Microsoft is betting big on AMD silicon to power its next round of cloud services. Under an expanded partnership, the tech giant is deploying the new AMD Helios Rackscale Solution to manage heavy artificial intelligence workloads on Microsoft Azure. The massive hardware rollout includes custom server designs, updated processor architectures, and accelerated networking tech.
The core of this hardware deployment is the Helios platform. This setup integrates Instinct MI455X graphics chips, EPYC Venice processors, and Pensando networking units. Microsoft plans to use this unified architecture to run large scale frontier models and support its cloud artificial intelligence tools. Microsoft CEO Satya Nadella highlighted the goal of this integration, stating
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.
Beyond the custom server racks, Azure is introducing 2 new virtual machine options. These are the HDv2 series, which runs agentic artificial intelligence tasks, and the HXv2 series, designed for semiconductor engineering. Both series run on the 6th Gen AMD EPYC Venice processors. This silicon update brings more power and speed to demanding data pipelines. It represents a major performance jump for developers.
The partnership also touches the network layer of the cloud platform. The companies are integrating Azure Boost with Pensando data processing units to handle massive data connections. These networking changes are aimed at making the cloud architecture far more efficient. According to the official joint announcement, shipping for the Helios platform to customers will begin during the latter half of 2026. This timeline ensures that Microsoft and other enterprise clients will have these systems running before the end of the year.
