en.Wedoany.com Reported - HPE recently announced a series of new solutions for enterprise-grade agentic AI, designed to help customers deploy AI into production with enhanced security, control, and governance. These solutions enable the HPE AI Factory with NVIDIA to adapt to the next phase of AI development, where systems can self-adapt, evolve, and collaborate within established control frameworks.

Antonio Neri, President and CEO of HPE, stated that moving toward more autonomous AI requires a new architecture to ensure secure execution, responsible governance, and efficient scaling. Based on this, HPE and NVIDIA jointly provide end-to-end AI solutions to help customers transition smoothly and controllably from experimentation to production.

Jensen Huang, Founder and CEO of NVIDIA, noted that every layer of the computing stack is being transformed for the era of AI agents. Through the partnership with HPE, the two companies are building an AI factory based on NVIDIA Vera CPUs, accelerated infrastructure, and security software, with the goal of converting enterprise data into intelligent actions.

Enterprises across various industries are exploring the potential of AI agents in large-scale production environments to automate processes and improve decision-making quality. To simplify this process, HPE provides technologies that streamline adoption, enhance security, and boost performance.

HPE Private Cloud AI, a one-stop solution co-developed with NVIDIA, now includes new features to support the reliable deployment of agentic AI, offering greater control, visibility, and efficiency. Highlights include the use of the NVIDIA Agent Toolkit (featuring open models like NVIDIA Nemotron and tools such as NemoClaw and OpenShell), providing an "operating system" for AI agents. This enables monitoring of agent behavior, application of policies, and reduction of deployment risks.

This solution is based on the HPE ProLiant Compute DL394 Gen12 server powered by NVIDIA Vera CPUs, optimized for AI and data-intensive processing, and integrates advanced security and management features. Additionally, HPE Zerto Software can detect anomalous agent behavior and initiate continuous data protection. It also includes secure local logs for approving AI models and tools under centralized governance policies.

Since data is critical in AI projects and often becomes a bottleneck, HPE Private Cloud AI can convert unstructured data into ready-to-use AI pipelines within minutes. Using HPE Alletra Storage MP X10000, metadata and governance policies are automatically applied, reducing token-related response times by up to 20 times.

The platform also improves operational efficiency by optimizing prompt processing, increasing token throughput by up to 20%. Meanwhile, HPE Data Fabric Software expands data access for agent workflows, including support for the Model Context Protocol in Apache Airflow and an enterprise AI inventory that enriches distributed data with metadata. A standalone version is also available to simplify deployment on HPE ProLiant servers.

In terms of cost and scalability, the solution controls token expenditure, maximizes GPU utilization, and scales efficiently. New features include a unified model gateway, dynamic workload prioritization, and multi-node inference with up to 256 GPUs. Additionally, model fine-tuning (including NVIDIA Nemotron) is performed securely via NVIDIA NeMo.

HPE has also strengthened the security of its HPE AI Factory and HPE Sovereign AI Factory solutions, adding features that integrate NVIDIA Confidential Computing. This technology protects runtime models and data through encryption and cryptographic attestation, ensuring the trust chain verifies hardware, software, and data, while meeting specific regional or industry regulatory requirements.

Furthermore, technologies such as NVIDIA BlueField and NVIDIA DOCA enable zero-trust architectures, real-time threat detection, and network encryption, protecting workloads, agents, and data without compromising performance.

On the infrastructure front, the HPE AI Factory and its sovereign versions will adopt advanced NVIDIA components, including RTX PRO 6000 Blackwell Server Edition GPUs, Ethernet Spectrum-X, BlueField-3 DPUs, and ConnectX-8 SuperNICs. Based on NVIDIA reference architectures, this solution covers the entire process from development to large-scale deployment, integrating software like NVIDIA AI Enterprise and solutions from HPE Unleash AI ecosystem partners.

Finally, HPE has expanded its portfolio with new high-performance systems, including the NVIDIA Vera Rubin NVL72 by HPE, the HPE Compute XD700 with NVIDIA HGX Rubin NVL8, the HPE Cray Supercomputing GX240 blade equipped with NVIDIA Vera CPUs, and the HPE Cray Supercomputing GX5000 supporting NVIDIA Quantum-X800 InfiniBand.

This article is compiled by Wedoany. All AI citations must indicate the source as "Wedoany". If there is any infringement or other issues, please notify us promptly, and we will modify or delete it accordingly. Email: news@wedoany.com