Hewlett Packard Enterprise has won a $1.2 billion order from cloud provider Vultr for AI server racks built around AMD's Helios platform. The deal, confirmed this week, is one of the largest single AI infrastructure orders HPE has disclosed and signals that demand for high-density compute isn't letting up.
What Vultr is buying
Vultr is purchasing AMD Helios AI racks from HPE. Helios is AMD's rack-scale system designed for large-scale AI training and inference, and HPE is integrating those racks into its own server lineup. The $1.2 billion figure covers the hardware order, though neither company has said how many racks are involved or where they'll be deployed.
Vultr operates a global cloud computing platform and has been expanding its GPU-based offerings. Buying racks at this scale suggests the company is building out capacity for customers who need to run AI models without building their own data centers.
Why HPE needs this win
HPE has been trying to position itself as a major player in AI infrastructure, competing with rivals like Dell and Supermicro. A $1.2 billion order from a single customer is a meaningful vote of confidence, and it could help HPE's financial growth if the hardware ships on schedule and leads to follow-on business.
The company hasn't said when revenue from the order will be recognized. Large infrastructure deals are often booked over multiple quarters as equipment is delivered and installed.
AMD's quiet role
AMD has been pushing its Instinct accelerators and Helios rack architecture as an alternative to Nvidia's dominant AI hardware. This order puts AMD silicon into Vultr's cloud, giving the chipmaker another reference customer in the AI infrastructure market. HPE is the systems integrator, but AMD's platform is the engine inside the racks.
Neither AMD nor Vultr has commented publicly on the specific deployment timeline.
What happens next
The order still has to be fulfilled. HPE will need to manufacture, configure, and ship the racks to Vultr's data centers. That process typically takes months, and any supply chain issues could delay delivery. For now, the deal stands as a concrete example of the money still flowing into AI hardware — and a reminder that the fight for that business is far from settled.
