HPE announced a multi‑year global networking contract with Oracle that will see HPE‑Juniper routers and switches installed in what the company calls one of the industry’s largest AI cloud infrastructure projects.
Deal details and technology roadmap
During HPE’s third‑quarter earnings call, CEO Antonio Neri said Oracle will use HPE‑Juniper networking gear across its AI data centers. The agreement covers deployment of HPE’s QFX data‑center switches, which rely on Broadcom’s Tomahawk silicon, and the PTX routing platform built on HPE’s own silicon.
According to the filing, the network will be based on Juniper’s Express 5 ASIC, a design HPE introduced a few years ago. Neri described the silicon as “unique” and noted that only a handful of vendors have achieved similar scale.
Executive Vice President of Oracle Cloud Infrastructure Mahesh Thiagarajan said the HPE‑Juniper solution “has been an important part of our cloud architecture and now plays a key role in our AI clusters, regional data centers, and edge networks with the scale, performance, and reliability our customers expect.”
Related: Pythian’s AI Model Drives Million-Dollar Outcomes
Strategic implications for both companies
The contract is expected to showcase HPE’s ability to deliver gigawatt‑class networking capacity, a claim the company has highlighted in past product launches. Neri said the deal could act as a high‑profile example that drives further AI‑focused networking sales for HPE.
He added that the partnership proves HPE has a “scalable set of products that deliver the performance with the AI capabilities that everybody is now looking for with self‑driving network operations.” The CEO also pointed to HPE’s early market entry with 1.6‑terabit routing capabilities that use air‑cooled and direct liquid‑cooling technologies.
While Oracle has not responded to requests for comment, the press release indicates the networking hardware will be integrated into AI clusters that require high bandwidth and low latency. The architecture is described as “scale‑out,” suggesting modular expansion as demand grows.
Compared with earlier AI infrastructure rollouts, this deployment appears larger in power and scope. Past projects have often focused on single‑region data centers, whereas the current plan spans multiple regional sites and edge locations, reflecting a broader commitment to AI workloads.
Industry observers note that the timing aligns with a surge in demand for AI‑ready cloud services. Companies are seeking networking solutions that can handle the intense data movement typical of large language model training and inference.
Related: Dell Matches HPE’s 30-Day Quote Validity
HPE’s statement that the deal will generate “significant interest” in its networking portfolio shows the competitive pressure to offer high‑performance, AI‑optimized hardware. The company’s emphasis on time‑to‑market suggests it hopes to stay ahead of rivals that may be slower to introduce comparable silicon.
Analysts will likely watch how quickly Oracle can bring the new equipment online, as deployment speed can affect both cost efficiency and customer adoption. The contract’s multi‑year nature gives both firms a runway to refine the architecture as AI workloads evolve.
In the meantime, HPE’s marketing materials highlight the partnership as a validation of its strategy to embed AI capabilities across the entire product line, not just in campus or branch solutions.
The agreement marks a notable expansion of HPE’s relationship with Oracle, building on previous collaborations that involved cloud infrastructure components. As the AI market matures, such large‑scale networking deals may become a benchmark for future cloud provider partnerships.
