Breaking News | AMD Lands Microsoft Azure Deal for Full-Stack AI Systems

Microsoft announced it will deploy AMD's Helios rack-scale platform across Azure data centers to power AI inference workloads. The platform bundles Instinct MI455X GPUs, 6th Gen EPYC Venice CPUs, Pensando networking hardware, and ROCm software into integrated racks. Shipments begin in the second half of 2026.

Microsoft joins Meta, OpenAI, and Oracle as early Helios adopters. Azure already runs Nvidia GPUs and its own Maia chips; Helios adds a third vendor rather than replacing either.

AMD expects server CPU revenues to grow more than 70% year over year in Q2 2026, with continued growth through 2027 as Venice processors ramp. The company doubled its 2030 server CPU total addressable market estimate from $60 billion to more than $120 billion, citing inferencing and agentic AI driving CPU demand for orchestration and data movement.

AMD's Venice CPU began production ramp in May, with more customers validating it than any earlier EPYC generation. Management emphasized customers are moving from pilot projects to multi-generation production deployments rather than one-time hardware purchases.

Rosenblatt and UBS raised price targets on AMD to $655 and $700, respectively. Jefferies analysts told clients they believed Microsoft was already using AMD's MI450 chips based on industry checks before Monday's official announcement.

• Second half 2026: Helios shipments begin to Microsoft and other hyperscalers

• Server CPU revenue growth trajectory through 2027 as Venice ramp continues

• AMD's ability to convert other Nvidia-dependent customers into full-stack Helios deployments

• Competitive response from Nvidia's Blackwell platform and Broadcom's custom AI accelerators