AMD and Microsoft Expand Partnership: Helios Chips Coming to Azure in Second Half of 2026

AMD Helios combines Instinct MI455X graphics processors, EPYC “Venice” processors, Pensando network technologies and ROCm software into a single platform intended for training and inference of large-scale AI models. On Azure, it will be used for inferential workloads that include state-of-the-art models, Azure AI services, and consumer applications. Delivery to users, including Microsoft, is planned for the second half of 2026.

Along with Helios, Azure is adding two new series of virtual machines powered by sixth-generation AMD EPYC “Venice” processors – HDv2 for agent-based artificial intelligence and data pipelines, and HXv2 for semiconductor design. The partnership extends to the network layer through the integration of AMD Pensando DPUs with the Azure Boost solution.

“AMD and Microsoft have been building high-performance infrastructure together for years, and today we are expanding that partnership to the entire technology platform of AMD’s AI solutions on Azure,” said Dr. Lisa Su, President and Chief Executive Officer of AMD.

“With AMD, we’re expanding the Azure infrastructure portfolio with AMD Helios to provide customers with the performance, scalability and choice needed to develop and run the next generation of AI applications,” added Satya Nadella, president and chief executive officer of Microsoft.