Microsoft says it will deploy AMD’s Helios rack-scale AI infrastructure and next-generation EPYC processors on Azure, extending a long-running partnership aimed at large-scale AI inference, Azure AI services and high-performance computing workloads.
Microsoft will deploy AMD’s Helios rack-scale AI infrastructure and next-generation EPYC processors on Azure, the companies said on July 20.
In an official announcement, AMD said Microsoft plans to deploy the AMD Helios Rackscale Solution to support frontier-model AI inference for Microsoft, Azure AI services and Microsoft AI customers. Microsoft confirmed the plan in a company blog post, saying Azure is adding AMD Helios rack-scale infrastructure and next-generation AMD EPYC processors for large-scale inference and demanding AI workloads.
The companies framed the move as an expansion of a long-term strategic partnership rather than a one-off procurement. AMD said the deployment will combine its Instinct accelerator roadmap with EPYC server processors, while Microsoft said the additions are intended to strengthen Azure’s AI and high-performance computing infrastructure.
AMD describes Helios as a rack-scale AI platform designed around its next-generation Instinct accelerators and EPYC processors. The goal, according to AMD’s press release, is to provide infrastructure for large AI models that require substantial compute capacity for inference.
Tom’s Hardware reported that Microsoft will make AMD’s Helios-based infrastructure available through Azure for frontier-model workloads, including AI training and inference serving for cloud customers. ITPro also reported that Microsoft is joining AMD’s growing list of Helios customers and said the Azure deployment is part of AMD’s broader push into AI infrastructure.
Microsoft’s blog post emphasized both AI and HPC use cases. The company said the new AMD-based systems are meant to support Azure AI services, Microsoft AI workloads and customers running compute-intensive applications in the cloud.
The announcement comes as cloud providers continue to diversify the hardware behind AI services. Microsoft already operates large AI systems on Azure and has been adding multiple types of accelerators and processors to meet demand from internal teams and external customers.
For AMD, the Microsoft deployment is a notable customer win for its next-generation Instinct and EPYC platforms. AMD’s official statement said the partnership builds on years of collaboration with Microsoft across cloud, AI and high-performance computing.
The companies did not disclose financial terms or a specific public availability date for all Helios-based Azure capacity in the cited announcements. Microsoft’s statement focused on the planned infrastructure expansion and the workloads it expects the systems to support.
Microsoft will deploy AMD’s Helios rack scale AI infrastructure and next generation EPYC processors on Azure, the companies said on July 20.
Microsoft confirmed the plan in a company blog post, saying Azure is adding AMD Helios rack scale infrastructure and next generation AMD EPYC processors for large scale inference and demanding AI workloads.
The companies framed the move as an expansion of a long term strategic partnership rather than a one off procurement.
Continue reading
