Mirror Review
July 21, 2026
Microsoft and AMD expanded their strategic collaboration to deploy the AMD Helios rackscale platform and next-generation 6th Gen AMD EPYC “Venice” CPUs across Microsoft Azure data centers.
This enhances Azure’s high-performance computing capabilities to power large-scale AI inference, agentic workloads, and custom chip designs.
Under this Microsoft AMD Partnership, Azure will integrate AMD Helios racks, power new HDv2 and HXv2 virtual machines, and scale AMD Pensando DPUs to optimize cloud networking performance.
This initiative gives enterprise customers greater performance, choice, and energy efficiency for demanding workloads.
Microsoft AMD Deal Expands Infrastructure Options
The expanded Microsoft AMD Deal directly addresses the growing compute requirements of frontier AI models, automated agents, and complex data processing pipelines.
By integrating AMD’s full hardware and software stack into Azure, Microsoft expands Azure AI infrastructure to give enterprise customers alternatives for large-scale deployments.
As noted in the official statement, Microsoft CEO Satya Nadella highlighted the goal of this collaboration:
“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.”
Microsoft Azure & AMD Integration
| Technology Component | Target Application on Microsoft Azure |
| AMD Helios Rackscale | Frontier model AI inference & Azure AI services |
| Azure HDv2 VMs | Agentic AI, search & data processing pipelines |
| Azure HXv2 VMs | Semiconductor design (EDA) & scientific computing |
| AMD Pensando DPUs | Accelerated cloud networking & Azure Boost integration |
AMD Launches Helios to Power Production AI Inference
The centerpiece of the latest Microsoft AMD Collab is the deployment of AMD’s Helios rackscale system.
AMD launches Helios as an integrated hardware and software solution designed to handle massive inference demands for enterprise AI services.
Use up and down arrow keys to resize the meta box pane.
The Helios platform combines four core technologies developed by AMD:
Use up and down arrow keys to resize the meta box pane.
- AMD Instinct MI455X GPUs: Built specifically for massive compute tasks and high-bandwidth model execution.
- 6th Gen AMD EPYC CPUs: Code-named “Venice,” these processors manage heavy host node compute and workload distribution.
- Pensando DPUs: Hardware-accelerated networking chips that manage connection scaling and data transport.
- ROCm Open Software: An open-source software layer providing optimization for frontier AI models.
Microsoft plans to deploy Helios in the second half of 2026 to serve internal AI workloads, Azure AI services, and external developers through Azure Foundry.
In an official statement regarding the rollout, Dr. Lisa Su, Chair and CEO of AMD, stated:
Use up and down arrow keys to resize the meta box pane.
“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.”
Use up and down arrow keys to resize the meta box pane.
New Azure Virtual Machines Focus on Specialized Compute
Alongside the Helios rack system, Microsoft is introducing two specialized Azure Virtual Machine (VM) families powered by 6th Gen AMD EPYC processors.
These offerings target specific compute bottlenecks in agentic workflows and silicon engineering.
Azure HDv2 VMs for Data Processing and AI Agents
The HDv2 VM series resolves data pipeline bottlenecks for AI agents and search applications. Each host system includes:
- Nearly 500 physical 6th Gen AMD EPYC CPU cores
- 4 terabytes of RAM
- 32 terabytes of local NVMe storage
- 400 Gb Azure Boost network throughput
Azure HXv2 VMs for Silicon Design and Engineering
Use up and down arrow keys to resize the meta box pane.
The HXv2 VM series serves engineering firms running Electronic Design Automation (EDA) and high-performance computing (HPC) simulations. These instances feature:
Use up and down arrow keys to resize the meta box pane.
- 176 AMD EPYC cores operating at frequencies above 5 GHz
- 3D V-Cache technology with 50% more addressable cache per core
- Up to 4 terabytes of system memory
- 800 Gb InfiniBand interconnect support for large-scale parallel tasks
According to Mark Papermaster, Executive Vice President and CTO at AMD:
“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. Azure HX is an important platform for scaling complex EDA workloads, and we’re excited about Azure HXv2, which is designed to deliver even greater performance and scalability.”
Microsoft and AMD’s Longstanding Partnership
Microsoft and AMD have maintained a hardware relationship spanning several product generations. AMD silicon has powered Microsoft Xbox consoles and Surface PCs for years, while Microsoft was the first major cloud provider to deploy AMD’s MI300X GPUs in 2023.
The current data center hardware landscape is highly competitive.
According to estimates from the Futurum Group, Nvidia holds over 95% of the data center GPU market, while AMD holds roughly 4.5%.
However, industry analysts predict that integrated solutions like Helios could allow AMD to expand its market share significantly as cloud providers search for cost-effective capacity.
The expansion also strengthens cloud networking. Microsoft is integrating its proprietary Azure Boost software with AMD Pensando Data Processing Units (DPUs).
This hardware-software integration offloads storage and networking tasks from main host processors, freeing compute capacity for core applications.
Looking Ahead at Cloud AI Competition
The expanding Microsoft AMD Partnership represents an effort to build scalable, open cloud architecture for the generative AI era.
As model training and inference demands grow, cloud operators rely on diverse, heterogeneous hardware environments to maintain cost efficiency and operational reliability.
Deploying the Helios platform alongside new AMD EPYC virtual machines positions Azure to support complex enterprise workflows at scale.
Maria Isabel Rodrigues
