Microsoft is set to deploy AMD’s Helios Rackscale Solution for AI inference across its Azure cloud platform, thereby expanding a long-term partnership.
Under the expanded collaboration, the AMD Helios system will become a core component in powering large-scale AI models and Azure AI services. Deployment to Microsoft and other customers is expected to begin in the second half of 2026.
Go deeper with GlobalData
Access deeper industry intelligence
Experience unmatched clarity with a single platform that combines unique data, AI, and human expertise.
The Helios Rackscale Solution incorporates AMD Instinct MI455X GPUs, Pensando networking, AMD EPYC “Venice” CPUs, and ROCm software into an integrated platform designed for extensive AI training and inference tasks.
Microsoft plans to use Helios as the basis for AI workloads, including those involving frontier models and customer-specific applications.
In addition to the Helios deployment, Microsoft Azure will introduce two new virtual machine series powered by 6th Gen AMD EPYC “Venice” processors.
The Azure HDv2 series will target agentic AI and data pipeline workloads, while the HXv2 series is aimed at semiconductor design.
These new offerings are intended to extend Azure’s AMD EPYC portfolio and address a broader range of AI, data, and engineering requirements. Network infrastructure will also be affected by the collaboration.
Azure plans to broaden the use of Pensando DPUs to support networking services and will integrate Azure Boost with AMD technology. This integration aims to enhance networking performance, efficiency, and the processing of cloud-scale connections.
AMD chair and CEO Lisa Su said: “AMD and Microsoft have spent years building high-performance infrastructure together, and today we’re extending that partnership across the full stack of AMD AI solutions on Azure.
“Microsoft’s new AMD deployments mark an important milestone as we deliver leadership compute solutions to Azure customers and scale the next generation of AI infrastructure together.”
The firms indicate that this expanded strategic partnership will grant broader access to AMD-based AI infrastructure on Azure. This includes support for developers building large frontier models and enterprises scaling AI workloads through Azure Foundry Managed Compute.
Microsoft chairman and CEO Satya Nadella said: “Customers are looking for AI infrastructure that is optimised for a wide range of workloads, from training and inference to data preparation, search, and reinforcement learning.
“Through our collaboration with AMD, we are expanding the Azure infrastructure portfolio with AMD Helios to give customers the performance, scale and choice they need to build and run the next generation of AI applications.”
Last month, AMD announced the purchase of MEXT, a company focused on memory optimisation technology powered by AI.
