HPE unveils innovative enhancements to its Private Cloud service, featuring Nvidia Blackwell GPUs
HPE Strengthens Partnership with Nvidia, Introducing New AI-Focused Servers and Private Cloud Platform
HPE has made significant strides in its partnership with Nvidia, expanding its Nvidia Computing by HPE portfolio to include the latest Nvidia hardware designed for AI workloads. This move is aimed at empowering enterprises to make the most of enterprise AI deployment.
Two new HPE ProLiant Compute servers will be available: the HPE ProLiant DL385 Gen11 and the HPE ProLiant Compute DL380a Gen12. The HPE ProLiant DL385 Gen11 server supports up to two NVIDIA RTX PRO 6000 Blackwell Server Edition GPUs in a compact 2U air-cooled form factor, designed for easy deployment in existing data center racks. On the other hand, the HPE ProLiant Compute DL380a Gen12 server supports up to eight RTX PRO 6000 GPUs in a 4U chassis, with shipping scheduled for September 2025.
The intended workloads for these servers include inference for generative and agentic AI models, physical AI applications such as robotics, industrial automation, and simulation, visual computing, 3D modeling, and digital twins, as well as AI model development and deployment throughout the AI lifecycle.
The expanded portfolio also includes the newest version of HPE Private Cloud, which will support Nvidia's Nemotron, a set of open source reasoning models based on Meta's Llama, and Cosmos Reason, a vision language model intended for robotics and physical AI devices. This new version will integrate the newest NVIDIA AI models and NVIDIA Blueprints (including Nemotron, Cosmos Reason, and VSS 2.4 Blueprint) for streamlined AI application development and deployment.
Moreover, the HPE Private Cloud AI platform is designed to accelerate the development and deployment of both agentic AI (autonomous AI agents) and physical AI workloads. It incorporates advanced multi-layered security features such as HPE Integrated Lights Out (iLO) 7 Silicon Root of Trust, secure enclave, quantum-resistant firmware signing, ensuring tamper-resistant protection. The private cloud AI platform will be available in the second half of 2025 alongside the new ProLiant servers.
In addition, HPE GreenLake now supports agentic AI, providing organizations with baked-in investment protection to ensure access to the latest Nvidia GPUs as they become available. This means customers can opt for air-gapped Private Cloud AI services if their data privacy rules demand strict controls.
In summary, HPE’s expanded Nvidia Computing portfolio strengthens support for AI workloads from experimentation to production, combining new Nvidia Blackwell GPUs in versatile server form factors with a secure private cloud AI platform co-developed with Nvidia to serve enterprises’ growing AI compute and deployment needs.
[1] HPE Press Release: [Link to the Press Release] [2] Nvidia Press Release: [Link to the Nvidia Press Release] [3] HPE Product Page: [Link to the HPE Product Page] [4] Nvidia Product Page: [Link to the Nvidia Product Page]
The HPE ProLiant DL385 Gen11 server, introduced as part of HPE's expanded Nvidia Computing portfolio, supports up to two NVIDIA RTX PRO 6000 Blackwell Server Edition GPUs and is designed for enhanced cybersecurity infrastructure in data-and-cloud-computing environments.
With the new HPE Private Cloud AI platform, scheduled for release in the second half of 2025, organizations can expect advanced technology integration, including NVIDIA's Nemotron and Cosmos Reason, for secure data processing and facilitating physical AI devices.
