AMD Helios Powers Azure AI Inference

Microsoft is deploying AMD's Helios AI inference platform on Azure, integrating new EPYC CPUs and Instinct GPUs to power frontier models and AI services.

AMD Helios Rackscale Solution hardware components
AMD's Helios platform integrates GPUs, CPUs, and networking for AI.· AMD
Visual TL;DR
Azure AI InferenceDriver
From the article 6 mentionsThis move aims to bolster AI inference capabilities for Microsoft's own frontier models, its growing roster of AI customers, and its suite of Microsoft Azure AI services.
AMD Helios PlatformCore
From the article 4 mentionsThe integration centers on the AMD Helios platform, a comprehensive solution designed for large-scale AI training and inference.
Instinct MI455X GPUsCore
integrating powerful AMD Instinct GPUs for accelerated AI computations
From the articleHelios combines AMD Instinct MI455X GPUs, AMD Instinct MI455X accelerators, AMD EPYC "Venice" CPUs, Pensando networking technology, and the ROCm software stack.
EPYC "Venice" CPUsCore
deploying 6th Gen AMD EPYC processors for core computing power
From the article 2 mentionsHelios combines AMD Instinct MI455X GPUs, AMD Instinct MI455X accelerators, AMD EPYC "Venice" CPUs, Pensando networking technology, and the ROCm software stack.
ROCm Software StackCore
utilizing AMD's open software for GPU programming and AI development
From the articleHelios combines AMD Instinct MI455X GPUs, AMD Instinct MI455X accelerators, AMD EPYC "Venice" CPUs, Pensando networking technology, and the ROCm software stack.
Enhanced AI ServicesOutcome
From the articleThis move aims to bolster AI inference capabilities for Microsoft's own frontier models, its growing roster of AI customers, and its suite of Microsoft Azure AI services.
New Azure VMsEffect
From the article 5 mentionsThese VMs, Azure HDv2 and HXv2, are targeted at agentic AI, data pipelines, and semiconductor design workloads, respectively, expanding Azure's AMD EPYC portfolio.
Strategic PartnershipContext
deepening collaboration between Microsoft and AMD for future innovations
From the articleMicrosoft is deepening its strategic partnership with AMD, announcing a major deployment of AMD's Helios Rackscale Solution on Azure.

Microsoft is deepening its strategic partnership with AMD, announcing a major deployment of AMD's Helios Rackscale Solution on Azure. This move aims to bolster AI inference capabilities for Microsoft's own frontier models, its growing roster of AI customers, and its suite of Microsoft Azure AI services.

The integration centers on the AMD Helios platform, a comprehensive solution designed for large-scale AI training and inference. Helios combines AMD Instinct MI455X GPUs, AMD Instinct MI455X accelerators, AMD EPYC "Venice" CPUs, Pensando networking technology, and the ROCm software stack.

Microsoft will also introduce two new virtual machine series on Azure, powered by 6th Gen AMD EPYC "Venice" processors. These VMs, Azure HDv2 and HXv2, are targeted at agentic AI, data pipelines, and semiconductor design workloads, respectively, expanding Azure's AMD EPYC portfolio.

The collaboration extends to the network layer, with Microsoft increasing its use of AMD Pensando DPUs. This integration, coupled with Azure Boost, is designed to enhance cloud networking performance and efficiency.

AMD expects to begin shipping Helios to customers, including Microsoft, in the latter half of 2026.

© 2026 StartupHub.ai. All rights reserved. You may not republish this article in full without a license. Search engines and AI research tools may crawl and summarize for reference. Bulk reproduction or model training requires a license. See our terms.
Daniel Singer

Written by

Daniel Singer

Editor, StartupHub.ai

Daniel Singer is the editor of StartupHub.ai, a technology expert and thought leader on AI and its applications across sectors, from fintech and healthcare to developer tooling and consumer software. He writes and tests the tools covered here thoroughly and regularly, and built StartupHub.ai to give founders, operators and buyers a clearer read on what they are actually being sold.

More from Daniel Singer