Hewlett Packard Enterprise (HPE) has secured a $1.2 billion order from cloud infrastructure provider Vultr for AMD Helios AI Rack by HPE systems, marking HPE’s first order for the new platform.
Vultr plans to deploy the systems across its cloud data center locations in the United States to support service providers and enterprises running AI model training and inference workloads.
What the Vultr Order Includes
Each AMD Helios AI Rack integrates 72 AMD Instinct MI455X GPUs with AMD EPYC “Venice” CPUs, AMD Pensando Vulcano AI NICs and AMD ROCm software. HPE’s configuration also includes six Juniper Networking QFX5252 scale-up Ethernet switch trays per rack, connecting the GPUs through high-bandwidth, low-latency Ethernet, alongside rack-scale engineering and liquid-cooling support.
For Vultr, the infrastructure is intended to expand capacity for increasingly demanding AI workloads. In HPE’s announcement, Vultr said demand for high-performance AI infrastructure is outpacing available capacity, making compute density, memory capacity, networking performance and energy efficiency increasingly important for large training and inference deployments.
The order also builds on nearly three years of collaboration between Vultr and Juniper Networks, which is now part of HPE. That existing networking relationship gives HPE a broader role in the deployment as Vultr adds rack-scale compute, networking and cooling infrastructure.
Why the Deal Matters for HPE
HPE announced the Vultr order alongside an upgraded outlook for its networking business. The company raised its fiscal 2027 Networking segment revenue-growth outlook to the high-teens-to-low-20s percentage range and expects its Data Center Networking revenue to grow at a low-to-high-50s percent compound annual growth rate through fiscal 2029.
HPE also increased its target for annual cost savings related to its Juniper Networks acquisition to $800 million by the end of fiscal 2028, up from its previous target of at least $600 million.
A Broader Move Toward Rack-Scale AI Systems
The Helios platform reflects a broader shift in AI infrastructure from individual accelerator purchases toward integrated rack-scale systems. Its design combines GPUs, CPUs, networking, software and cooling within a coordinated architecture intended for large training and inference workloads.
HPE’s announcement did not specify how many Helios racks are included in Vultr’s $1.2 billion order or provide a detailed deployment schedule. The disclosed value therefore shows the commercial size of the agreement but does not by itself indicate the total amount of computing capacity Vultr will add.
With Vultr as its first Helios customer, HPE now has an early commercial deployment for its AMD-based rack-scale platform while Vultr gains additional infrastructure for expanding AI training and inference services.
