4 Takeaways from Antonio Neri's Keynote at HPE Discover 2025

Hewlett Packard Enterprise recently held its HPE Discover user event in Las Vegas. One of the highlights of the event was the keynote delivered by Antonio Neri, president and CEO at HPE, in the iconic Sphere for the second consecutive year.

Some keynotes are visionary, and others are technical. I thought Neri's keynote delivered a compelling vision of an AI-enabled future mixed with HPE innovation and real customer use cases.

Below are four takeaways that define HPE's strategic direction and its role in an increasingly AI-driven world.

It's All about AI, but that Comes in Many Flavors

Early in Neri's keynote, he made the statement that "we are standing at the edge of a new Renaissance, the AI era," which is characterized by new possibilities that will change the way we live, work and create. The internet had a similar effect, as it refined almost every aspect of our lives. I expect AI to do the same, in a much bigger way.

The AI era isn't merely about smarter machines but about a fundamental shift across three distinct yet interconnected facets of AI:

Related:Nvidia Announces New and Expanded Products at SIGGRAPH 2025

Related:2 Ways to Think about AI and Networking -- Without the Hype

Neri said these three dimensions of AI represent a "profound shift" leading to an "AI-attributed business and society transformation," underscoring the urgency for enterprises to embrace this new reality.

HPE's Core Strategy: Build a Modern IT Foundation with Networking, Hybrid Cloud and AI

At last year's HPE Discover event, Neri's keynote had a good amount of network content. He was emphatic that HPE was shifting to a network-first company, and his narrative in the 2025 keynote reflected that shift. Networking got equal billing along with hybrid cloud and AI, the three of which comprise the building blocks for modern IT.

Networking

According to Neri, networking serves as the "core foundation for designing and deploying a modern, secure connectivity fabric across your enterprise." The exponential growth in users, devices, applications and data has strained traditional networks.

HPE addresses this change with networking for AI as well as AI for networking. Networking for AI upgrades infrastructure to handle AI's demanding data flows, while AI for networking embeds AI to manage, remediate and optimize the network itself.

While the two sides of the AI networking coin are not unique to HPE, the company announced the most network-focused innovation I've seen at Discover. Neri highlighted private 5G, Wi-Fi 7 access points and the agentic AI mesh in Aruba Networking Central with a networking co-pilot .

Now that the U.S. Department of Justice and HPE have come to an agreement about the Juniper Networks acquisition , HPE should be able to further its networking strategy moving forward.

Hybrid cloud

Neri positioned hybrid cloud as the go-to architecture for the AI era, which is consistent with what one might hear at other vendor events. AI creates unique requirements for privacy, security and sovereignty, and hybrid clouds can handle those better than public cloud only.

For HPE, GreenLake enables customers to securely run their workloads where it makes the most sense, including at the edge. At Discover, HPE announced GreenLake Intelligence , an agentic AI framework designed to simplify how businesses manage and run IT. This framework features intelligent agents powered by domain-specific LLMs that collaborate across the entire hybrid IT estate, providing continuous optimization, cost savings and full-stack control across compute, storage, networking and software.

AI

AI was the third pillar of HPE's modern IT strategy. During his keynote, Neri commented that "AI is only as good as the data infrastructure behind it," adding that HPE had revamped almost all its products for the AI era. For example, HPE announced its disaggregated intelligent system, HPE Alletra Storage MP, built for data-centric and AI workloads. It offers independent scaling of compute and storage nodes to deliver up to 40% lower costs and 45% reduced energy consumption.

Private Cloud is the Way Forward for AI, but Costs Need Containment

Technology trends ebb and flow. A handful of years ago, experts predicted that everything would move to public clouds. Fast forward to the AI era, where data sovereignty, privacy and control of data matter, and the industry is seeing a shift back to private clouds. A big part of Neri's keynote addressed the escalating and unpredictable costs associated with virtualization, licensing fees and rigid cloud models. Private cloud might be the way forward, but costs are starting to run amok.

HPE's response to rising costs is its HPE Private Cloud stack, powered by Morpheus software. The stack is designed to give businesses more control and cut costs through infrastructure efficiencies. Discover attendees applauded for Neri's claim that HPE Private Cloud can reduce VM licensing by up to 90% on a per-socket basis. The HPE Morpheus suite includes VM Essentials Software to manage virtualized workloads across VMware and HPE hypervisors as well as software for unifying operations across on-premises and public clouds.

HPE also announced ProLiant Gen12 servers , specifically built for private cloud and AI workloads. These servers have the equivalent performance of seven Gen10 servers. This reduces space and can cut power costs, both of which are important as companies look to scale AI without breaking the bank.

Ecosystems, Partnerships and Production-Ready Options Accelerate AI Deployments

While Neri's keynote largely focused on HPE technology, he did talk about the power of the ecosystem and how that can help get customers to production-ready AI faster. Specifically, he discussed the following areas.

The NVIDIA AI Computing by HPE initiative, first launched at the Discover 2024 event , continues to be a cornerstone of HPE's ecosystem strategy. HPE's new Private Cloud AI stack is designed to support the latest NVIDIA GPUs, including the H200 and RTX 6000 models. Key enhancements include multi-tenancy capabilities for secure tenant isolation and full air-gapped performance options for stringent data governance.

HPE's "Unleash AI" ecosystem is about bringing AI vision to life while minimizing risk. This initiative includes partnerships with several ISVs such as H2O.ai for agentic application creation, Dataiku for data science development and Frameworks Labs for industry-specific models.

Partnerships with systems integrators can help customers adopt AI and align them to business outcomes. Deloitte's Zora AI on HPE Private Cloud provides a ready-to-deploy, turnkey system for AI agents. Similarly, HPE's partnership with Accenture brings the AI Refinery platform to market on HPE Private Cloud.