HPE Vultr $1.2 Billion AI Order, Networking Outlook Raised

Hewlett Packard Enterprise just landed one of the biggest AI infrastructure deals of the year, and it used the moment to tell Wall Street the AI boom still has a long runway. On September 30, HPE announced the HPE Vultr $1.2 billion AI order: cloud company Vultr is buying its new AMD Helios AI server racks, while raising its long-term growth forecast for its networking business. HPE shares jumped nearly 6% in premarket trading, according to Reuters.

The HPE Vultr $1.2 billion AI order is the first order for AMD Helios AI Rack by HPE systems. Vultr, described by HPE as the world's largest privately-held cloud infrastructure company, will deploy the systems across its US data centers to serve customers running AI training and inference workloads. The announcement came during HPE's networking investor day, where executives laid out sharply higher targets for the networking unit that HPE expanded through its acquisition of Juniper Networks last year.

What HPE is actually selling Vultr

Each AMD Helios rack is a rack-scale AI computer rather than a stack of independent servers. According to HPE's announcement, every rack packs 72 AMD Instinct MI455X GPUs alongside AMD EPYC Venice CPUs and AMD Pensando Vulcano networking hardware, tied together by six HPE Juniper Networking scale-up Ethernet switch trays. HPE says the design supports open rack-scale fabrics and ships with direct liquid cooling, which matters because AI accelerators run extremely hot and dense.

The networking piece is the strategic core of the pitch. Training a large AI model requires thousands of GPUs exchanging data constantly; if the network between them is slow, expensive accelerators sit idle. That is why HPE, which competes with Dell and Super Micro in AI servers, is leaning on the Juniper portfolio to sell the complete package: compute, scale-up networking inside the rack, and the software to run it. "AI is reshaping the technology stack, making the network more strategic and driving significant new demand from enterprises and service providers," said Rami Rahim, head of HPE's networking business. The HPE Vultr $1.2 billion AI order is an early proof point for that thesis, since Vultr chose HPE specifically for the bundled networking in the racks.

For Vultr, the deal is about capacity. "Demand for high-performance AI infrastructure continues to outpace available capacity," said Vultr CEO J.J. Kardwell, adding that the Helios systems bring the memory capacity, interconnect performance, and energy efficiency needed for large training and inference jobs. The HPE Vultr $1.2 billion AI order builds on nearly three years of collaboration between Vultr and Juniper Networks that carried over into HPE after the acquisition.

The forecast upgrade is the bigger story

Beyond the single order, HPE raised its expectations for the whole networking segment. The company now expects networking revenue to grow at a compound annual rate in the high teens from fiscal 2026 through fiscal 2029, up from its previous forecast of 5% to 7% annual growth. For fiscal 2027 specifically, HPE lifted its networking growth target to the high teens to low 20s percent range, up from 14% to 17%.

HPE also increased its target for cost savings from the Juniper acquisition to $800 million in annual run-rate savings by the end of fiscal 2028, up from at least $600 million previously. To keep up with demand, the company said it doubled its networking supply purchase commitments in its fiscal third quarter compared with the prior quarter, a sign it expects the order book to keep growing. If the HPE Vultr $1.2 billion AI order is any indication, that demand is already arriving.

"AI is driving the largest infrastructure buildout in history and together with AMD and Vultr, we are building an open foundation to accelerate the next phase of AI," said HPE CEO Antonio Neri. The emphasis on "open" is a deliberate contrast with closed, single-vendor AI stacks: HPE is betting customers want Ethernet-based fabrics and mix-and-match hardware rather than being locked into one supplier's ecosystem.

Why this matters beyond Wall Street

Deals like the HPE Vultr $1.2 billion AI order are where the AI boom turns into physical reality. Every chatbot answer, generated image, and AI video starts in a data center full of racks like these, and the $1.2 billion price tag shows how much cloud providers are willing to pay to keep up with demand. That spending eventually flows into the cost and availability of the AI tools millions of people use every day, from free chatbots to the cloud GPUs that startups rent by the hour.

The competitive picture is shifting too. For years, the AI infrastructure conversation was dominated by one chipmaker's GPUs and proprietary networking. HPE's Helios push, built on AMD silicon and Ethernet-based scale-up switching, is part of a broader industry move toward open alternatives. If that bet pays off, it could mean more suppliers, more capacity, and eventually cheaper AI compute, which is good news for anyone building with these tools rather than just buying the stocks.