Microsoft is set to provide cloud computing customers with the option of choosing AMD chips due to increased demand for Nvidia AI chips

Microsoft to Offer AMD AI Chips for Cloud Computing Customers, Challenging Nvidia’s Dominance

Microsoft is making waves in the cloud computing industry with its latest move to offer its customers a platform of AMD artificial intelligence (AI) chips. This new development, as reported by Reuters, will bring Microsoft head-to-head with the reigning champion of the AI chip market, Nvidia.

The clusters of AMD’s flagship MI300X AI chips will be available to customers through Microsoft’s Azure cloud computing service, offering a compelling alternative to Nvidia’s dominant H100 family of GPUs. With the high demand for Nvidia’s chips making it difficult for customers to obtain, Microsoft’s collaboration with AMD comes as a welcome solution.

In the world of AI, multiple GPUs are typically needed to build and run complex AI models and apps, as a single processor often can’t handle the necessary data and computation. AMD’s powerful chips are expected to generate $4 billion in AI chip revenue this year, boasting the capability to train and run large AI models efficiently.

Microsoft is poised to unveil more details about the AMD chips at its upcoming developer conference, Build. In addition to the AMD AI chip clusters, Microsoft’s cloud computing unit also offers access to its in-house AI chips, Maia.

But the excitement doesn’t end there – Microsoft is also set to launch a preview of its new Arm-based Cobalt 100 custom processors at the upcoming developer event. These chips are said to offer a 40% performance improvement over other processors based on Arm Holdings’ technology, with companies like Snowflake already jumping on board to leverage their capabilities.

The Cobalt chips, designed to power Microsoft’s messaging tool for businesses, Teams, will compete with Amazon’s in-house Graviton CPUs. This move marks Microsoft’s commitment to innovation and competition in the rapidly evolving cloud computing landscape.

Scroll to Top