Hewlett Packard Enterprise is expanding its role in the AI infrastructure market after securing a $1.2 billion order from cloud provider Vultr, marking a significant deployment of AMD-powered rack-scale systems. The HPE $1.2 billion Vultr AI order is one of the largest recent investments in AI infrastructure.
The order covers HPE systems based on AMD’s Helios architecture, which brings together AI accelerators, CPUs, networking and other infrastructure components in a rack-scale design. Vultr plans to use the systems to expand its AI cloud capacity as demand grows for training and inference workloads.
The deal also arrives as HPE raises its long-term expectations for its Networking business, reflecting the growing importance of high-performance networking in large AI data centers.
HPE Secures $1.2 Billion AI Infrastructure Order From Vultr
The agreement with Vultr gives HPE a major customer deployment for its AMD Helios-based infrastructure. Rather than supplying conventional servers individually, HPE is providing a larger integrated system designed to support demanding AI workloads across connected racks. For Vultr, the investment adds substantial capacity to its cloud infrastructure as enterprises and AI developers look for access to increasingly powerful computing resources.
HPE said the order is its first for an AMD Helios AI Rack system. The infrastructure combines AMD compute technology with HPE networking and data-center capabilities, creating a system designed around the demands of large-scale AI workloads.
Vultr has been expanding its AI infrastructure strategy as demand grows for accelerated computing. Its cloud platform supports workloads including AI model training, inference, fine-tuning and newer agentic AI applications.
AMD Helios Brings Compute and Networking Into One AI System
The Helios architecture reflects a broader change in AI infrastructure. Large AI workloads increasingly require multiple accelerators to communicate rapidly, which makes networking performance a critical part of overall system performance rather than a separate consideration.
AMD’s Helios platform combines 72 AMD Instinct MI455X GPUs with AMD EPYC processors, AMD Pensando networking technology and ROCm software. The architecture is designed for large-scale AI training, inference and fine-tuning, with the individual components engineered to operate as part of a rack-scale system.
That approach is particularly relevant for cloud providers such as Vultr. Instead of building AI clusters from disconnected hardware components, providers can deploy systems designed around the communication requirements of large accelerator clusters.
HPE Raises Its Networking Growth Outlook
HPE’s Vultr announcement comes at the same time the company is placing a larger financial target on its Networking business. The company expects Networking revenue to grow at a high-teens compound annual growth rate between fiscal 2026 and fiscal 2029, with operating margins expected to remain in the mid-to-high 20% range during the later part of that period.
The outlook reflects HPE’s broader strategy following its acquisition of Juniper Networks. The company now has a larger portfolio spanning data-center networking, routing, campus and branch infrastructure, and security.
AI is an important part of that strategy because modern AI clusters generate enormous volumes of data traffic between processors. As clusters grow, the network connecting those processors can become a major determinant of how efficiently the infrastructure operates.
HPE’s Networking Business Is Already Expanding
The new growth target comes after a period of strong performance in HPE’s Networking segment. In its fiscal third quarter, HPE reported $2.9 billion in Networking revenue, representing a substantial year-over-year increase.
Data Center Networking revenue increased sharply during the quarter, while Routing also recorded significant growth. The figures include contributions from the broader networking portfolio following the Juniper acquisition.
That performance gives HPE a larger base from which to pursue AI infrastructure opportunities. The company is increasingly positioning networking as a core component of modern data centers rather than simply an additional layer around servers and storage.
Vultr Expands Its Push Into AI Cloud Infrastructure
Vultr’s investment comes as cloud providers compete to offer more powerful AI infrastructure to developers and businesses. The company has been building out access to accelerated computing and has positioned AMD technology as one part of its broader AI infrastructure offering.
The company has highlighted AMD Helios for workloads such as model training, inference, fine-tuning and agentic AI. These applications can place different demands on computing resources, particularly when models become larger and applications require more real-time processing.
The HPE order therefore gives Vultr additional infrastructure for a market that is moving beyond traditional cloud workloads. AI customers increasingly need dedicated accelerator capacity, high-bandwidth networking and infrastructure capable of handling large distributed workloads.
AI Infrastructure Is Moving Beyond the GPU
The HPE-Vultr deal also highlights a larger change taking place across the AI hardware market. GPUs remain central to AI computing, but large deployments require much more than accelerators alone.
AI infrastructure now involves processors, networking, memory, storage, power delivery, cooling and software working together. At large scale, weaknesses in any one of those areas can affect the performance of the entire cluster.
Rack-scale architectures such as AMD Helios are designed around that reality. Instead of treating each server as an independent unit, they treat the rack as an integrated computing system.
For HPE, this creates an opportunity to sell more of the infrastructure surrounding AI compute. For cloud providers, it offers a path toward deploying large AI clusters without designing every component from scratch.
What the $1.2 Billion Deal Means for HPE and Vultr
The $1.2 billion Vultr order gives HPE a significant customer deployment for its AMD Helios infrastructure while reinforcing the company’s push into AI-focused networking.
Vultr, meanwhile, gains access to additional rack-scale AI infrastructure as it expands its cloud platform and responds to demand for accelerated computing.
The deal also shows how the AI infrastructure market is changing. The competition is no longer only about which company provides the fastest AI accelerator. Networking, system integration, cooling, power and software are becoming equally important as AI clusters become larger and more complex.
HPE’s higher networking growth expectations suggest the company sees that shift as a long-term opportunity.
Sources
- Reuters — HPE boosts networking growth outlook, gets $1.2 billion AI order from Vultr
https://www.reuters.com/business/hpe-boosts-networking-growth-outlook-gets-12-billion-ai-order-cloud-firm-vultr-2026-09-30/ - HPE — HPE secures first AMD Helios order in $1.2 billion deal with Vultr
https://www.hpe.com/us/en/newsroom/press-release/2026/09/hpe-secures-its-first-amd-helios-order-in-12-billion-deal-with-vultr.html - HPE — Networking priorities and long-term growth outlook
https://www.hpe.com/us/en/newsroom/press-release/2026/09/hpe-to-outline-networking-priorities-for-long-term-value-creation.html - AMD — Helios Rack-Scale AI Infrastructure
https://www.amd.com/en/products/rackscale-solutions/helios.html - Vultr — AMD Instinct MI455X GPU and Helios Rack-Scale AI Solution
https://blogs.vultr.com/amd-instinct-mi455x-gpu-helios-rackscale-solution-ai
