration shows GPU-accelerated Fabric Data Warehouse in action. The session concludes with customer success stories, public preview availability, and a recap.

Session chapters

Resources and session details