AMD announced Monday that it is significantly expanding its strategic partnership with Microsoft, with Microsoft Azure set to deploy AMD's new Helios rack-scale AI systems to power frontier AI inference and Azure AI services. Following the announcement, AMD shares rose about 2% in late trading.

Helios is AMD's first rack-scale AI platform, integrating Instinct MI455X GPUs, sixth-generation EPYC "Venice" processors, Pensando networking technology, and the ROCm software stack. The platform is designed for large-scale AI training and inference. AMD said shipments of Helios systems to customers, including Microsoft, will begin in the second half of 2026.

The collaboration also extends to CPUs and cloud networking. Azure will introduce two new virtual machine families powered by EPYC Venice processors. The HDv2 series is optimized for agentic AI and data pipeline workloads, while the HXv2 series targets semiconductor design applications. Microsoft will also expand deployment of Pensando DPUs across Azure's network infrastructure and integrate them with Azure Boost technology to improve cloud networking performance.

Microsoft Chairman and CEO Satya Nadella said the expanded partnership with AMD will strengthen Azure's AI infrastructure portfolio, giving customers the performance, scale, and flexibility needed to build and run next-generation AI applications. AMD Chair and CEO Lisa Su described Microsoft's Helios deployment as a major milestone that highlights the companies' joint commitment to advancing next-generation AI infrastructure.

Microsoft becomes the latest major customer to publicly adopt Helios, joining Meta, OpenAI, and Oracle. Industry observers view the deployment as AMD's most direct challenge yet to NVIDIA's dominance in the rack-scale AI infrastructure market. According to research firm Futurum Group, successful adoption of Helios could increase AMD's data center GPU market share from roughly 4.5% today to between 20% and 25% over time.