Hewlett Packard Enterprise (HPE) is undergoing a significant transformation from a traditional technology provider to a principal architect of sovereign, AI-native infrastructure aimed at enterprises and cloud service providers. This shift was prominently showcased at HPE Discover 2025, held in Barcelona, where keynotes, breakout sessions, and media briefings laid out a new blueprint for AI-ready data centres that are modular, compliant, intelligent, and seamlessly integrated across compute, networking, storage, security, and operations.
At the core of this transformation is HPE’s initiative to create “AI factories.” These environments are designed to provide compliant, high-performance spaces where enterprises can develop, validate, and deploy next-generation AI models. HPE’s partnerships with key industry players such as NVIDIA, AMD, Broadcom, CrowdStrike, and Fortanix further reinforce this strategy, complemented by a revamped partner ecosystem under the HPE Partner Ready Vantage programme.
HPE’s sovereign AI factories create compliant, high-performance environments where enterprises can develop, validate, and deploy next-generation models.
Antonio Neri, President and CEO of HPE, emphasized the importance of this moment, calling it a generational inflection point. “With the new AMD ‘Helios’ and our purpose-built HPE scale-up networking solution, we are providing our cloud service provider customers with faster deployments, greater flexibility, and reduced risk in how they scale AI computing in their businesses,” he stated, highlighting the comprehensive nature of HPE’s strategy.
Central to HPE’s vision is the development of AI Factory Labs, with the first two unveiled in Grenoble, France, and London. The AI Factory Lab in Grenoble is designed for rigorous experimentation, validation, and compliance testing, integrating NVIDIA AI Enterprise software, HPE Alletra storage, ProLiant servers, and accelerated compute solutions. This facility aims to enable European enterprises to test AI workloads while adhering to regulatory requirements.
HPE positions these labs as critical bridges between innovation and governance, allowing enterprises to refine and certify their AI models before deploying them into production environments. Neri noted that the collaboration with NVIDIA delivers “true full-stack AI infrastructures that provide enterprises with a greater range of performance for more diverse workloads.”
Networking as the Backbone of AI
Networking plays a vital role in this new infrastructure. Following its acquisition of Juniper Networks, HPE introduced a unified, AI-native networking fabric that integrates Aruba Central and Juniper Mist into a single operating experience. Rami Rahim, Executive Vice President of HPE Networking, described this integration as strategic and inevitable, stating that the distinction between the two platforms will eventually disappear as customers benefit from an identical experience across both.
Our unified, AI-native fabric ensures customers gain the same advanced experience across Aruba and Mist, with no platform left behind. – RAMI RAHIM, EVP, President & General Manager, Networking, HPE
The new QFX5250 switch, equipped with Broadcom’s Tomahawk 6 silicon, was also showcased, delivering 102.4 Tbps throughput and supporting Ultra Ethernet Transport. This innovation positions HPE to extend Ethernet’s role in both large-scale GPU clusters and edge inference use cases.
HPE’s compute announcements reflect a broad spectrum of AI workloads, from high-density inference to exascale-class training. The introduction of the NVIDIA GB200 NVL4 platform aims at efficient large language model serving, while the anticipated AMD Helios AI rack, set for release in 2026, promises an open, exascale-class platform for training trillion-parameter models.
By complementing NVIDIA’s Spectrum-X with its own Juniper-based Ethernet solutions, HPE is positioning itself as one of the few vendors capable of creating end-to-end AI factories, spanning lab, core, and cloud environments.
As AI models increase in complexity, the preparation and movement of data become crucial for efficient training. To that end, HPE introduced the Alletra Storage MP X10000 Data Intelligence Node, a platform that enhances storage capabilities to support real-time analytics and accelerate data preparation tasks. In a similar vein, HPE has updated its StoreOnce 7700 array and expanded recovery capabilities to bolster cyber resilience, which is essential given the evolving threat landscape.
HPE’s emphasis on security extends throughout its architecture. “Security has to be baked into every piece of the infrastructure. You cannot bolt it on afterwards,” Russo stated, reflecting the company’s philosophy. This holistic approach to security includes Confidential Computing initiatives, post-quantum-resilient features, and integration with CrowdStrike’s AI-native security platform.
HPE’s hybrid cloud strategy also plays a critical role, with Morpheus serving as the unified control plane across its portfolio. This integration streamlines complex transitions, particularly in environments that include virtual machines, containers, and legacy workloads. Maeve Culloty, CEO of HPE Financial Services, described GreenLake’s evolution as foundational, evolving from a financial model to a comprehensive platform for managing all types of infrastructure.
Financing and lifecycle services reduce barriers to AI adoption, helping customers modernise while managing evolving hybrid infrastructures. – MAEVE CULLOTY, President & CEO, HPE Financial Services
HPE’s announcements signal a strategic commitment to full-stack, sovereign AI infrastructure at an industrial scale. By unifying AI Factory Labs, AI-native networking, rack-scale compute, intelligent storage, integrated security, and lifecycle-aware financing, HPE aspires to be a leading architect of AI-ready digital infrastructure. This approach aims to not only support AI workloads but also to industrialise the foundational infrastructure that will shape the future of enterprise computing.
See also
Tesseract Launches Site Manager and PRISM Vision Badge for Job Site Clarity
Affordable Android Smartwatches That Offer Great Value and Features
Russia”s AIDOL Robot Stumbles During Debut in Moscow
AI Technology Revolutionizes Meat Processing at Cargill Slaughterhouse
Seagate Unveils Exos 4U100: 3.2PB AI-Ready Storage with Advanced HAMR Tech
