Surprisingly Useful AI Article Enhancements
AMD and Microsoft have expanded their strategic partnership on Azure, with Microsoft committing to deploy AMD’s Helios rackscale AI platform and introduce new virtual machine series powered by 6th Gen AMD EPYC processors.
The expanded collaboration spans GPUs, CPUs, networking and software across Microsoft’s Azure cloud platform. AMD will begin shipping Helios to customers including Microsoft in the second half of 2026.
Helios On Azure
At the centre of the expansion is Microsoft’s planned deployment of the AMD Helios Rackscale Solution to power AI inference workloads. The deployment will cover frontier model inference for Microsoft’s own services, its AI customers and Azure AI services.
Helios combines AMD Instinct MI455X GPUs, AMD EPYC “Venice” CPUs, Pensando networking and ROCm software in an integrated rackscale platform built for large-scale AI training and inference. On Azure, the Helios-powered offering will appear as the ND MI455X v7 VM series, targeting reasoning, search and agentic workloads behind production AI services.
Dr. Lisa Su, Chair and CEO of AMD, pointed to the breadth of the expanded relationship.
“AMD and Microsoft have spent years building high-performance infrastructure together, and we’re extending that partnership across the full stack of AMD AI solutions on Azure,” Su commented. “Microsoft’s new AMD deployments mark a milestone as we deliver compute solutions to Azure customers and scale the next generation of AI infrastructure together.”
Satya Nadella, Chairman and CEO of Microsoft, framed the expansion around customer choice.
“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,” Nadella stated. “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.”
Azure HDv2 For Agentic AI
The first of the two new VM series is Azure HDv2, designed for agentic AI workloads and data pipelines. Microsoft describes it as built to handle data preparation, search, reinforcement learning and agent coordination at scale.
HDv2 VMs feature close to 500 physical 6th Gen AMD EPYC CPU cores, 4 terabytes of RAM, 32 terabytes of local NVMe storage and 400 Gb Azure Boost networking.
Microsoft positions HDv2 as addressing a specific bottleneck in AI infrastructure: GPU-based accelerators depend on high-density CPU compute to process data, coordinate workloads and keep pipelines running. Without sufficient CPU capacity, training jobs lack data throughput and agents lack the compute needed to perform tasks on behalf of customers.
Azure HXv2 For Semiconductor Design
The second new series is Azure HXv2, which targets semiconductor design and technical computing. The VM series builds on the existing HX series launched in partnership with AMD in 2023, which has seen adoption among silicon design firms using Azure for electronic design automation (EDA) workloads.
HXv2 VMs feature 176 AMD 6th Gen EPYC CPU cores with a clock frequency above 5 GHz, 50 per cent more addressable cache per core compared to the previous generation, and VM sizes with close to 2 or 4 terabytes of RAM. The VMs include 800 Gb InfiniBand connectivity for large-scale MPI-based simulations.
The series continues to use AMD’s 3D V-cache technology for RTL simulation workloads and extends into scientific simulation, engineering analysis and other distributed memory applications.
AMD As A Customer Of Its Own Hardware
AMD itself uses the HX series for its own chip design work. Mark Papermaster, Executive Vice President and CTO of AMD, described the platform’s role in the company’s engineering operations.
Related: Best Business Laptops for work & school
Related: Best Gaming Laptops
Related: Best Portable Laptop
“Engineering teams are pushing the limits of simulation, chip design and scientific computing. At AMD, we experience those demands firsthand as we design future AMD EPYC CPUs and AMD Instinct GPUs,” Papermaster explained. “Azure HX is an important platform for scaling complex EDA workloads, and we look forward to continuing our collaboration with Microsoft as we help advance infrastructure for engineering and scientific workloads.”
Synopsys, the EDA software vendor, also weighed in on the HXv2 announcement. Shankar Krishnamoorthy, Chief Product Development Officer at Synopsys, pointed to the role of cloud-based compute in extending chip design beyond on-premises infrastructure constraints.
“As AI compute continues to push the limits of semiconductor design, our collaboration with Microsoft on the Azure HX-series demonstrates a shared vision for enabling customers to deliver next-generation AI systems with precision and scale in accelerated design cycles,” Krishnamoorthy commented. “These systems have enabled Synopsys customers to reliably and efficiently leverage cloud-based compute, extending EDA workloads beyond traditional infrastructure constraints.”
Networking Infrastructure
The partnership also extends into the networking layer that connects and scales Azure infrastructure. Microsoft has deployed AMD Pensando DPUs in its backend networking and select Azure services.
The two companies are integrating Azure Boost with AMD silicon to scale cloud networking performance, efficiency and connection processing across the Azure fleet.
Broader Access To AMD-Based AI Infrastructure
The collaboration broadens access to AMD-based AI infrastructure on Azure. Frontier model builders can use AMD-powered systems to train and serve large-scale AI models, while enterprise customers can deploy production AI workloads through Azure Foundry Managed Compute.
The expansion sits within Microsoft’s broader approach to AI infrastructure, which encompasses its own purpose-built silicon alongside hardware from partners including AMD. Microsoft has framed this as providing an open, heterogeneous platform rather than locking customers into a single hardware stack.
For AMD, the partnership represents a direct route to large-scale cloud deployment for its Helios platform, Instinct GPUs and EPYC processors at a time when demand for AI inference infrastructure is growing across the cloud computing market.
Last Updated on July 21, 2026 by Nick Ross
