© ROOT-NATION.com - Use of content is permitted with a backlink.
AMD announced the expansion of its strategic partnership with Microsoft, which includes the integration of AMD’s graphics and central processing units, networking infrastructure, and software into the Microsoft Azure cloud ecosystem.
A cornerstone of this collaboration is Microsoft’s integration of AMD’s sustainable Helios (Rackscale Solution) to power inference computations for leading AI models – both for the company’s own needs and those of its AI customers, as well as to support Azure AI services. In addition, two new series of virtual machines based on AMD EPYC processors will be introduced in the Azure cloud, and the deployment of Pensando DPU accelerators will be expanded to support network services. AMD plans to begin shipping Helios systems to customers, including Microsoft, in the second half of 2026.
See also: AERONAUT – everything that flies above the ground: aviation, UAVs and drones, rockets, and space
The AMD Helios platform combines AMD Instinct MI455X graphics accelerators, AMD EPYC Venice central processing units, Pensando networking solutions, and ROCm software within an open, integrated rack-based architecture designed for large-scale AI training and inference. When running on Azure, the Helios system will be used for inference workloads, including cutting-edge models, Azure AI services, and a variety of client applications.
“For many years, AMD and Microsoft have been collaborating to build high-performance infrastructure, and today we are expanding our alliance to encompass the entire stack of AMD AI solutions on Azure,” said Dr. Lisa Su, Chair of the Board of Directors and CEO of AMD. “Microsoft’s adoption of AMD’s latest innovations marks an important milestone in our collaboration, as we deliver cutting-edge computing solutions to Azure users and jointly scale the next generation of AI infrastructure.”
“Customers are looking for AI infrastructure optimized for a wide variety of workloads – from training and inference to data preparation, search, and reinforcement learning,” noted Satya Nadella, Chairman of the Board and CEO of Microsoft. “Through our collaboration with AMD, we are expanding the Azure infrastructure solutions portfolio with the AMD Helios system to provide customers with the performance, scale, and freedom of choice needed to build and run next-generation AI applications.”
This collaboration expands access to AMD’s AI infrastructure across the entire Azure platform. From now on, developers of leading models will be able to use infrastructure based on AMD solutions to train and deploy large-scale AI systems. At the same time, enterprise customers will be able to deploy and scale AI workloads through the Azure Foundry Managed Compute service.
The latest series of Azure virtual machines – Azure HDv2, designed for agent-based AI and data processing pipelines, and Azure HXv2, designed for semiconductor design – will be powered by 6th-generation AMD EPYC processors (codenamed Venice). Together, the new virtual machine series enhance the Azure portfolio of AMD EPYC-based solutions for tasks in the fields of AI, data processing, and engineering calculations.
In addition, this partnership extends to the networking layer, which connects and scales the Azure infrastructure. The companies are integrating Azure Boost technology with AMD solutions to enhance network performance, energy efficiency, and connection processing speeds at cloud scale, building on the widespread deployment of AMD Pensando DPU accelerators across Microsoft’s infrastructure.
Given the rapid growth in demand for artificial intelligence, AMD and Microsoft will continue to provide an open and high-performance infrastructure that ensures customers have the flexibility, efficiency, and scale to develop the technological solutions of the future.
Read also: Microsoft announces a radical restructuring of Xbox: What will change for studios and employees