Microsoft Azure Expanding AI Infra Choice with AMD Helios™

Artificial intelligence is entering a new era.

As organizations move from AI experimentation to production deployment, infrastructure requirements are changing rapidly. AI workloads now extend far beyond model training to include inference, agentic AI, search, reinforcement learning, simulation, data preparation, and advanced engineering workloads. These emerging requirements demand infrastructure that is purpose-built for how AI is deployed at scale—not simply larger clusters of accelerators. This perspective aligns closely with the AMD and Microsoft joint announcement highlighting a broader portfolio approach spanning accelerators, CPUs, networking, and software. 

Today, AMD and Microsoft are announcing an expanded collaboration to deliver AI infrastructure designed for production-scale deployments in Azure. The collaboration spans AMD Instinct™ accelerators, AMD EPYC™ processors, AMD Pensando™ networking technologies, and the software ecosystem needed to help customers deploy AI with greater choice, flexibility, and efficiency. Microsoft is also adopting AMD Helios rack-scale AI infrastructure powered by AMD Instinct™ MI455X accelerators as part of its ongoing investment in Azure AI infrastructure. 

Building Infrastructure for the Next Generation of AI

The AI industry has historically focused on individual components: CPUs, GPUs, networking devices, and software frameworks. But as AI models scale and enterprise adoption accelerates, infrastructure performance increasingly depends on how all those elements work together. That is why AMD is taking a system-level approach.

AMD Helios represents AMD’s vision for rack-scale AI infrastructure: a fully integrated platform designed around AMD compute, networking, software, power, and cooling technologies. Helios brings together AMD Instinct™ MI455X accelerators, next-generation AMD EPYC™ processors, AMD Pensando™ networking technology, and the ROCm™ software ecosystem in a purpose-built architecture designed for large-scale AI training, inference, fine-tuning, and emerging agentic AI workloads. 

Rather than optimizing for a single workload, AMD Helios is designed to support the growing diversity of AI applications that organizations must deploy in production environments.

Expanding AI Infrastructure Choice in Azure

AMD and Microsoft share a common belief that customers benefit from infrastructure choice. Organizations are increasingly looking for infrastructure tailored to their specific workload requirements, performance objectives, and cost considerations. The AMD and Microsoft collaboration helps address that need by expanding Azure's portfolio of AI infrastructure offerings and enabling workload-optimized architectures across the AI stack. This collaboration extends well beyond AI accelerators alone.

Microsoft is also introducing new Azure infrastructure offerings built on 6th Generation AMD EPYC processors, designed to support demanding workloads ranging from AI data systems to advanced semiconductor design. Together, these technologies provide customers with more options for matching the right infrastructure to the right workload. 

As AI adoption continues to accelerate, delivering choice at every layer of the infrastructure stack becomes increasingly important. AMD and Microsoft are committed to helping customers deploy AI systems that balance performance, efficiency, scalability, and operational simplicity

From Silicon Innovation to AI Platforms

A key theme underlying this collaboration is the evolution from components to platforms. AMD delivers the foundational technologies—accelerators, CPUs, networking, and software—that power modern AI infrastructure. Microsoft brings those technologies together within Azure through enterprise-grade cloud operations, developer services, security capabilities, and AI platform services. This reflects the joint AMD and Microsoft positioning that rack-scale infrastructure becomes more valuable when combined with a complete cloud platform that customers can consume and deploy at scale. 

Together, AMD and Microsoft are working to enable AI infrastructure that helps customers move from experimentation to real-world production deployment.

Open Software, Built for Scale

The future of AI will be defined by openness, flexibility, and customer choice. AMD continues to advance an open ecosystem for AI through AMD ROCm™, open software frameworks, broad model compatibility, and partnerships spanning cloud providers, enterprises, developers, researchers, and startups. Combined with Microsoft Azure’s AI platform capabilities, the result is an environment where customers can innovate faster while maintaining flexibility across models, tools, and deployment strategies.

The collaboration announced today represents another step towards that vision—helping organizations deploy the infrastructure needed to support the next generation of AI workloads while providing greater choice across the AI stack.

Join AMD and Microsoft at Advancing AI 2026

Want to learn more about the AMD–Microsoft collaboration and the future of AI infrastructure? Visit the following Microsoft sessions and activities at Advancing AI:

Right Silicon, Right Workload: Azure and AMD for AI at Scale

Speakers: Alistair Speirs (Microsoft), Igal Figlin (Microsoft), and Scott Tease (AMD)

Learn how Microsoft and AMD are approaching sovereign AI, AI economics, and enterprise-ready infrastructure designed to support production-scale AI deployments. 

Meet the Experts: What’s Next for AI Infrastructure in the Cloud?

Speakers: Andrew Black (Microsoft) and Tyler Hofstede (AMD)

Join an interactive discussion on AMD Helios, AMD Instinct™ MI455X, rack-scale AI infrastructure, and how AMD and Microsoft are working together to enable the next generation of AI workloads in Azure. 

Microsoft at theCUBE

As a part of overall coverage by theCUBE at Advancing AI, Microsoft executives and technical leaders will discuss Azure AI infrastructure strategy, platform innovation, and the evolving role of AI ecosystems that are open.

Visit Microsoft at the Advancing AI Expo

See demonstrations of Microsoft Azure technologies, learn more about the AMD and Microsoft collaboration, and explore a virtual Helios experience at the Microsoft booth.