Microsoft Deploys AMD Helios to Power Next-Generation AI on Azure

Dubai | EcoPulse24

Microsoft and AMD have significantly expanded their long-term strategic partnership, broadening their collaboration across AI accelerators, CPUs, networking technologies and software as Microsoft strengthens Azure's infrastructure for the next generation of artificial intelligence.

At the center of the expansion is Microsoft's decision to deploy the AMD Helios Rackscale Solution at scale across Azure to power frontier AI model inference for Microsoft, its AI customers, and Azure AI Services.

The announcement also includes the introduction of two new Azure virtual machine families powered by AMD's upcoming 6th Generation EPYC "Venice" processors, broader deployment of AMD Pensando Data Processing Units (DPUs) across Azure's AI networking infrastructure and selected Azure services, as well as deeper integration between AMD technologies and Azure Boost to improve cloud networking performance at hyperscale.

AMD said it expects to begin shipping the Helios platform to customers, including Microsoft, during the second half of 2026.

Microsoft Expands Its AI Infrastructure with AMD Helios

The AMD Helios Rackscale Solution represents AMD's next-generation open, integrated rackscale platform, designed to support both large-scale AI training and inference.

The platform combines:

• AMD Instinct™ MI455X AI accelerators

• 6th Generation AMD EPYC™ "Venice" processors

• AMD Pensando™ networking

• ROCm™ open software platform

into a unified AI infrastructure built for hyperscale deployments.

Microsoft will deploy Helios across Azure to support frontier model inference, Azure AI Services, and customer AI applications, providing additional compute capacity for some of the industry's most demanding AI workloads.

Azure Broadens AMD-Powered Infrastructure

The expanded partnership significantly increases the availability of AMD-powered AI infrastructure across Azure.

According to the companies, frontier model builders will be able to leverage the new infrastructure to train and serve large-scale AI models, while enterprise customers will gain access to production-ready AI infrastructure through Azure AI Foundry Managed Compute, enabling them to deploy and scale enterprise AI applications.

The expansion reflects Microsoft's strategy of offering customers broader infrastructure choices across different AI workloads while supporting increasingly complex model development.

New Azure Virtual Machines Powered by EPYC "Venice"

Azure will introduce two new virtual machine series based on 6th Generation AMD EPYC "Venice" processors.

Azure HDv2

Designed for:

• Agentic AI

• Data preparation

• Data pipelines

Azure HXv2

Optimized for:

• Semiconductor design

• High-performance engineering workloads

The new VM families expand Azure's AMD EPYC portfolio across AI, engineering and data-intensive computing environments.

Collaboration Extends into Azure Networking

The strategic partnership also expands into the networking layer that connects Azure's cloud infrastructure.

Building on Microsoft's existing deployment of AMD Pensando DPUs, Azure will further deploy the technology across AMD AI backend networking infrastructure as well as selected Azure services.

The two companies are also integrating AMD silicon with Azure Boost to:

• Improve cloud networking performance

• Increase networking efficiency

• Accelerate connection processing

• Scale networking performance across Azure's global cloud fleet

The networking expansion is designed to improve performance while supporting the rapidly increasing compute density required by modern AI infrastructure.

Executive Commentary

Dr. Lisa Su, Chair and CEO of AMD, 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."

Satya Nadella, Chairman and CEO of Microsoft, said:

"Customers are looking for AI infrastructure that is optimized 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."

AI Demand Drives the Next Phase of Azure Expansion

The companies said accelerating global demand for AI infrastructure continues to drive investment across compute, networking and software.

AMD and Microsoft said they will continue delivering open, high-performance AI infrastructure that provides customers with the flexibility, efficiency and scalability required to build the next generation of AI applications.

EcoPulse24 Analysis

The announcement represents far more than a hardware refresh. It signals Microsoft's continued strategy of diversifying Azure's AI infrastructure by expanding AMD's role across the entire technology stack - from GPUs and CPUs to networking and software.

Helios is particularly significant because it combines Instinct MI455X accelerators, EPYC Venice processors, Pensando networking and ROCm software into a single rackscale architecture designed for both AI training and inference. This gives Azure another hyperscale AI platform capable of supporting frontier models while offering customers additional architectural choice.

The partnership also extends beyond compute. Microsoft's expansion of Pensando DPUs, deeper Azure Boost integration, and the launch of new EPYC-powered virtual machines demonstrate that networking efficiency, infrastructure scalability and workload optimization are becoming as strategically important as raw AI compute performance.

As competition intensifies among hyperscale cloud providers, the broader deployment of AMD technologies across Azure positions Microsoft to expand its AI infrastructure while giving customers greater flexibility in how they build, train and deploy next-generation AI applications.

Key Highlights