AMD launches Helios, its first rack AI system to rival Nvidia, adding Microsoft as newest buyer

AMD has launched Helios, its first rack-scale artificial intelligence system, directly challenging Nvidia's dominance in the AI hardware market. The announcement also revealed that Microsoft has become the newest customer for AMD's AI chips, planning to deploy the Helios system on its Azure cloud platform.

The Helios system is designed for large-scale AI training and inference, combining AMD's Instinct MI300X accelerators and EPYC processors in a rack-mounted configuration. Microsoft will use these chips to power AI workloads on Azure, marking a significant expansion of the companies' long-term strategic partnership. This move gives cloud customers an alternative to Nvidia's widely used GPU-based systems.

The development matters because Nvidia currently controls an estimated 80% or more of the AI chip market, with its H100 and upcoming Blackwell GPUs being the standard for training large language models. AMD's entry with a complete rack system could increase competition, potentially lowering costs and accelerating AI adoption. Microsoft's endorsement as a major cloud provider adds credibility to AMD's AI push.

Specific details from the announcement include that the Helios system is AMD's first integrated rack-scale AI solution, and Microsoft will deploy both AMD Instinct MI300X accelerators and AMD EPYC processors on Azure for AI inference and high-performance computing. The companies did not disclose financial terms or deployment timelines.

Background: AMD has been steadily building its AI portfolio, launching the MI300X chip in late 2023 as a direct competitor to Nvidia's H100. The company has also been working with other cloud providers like Meta and Oracle, but Microsoft's commitment represents a major win. The partnership between AMD and Microsoft dates back years, with AMD providing chips for Xbox consoles and Azure servers.

Looking ahead, AMD expects the Helios system to begin shipping to customers in the coming quarters. Microsoft plans to integrate the system into its Azure infrastructure, offering it to enterprise customers for AI workloads. The success of this deployment could determine whether AMD can gain meaningful market share against Nvidia in the rapidly growing AI chip market.