Microsoft will deploy AMD's Helios rack system on Azure, with shipments due in the second half of 2026

Microsoft will run AMD's newest AI hardware in its data centers. In a statement on July 20, AMD said Microsoft would deploy the AMD Helios rack-scale system on Azure to power inference for large models, part of what both companies called an expansion of a long-standing partnership. Helios brings together AMD's Instinct MI455X accelerators, its EPYC Venice processors, Pensando networking and the ROCm software stack. Shipments begin later this year.

AMD said it will begin shipping Helios to customers, including Microsoft, in the second half of 2026. Azure will also add two virtual-machine families built on the sixth-generation EPYC Venice chips: HDv2, aimed at agentic AI and data pipelines, and HXv2, aimed at semiconductor design work. Lisa Su, AMD's chair and chief executive, and Satya Nadella, Microsoft's chairman and chief executive, are both quoted in the announcement. No dollar figure is attached. The release gives no capacity number either.

The framing matters. AMD sells this as an open, rack-scale option for AI training and inference, and Microsoft joins the list of cloud operators taking AMD's top-end parts rather than leaning on a single supplier. What the announcement does not do is turn any of that into live capacity today.