Hewlett Packard Enterprise has expanded its NVIDIA AI Computing by HPE portfolio, introducing new tools designed to support secure and scalable “AI factories,” improve data center connectivity for large-scale AI workloads, and launch its first EU-based AI factory lab for testing sovereign AI systems.

The expansion reflects a strengthening collaboration between HPE and NVIDIA as demand grows for infrastructure that can support cross-border AI operations, meet regulatory requirements, and run increasingly diverse workloads.

HPE president and CEO Antonio Neri said the companies are advancing “full stack AI infrastructures” to support secure AI development at scale. NVIDIA founder and CEO Jensen Huang described AI factories as the manufacturing engines of a new industrial era, noting that the joint lab will help organizations securely and at scale turn data into value.

As part of the expansion, HPE and NVIDIA plan to open an AI Factory Lab in Grenoble, France, providing organizations with a sovereign, air-cooled environment to test and refine their workloads. The facility integrates NVIDIA AI Enterprise software, HPE servers, HPE Juniper Networking systems, NVIDIA accelerated computing and networking technologies, and HPE Alletra storage.

HPE is also partnering with Carbon3.ai to establish a Private AI Lab in London, supporting UK enterprises that deploy solutions built on HPE Private Cloud AI and NVIDIA infrastructure.

To meet rising demand for sovereign AI environments in Europe, HPE has expanded its Private Cloud AI portfolio with new GPU options, hardened NVIDIA AI Enterprise software for air-gapped deployments, GPU fractionalization through Multi Instance GPU technology, and updated Datacenter Ops Agents. The company has also introduced system designs that align with country-specific compliance rules and new reference architectures with stronger security features.

HPE is improving AI factory performance through enhanced networking built on NVIDIA Spectrum X Ethernet and BlueField 3 DPUs, along with its Juniper Networking portfolio for high-capacity, low-latency links across distributed clusters.

To strengthen data pipelines, HPE launched the Alletra Storage MP X10000 Data Intelligence Nodes, which bring NVIDIA accelerated computing into the data layer for real-time processing. The company also introduced the NVIDIA GB200 NVL4 by HPE, a compact, lower-power platform designed for high-performance AI inference with dense Grace CPU and Blackwell GPU configurations.

On the security side, HPE has selected CrowdStrike as the cybersecurity platform for its Private Cloud AI customers, extending unified protection across hybrid and multicloud environments. It is also partnering with Fortanix to support NVIDIA Confidential Computing, enabling secure, sovereign AI workloads on HPE Private Cloud AI and HPE ProLiant servers with NVIDIA RTX PRO 6000 Blackwell GPUs, particularly for organizations in highly regulated sectors across EMEA.