Where HPE goes, Nvidia is rarely far away. HPE introduces new AI Factory solutions and an advanced agentive AI framework for hybrid cloud management during its Discover conference.
During Discover, HPE has stepped up its game with new AI solutions. These are built around the concept of “AI factories”. This concept was conceived by Nvidia to make itself even more indispensable in the AI market, and successfully so, as providers are eager to jump on board. Nvidia is therefore rarely far away in product announcements.
An AI factory can be seen as a modular system that integrates computing power, storage, and network capacity for organizations wanting to deploy AI on a large scale. HPE and Nvidia promote themselves as the builders of such AI factories.
HPE’s infrastructure incorporates Nvidia’s Blackwell GPU’s into its own compute solutions and is available in eight different configurations, including the new XD690 system with eight Blackwell Ultras.
Inseparable Bond with Nvidia
HPE builds further on the Private Cloud AI offering launched last year; also in collaboration with Nvidia. This solution supports hybrid and secure AI implementations and runs on HPE ProLiant Gen12 servers. The system supports multiple GPU generations and offers air-gapped management and multi-user functionality. Through a “try and buy” program, customers can test the solution in Equinix data centers.
read also
AI according to Dell, HPE and Lenovo: same story, available in three colors
Additionally, HPE introduces the Alletra Storage MP X10000, which receives support for the Model Context Protocol (MCP). HPE OpsRamp Software acts as an observability layer in the infrastructure and is validated for the Nvidia Enterprise AI Factory. For governments and sovereign institutions, HPE offers specific configurations with extra focus on data sovereignty, a sensitive issue in the European market.
Through HPE Services, the company supports customers in the design, construction, and management of AI infrastructure. Customers can also count on financial flexibility with, among other things, deferred payments for Private Cloud AI.
GreenLake Intelligence
In addition to AI infrastructure, HPE also renews the management of hybrid IT with GreenLake Intelligence. This technology forms an agentive AI framework that enables real-time analysis and action across multiple IT domains. Part of this is the integration of AI agents for observability, cloud costs, sustainability, workload optimization, and networking.
GreenLake Intelligence uses OpsRamp for IT automation. This software supports incident management, root cause analysis, and capacity planning. For storage, HPE introduces MCP support on the Alletra Storage MP X10000, enabling AI workflows to be better fed with unstructured data and metadata insights.
Network Agent
Network management also gets an update via HPE Aruba Networking Central. This solution includes an agentive mesh and a network copilot that can autonomously analyze and solve network problems.
HPE CloudOps Software bundles OpsRamp, Morpheus, and Zerto into one suite for managing multi-cloud infrastructures. Additionally, CloudPhysics Plus offers an assessment tool for hybrid modernization. Through the new Cloud Commit program, customers can enter into long-term purchasing commitments with associated benefits.
Most new solutions will be available in the second half of 2025. The Compute XD690 is expected in October, while the beta of GreenLake Copilot is planned for the third quarter.