Microsoft and AMD deepen AI collaboration; Azure introduces Helios rack-level AI system.

```

AMD and Microsoft expand their long-term cooperation, adding another important weight for AMD to challenge Nvidia's dominance in the AI chip market.

On July 20, it was reported that Microsoft will deploy AMD's Helios rack-level solution on its Azure cloud platform to power cutting-edge model AI inference workloads. AMD plans to begin shipments to customers including Microsoft in the second half of 2026.

This cooperation also covers two new series of virtual machines based on the sixth-generation AMD EPYC Venice processors on Azure, as well as the deployment of AMD Pensando Data Processing Units (DPUs) in AMD AI backend network infrastructure and certain Azure services.

Microsoft CEO Satya Nadella said, "We are expanding our Azure infrastructure portfolio with AMD Helios to provide customers with the performance, scale, and choice needed to build and run next-generation AI applications." This collaboration further enriches Microsoft's AI computing power sources and holds important strategic significance for Microsoft, which is accelerating self-developed model R&D and continuously increasing R&D computing power investment.

Helios: AMD's First Rack-Level AI System

Helios is AMD's first rack-level AI system and the most direct competitor so far to Nvidia's Grace Blackwell and Vera Rubin systems. The system integrates AMD's in-house developed GPU, CPU, networking chips, and software. Each compute tray comes with four Instinct GPUs and is powered by a single EPYC central processor, with each tray also equipped with up to 12 network chips developed based on technology acquired from AMD's 2022 Pensando acquisition.

Forrest Norrod, head of AMD's data center business, said the core competitiveness of Helios lies in "offering the best total cost of ownership and lowest cost per token," and noted that customer feedback validates this target. AMD CEO Lisa Su also previously stated that Helios has a "significant advantage" over Nvidia's rack-level systems in inference performance, memory bandwidth, and memory capacity.

In terms of costs, market research firm Futurum Group estimates the price of Helios at $5 million to $5.5 million, which is higher than the firm's estimate of $3.5 million to $4 million for Nvidia's second-generation rack-level system Vera Rubin.

Customer Base Continues to Expand, AMD Accelerates Market Share Capture

With Microsoft's addition as a Helios customer, AMD's buyer camp has further grown.

Previously, Meta announced it would use up to 6 GW of AMD GPUs over time, and would be the first to deploy 1 GW on Helios racks; OpenAI, Oracle, and India's largest IT company Tata Consultancy Services have also made major deployment commitments. AMD says 8 of the world's top 10 AI companies run workloads on its Instinct GPUs.

Currently, Nvidia holds more than 95% share of the data center GPU market, while AMD's share is about 4.5%. Futurum Group analyst Daniel Newman believes AMD is expected to boost its market share to 20-25%, "which would mean hundreds of billions of dollars in revenue."

AMD said that in Q1 2026, its data center business revenue grew 57% year-over-year, and it plans to achieve tens of billions of dollars in AI data center revenue from 2027 onward, most of which will come from Helios.

Long-term Cooperation Deepens, Software Ecosystem Remains a Key Variable

This collaboration is a continuation of AMD and Microsoft's long-term partnership. AMD chips have long powered Microsoft's Surface PCs and Xbox game consoles; in 2023, Microsoft was the first to adopt the AMD MI300X GPU for AI workloads, while also deploying its own Maia chip in its data centers.

According to analysts cited in the report, AMD has basically drawn level with Nvidia on the hardware front, but the software ecosystem remains a key gap. Counterpoint Research analyst Neil Shah said AMD's Helios chips are "comparable to Nvidia’s GPUs and CPUs in performance," but "the key is software and optimization," while Nvidia enjoys a more mature and vast ecosystem thanks to CUDA.

Newman raised another core question: "Did AMD win because of technological leadership, or simply because demand for computing power is so tight that anything that ships is sold?" The actual performance of Helios in early deployments will largely determine the answer.

Risk Warning and DisclaimerThe market has risks, and investment requires caution. This article does not constitute personal investment advice, nor does it take into account the special investment objectives, financial situation, or needs of any individual user. Users should consider whether any opinion, viewpoint, or conclusion in this article is suitable for their specific situation. Investment based on this is at your own risk. ```