HPE Accelerates Secure and Scalable Production AI Through New Innovations with NVIDIA

• New AI servers, turnkey systems, services, and software, powered by NVIDIA Blackwell and Rubin acceleration, are designed to help achieve secure, predictable, and reproducible AI deployment outcomes.

• HPE Alletra Storage MP X10000 has become the first to achieve NVIDIA-Certified Storage object-based platform certification, demonstrating that it meets the performance, security, and scale requirements demanded by AI.

* This release is a Japanese summary based on an English press release issued by HPE (Headquarters: Houston, Texas, USA) at NVIDIA GTC2026, held in San Jose, California, on March 16, 2026 (local time). Please refer to the original (full) release here.

HPE today announced a significant expansion of the NVIDIA AI Computing by HPE portfolio, redefining how enterprises deploy, operate, and scale AI. Through close partnership and co-engineering with NVIDIA, HPE provides advanced products and solutions consisting of integrated, verified systems that reduce time to value for AI while addressing scale, security, and governance requirements.

Antonio Neri, President and CEO of HPE, stated:

"Speed, scale, and reliability are key to the race toward AI adoption and utilization. HPE's leadership in cloud, networking, and AI enables our customers to run AI securely and efficiently at unprecedented scale. Through our collaboration with NVIDIA, HPE offers turnkey AI factories and networks, helping our customers turn their AI ambitions into valuable initiatives."

Jensen Huang, Founder and CEO of NVIDIA, stated:

"NVIDIA and HPE are setting a new standard for enterprise AI infrastructure. HPE's established leadership in private cloud, networking, and secure on-premises systems shows they are well-positioned to elevate AI into a core enterprise function. Together, we are building the AI factories and AI Grid that serve as the foundation for embedding intelligence into every workflow."

Supporting Smooth AI Adoption through Extended Secure, Reproducible, and Efficient AI Architectures

HPE is expanding HPE Private Cloud AI, a turnkey enterprise AI factory co-designed with NVIDIA, to improve performance, scalability, and flexibility in enterprise inference. Enhancements announced include:

• Addition of network expansion racks, enabling the deployment of HPE Private Cloud AI scalable up to 128 GPUs.

• Large systems of HPE Private Cloud AI offer air-gapped configurations to meet the demand for AI deployment in secure, disconnected, sovereign environments.

• The HPE ProLiant Compute DL380a Gen12 server and HPE Private Cloud AI systems based on it are seeking certification for Fortanix Confidential AI, a joint solution leveraging NVIDIA Confidential Computing to support secure on-premises deployment of AI models and safe processing of sensitive data.

• CrowdStrike is bringing agentic security to HPE Private Cloud AI, helping organizations operate agentic AI safely and at scale by providing AI-driven threat detection and response to protect AI infrastructure, models, and AI agents running in enterprise environments.

• HPE Private Cloud AI provides a pre-configured hardware and software stack featuring the latest NVIDIA AI Enterprise software and Blueprints. Updated NVIDIA AI-Q Blueprints for AI agents and new NVIDIA Omniverse for digital twins help developers build highly customizable AI agents that they can own, verify, and control themselves.

• HPE is advancing updates to HPE Private Cloud AI, the latest HPE ProLiant servers, and HPE AI factory solutions to support NVIDIA Nemotron open models. NVIDIA Nemotron, part of the NVIDIA Agent Toolkit, facilitates secure on-premises and sovereign infrastructure deployment and helps quickly realize scalable, production-ready outcomes.

• NVIDIA RTX™ PRO 6000 Blackwell Server Edition GPUs can be integrated into HPE Private Cloud AI Developer System, Small and Large configurations, and HPE's AI factory solution configurations.

New AI Solutions Added for Retail, Biomedical Research, and Manufacturing

New multi-workload solutions co-designed with NVIDIA facilitate the deployment of various AI use cases, such as edge AI, retail shopping assistants, video search and summarization, and biomedical research.

These solutions are based on the NVIDIA Enterprise Reference Architecture, combining HPE ProLiant Compute servers with NVIDIA accelerated computing, NVIDIA Spectrum-X Ethernet networking, BlueField DPUs, and Connect-X NICs. They also incorporate NVIDIA AI Enterprise software, NVIDIA CUDA-X libraries, Blueprints, Confidential Computing, NVIDIA Multi-Instance GPU (MIG), and NVIDIA virtual GPU (vGPU) technologies, along with chip-to-cloud security and AI-driven automation via HPE Compute Ops Management.

To complement multi-workload solutions in retail, biomedical research, and manufacturing, HPE will integrate NVIDIA RTX PRO 4500 Blackwell Server Edition GPUs into HPE ProLiant servers for edge deployments, small language models, vector databases, and data analytics workloads. HPE is also developing new solutions leveraging the NVIDIA RTX 4500 Blackwell GPU and integrating the NVIDIA Retail Shopping Assistant Blueprint to streamline retail deployments. Furthermore, HPE is expanding its portfolio of HPE ProLiant Compute servers with new NVIDIA RTX PRO 6000 Blackwell Server Edition GPUs.

Accelerating AI Networking and Large-Scale AI through Innovations at NVIDIA GTC 2026

At NVIDIA GTC, HPE announced a networking solution combining HPE Juniper Networking routers and coherent optical technology to connect distributed AI environments for service providers, sovereign entities, and large enterprises. HPE also announced a new generation of systems adopting the NVIDIA Vera Rubin architecture, designed to handle AI workloads with advanced requirements in large-scale environments and sovereign AI factories.

Strengthening AI Data Pipelines through Close Collaboration with NVIDIA

As AI infrastructure production deployments progress, it has become clear that data pipelines—particularly inference context—are performance bottlenecks. HPE is working closely with NVIDIA to accelerate every stage of the AI data lifecycle, from data ingestion and vectorization to inference and recovery. Furthermore, HPE is evolving the HPE Alletra MP X10000 to centralize intelligent data processing and optimize the ingestion, processing, and delivery of data for AI workloads.

The HPE Alletra MP X10000 object-based system has become the first to achieve NVIDIA-Certified Storage certification at the Foundation level. This certification indicates that NVIDIA has verified and benchmarked performance for loads of up to 128 GPUs, validated enterprise-grade availability and reliability features, and demonstrated that the storage layer efficiently delivers data to high-performance computing resources, resulting in faster model training, reduced latency during inference execution, and improved overall resource utilization.

To further strengthen collaboration with NVIDIA, HPE will support the new NVIDIA STX rack-scale reference architecture and develop new AI storage solutions leveraging NVIDIA Vera Rubin, BlueField-4, Spectrum-X networking, Connect-X NICs, and NVIDIA AI software.

New HPE Services, Financing, and Unleash AI Ecosystem Solutions to Support Smooth AI Adoption

• HPE Services is supporting the systematic and scalable enterprise deployment of agentic AI through the new Agents Hub. The hub is developing and verifying agents leveraging NVIDIA Nemotron™ models and is establishing reusable patterns to assist in the deployment and production operation of intelligent agents.

• To help optimize resource utilization and ROI across shared HPE AI factories, HPE Services and Protopia AI are jointly developing a Blueprint for realizing highly reliable, multi-tenant AI factories that allow sensitive data to be leveraged without exposure in highly regulated and sovereign environments.

• HPE Financial Services is supporting the advancement of AI and modernization projects through the new 90/9 Advantage financing program. There are no payments for the first 90 days, followed by nine months of payments at 1% of the monthly lease cost. This program covers the networking, hybrid cloud, and compute server portfolios.

• Availability

• Support for NVIDIA RTX PRO 4500 Blackwell Server Edition GPUs across the HPE ProLiant Compute server portfolio is scheduled to begin in Q1 and Q2 2026.

• Support for NVIDIA RTX PRO™ 6000 Blackwell Server Edition GPUs in air-gapped HPE Private Cloud AI Medium and Large, Developer System, Small, and Large configurations, as well as NVIDIA AI-Q Blueprints and NVIDIA Omniverse Blueprints, is available now.

• New network expansion racks enabling scale-up of up to 128 GPUs in HPE Private Cloud AI are scheduled to be available in Q3 2026.

• The secure Blueprint by HPE and Protopia for realizing reliable AI factories is scheduled to be available in Q2 2026.

• Support for Fortanix Confidential AI on HPE ProLiant DL380a Gen12 systems is scheduled for Q3 2026.

■ About HPE

HPE (NYSE: HPE) is a leading global edge-to-cloud company that helps organizations accelerate outcomes by unlocking value from all of their data, everywhere. Built on decades of reimagining the future and innovating to advance the way people live and work, HPE delivers unique, open, and intelligent technology solutions as a service. With products spanning Cloud Services, Compute, High Performance Computing & AI, Intelligent Edge, Software, and Storage, HPE provides a consistent experience across all clouds and edges, helping customers develop new business models, engage in new ways, and increase operational performance. For more information, visit: www.hpe.com

■ Press Room

https://www.hpe.com/jp/ja/newsroom/press-hub.html

# # #

Other product names used in this press release are for identification purposes only and may be trademarks of their respective companies.

[Customer Inquiries]

Customer Information Center

TEL: 0120-268-186 or 03-6743-6370

Website: https://www.hpe.com/jp/

© 2026 Hewlett Packard Enterprise Development LP

The terms and conditions of any warranty for products and services are set forth in the express warranty statements accompanying such products and services. Nothing herein should be construed as constituting an additional warranty. The Company shall not be liable for technical or editorial errors or omissions contained herein.