10 years ago HP decided to split into two parts. In this way, two giants were born that have continued to maintain leadership in their respective businesses. In this decade, HPE, the company focused on IT infrastructures for businesses, has expanded its portfolio to meet virtually all business needs. And this is what could be seen in a new edition of HPE Discover held last week in Barcelona.
Under this atmosphere of celebration, Antonio Neri, CEO of the multinational, offered in the opening speech of HPE Discover the ambitious roadmap that will define the firm’s philosophy in the coming months. The objective is for HPE to be the technological partner that IT departments are looking for in a context that will be marked by generative and agentic artificial intelligence and the hybrid cloud. But since HPE’s catalog is so broad, the company’s top executive emphasized the value that native, secure and sovereign AI networks will acquire, in a wink designed for the European audience present in the auditorium.
With more than 6,000 enrolled in HPE Discover, Neri covered all points. His speech combined vision of the future, announcements of new products and messages aimed at CIOs, CISOs and business managers, as well as some hint to a competitor: “The future belongs to those who are fast and the opportunity lies in those who are capable of modernizing the network, cloud and data at the pace that AI demands.”
Networks, cloud and Artificial Intelligence
As happens in any technological event, Artificial Intelligence, in any of its flavors, occupies a good part of the messages. And in this HPE Discover it couldn’t be missing either.
As Neri stated, the key is that “AI is no longer an isolated experiment, but a digital workforce made up of thousands of specialized agents that IT areas must govern in the same way that templates, applications and data centers are managed. And in this, companies have to emphasize that AI agents will only work correctly if the architecture that supports them does the same. That is why networks, cloud and data must evolve at the same speed as models and GPUs.”
HPE’s strategic framework that could be seen in HPE Discover is based on three pillars that define the technology of companies today. Those three pillars are networks, the cloud and artificial intelligence. And among those three, the fundamental basis is networks because “they are what connect people, infrastructure and data, even when organizations operate in multiple countries and with distributed data centers.” The Italian-Argentine is right, although the message has been amplified since the company acquired its rival Juniper Networks last July.
At another time, cloud and AI would have been the stars of HPE Discover, but after the acquisition of one of its main rivals, networks have taken most of the spotlight at the event.
That does not mean that the other two legs of the strategy were overshadowed. In fact, Neri assured that “the cloud (particularly the hybrid cloud) provides the flexibility to run workloads where it makes the most sense, with a coherent operating model in own data centers, public clouds and disconnected environments.”
For its part, and as Neri explained at HPE Discover, AI would be the third essential component and becomes the layer that extracts value from data to accelerate results: improve the customer experience, optimize operations, create new sources of income or drive advanced research. Neri stressed that “these three vectors cannot be addressed in isolation. Complexity, sovereignty and security requirements, and the increasing cost of the cloud and virtualization force us to think about a unified architecture where each decision in one area impacts the rest.”
Native AI networks, stars of HPE Discover
But the main protagonists were the networks. Neri has presented the integration of HPE Aruba Networking and Juniper as the foundation to “create the best networking company on the planet”, with a complete portfolio of native and AI-driven cloud solutions that combine connectivity and security in a single stack.
HPE’s goal shown in the HPE Discover is to move towards autonomous networks that are self-configuring, self-optimizing and self-healing and that are capable of supporting both “AI for networks” and “networks for AI”, that is, that can use artificial intelligence to operate the infrastructure while ensuring the latency, reliability and scale required by clusters of tens of thousands of GPUs. The CEO of HPE explained that “we are at a unique moment in which companies can completely rethink their network infrastructure and their solutions strategy, with real use cases that span healthcare, universities, retail, hospitality and large telecom operators already using HPE’s combined platforms to deliver secure connectivity and advanced digital experiences.
In this context, HPE has integrated the best of Aruba Networking Central and Juniper Networking Mist on a common artificial intelligence and microservices framework to unify operations and protect customer investments. Mist Large Experience Model (LEM), which analyzes billions of data points from applications like Zoom and Teams combined with digital twins to detect, resolve and predict video incidents, becomes available in Aruba Central, elevating the quality of the digital experience. In turn, Aruba’s Agentic Mesh technology is incorporated into Mist, reinforcing anomaly detection and root cause analysis through advanced reasoning and autonomous or assisted actions. Mist will adopt Aruba Central’s global NOC vision to deliver a consistent user experience across both platforms, complemented by new WiFi 7 access points and Aruba Networking Central On-Premises 3.0, which brings advanced AIOps, automation and analytics to secure on-premises environments.
GreenLake, private cloud and the alternative to traditional virtualization
One of HPE’s great successes in these ten years was the development of GreenLake. In this edition of HPE Discover, Neri also placed special emphasis on this platform and the company’s new generation of private cloud. Neri, in a clear jab at the new VMware, emphasized one of the main concerns of CIOs, which is the increase in virtualization costs, the rigidity of certain architectures and the imminent deadlines linked to changes in licensing models, which make it unfeasible to continue relying on a single provider.
Networks, Artificial Intelligence and cloud were the protagonists of HPE Discover
In response, HPE has introduced its next-generation private cloud, powered by HPE Morpheus software, which promises to regain control over workload placement and cloud spending, significantly reducing virtualization costs. This solution delivers a consistent cloud experience for virtual machines, containers, and AI workloads, with support for both VMware and HPE’s own hypervisor, and live migration, self-service, and multicloud orchestration capabilities.
Morpheus is the key: it works as a unified management layer to coordinate workloads in public, private clouds and on-premise environments, including virtual machines, bare metal and containers. The proposal is infrastructure agnostic and relies on policies to automate provisioning, governance and cost optimization, seeking to avoid new supplier lock-ins in the classic post-virtualization era.
When it comes to AI agents and hybrid cloud, HPE is advancing its hybrid cloud and Agentic AIOps strategy with advances that highlight one of the company’s great strengths: full-stack, multi-domain and multi-vendor intelligence, based on a shared resource model that spans from hardware to the public cloud. With these enhancements to HPE OpsRamp software and deeper integration with GreenLake, the company now brings together telemetry from HPE Compute Ops Management, HPE Aruba Networking Central, and HPE Juniper Networking Apstra to give IT teams a single place to see, interpret, and act on everything happening in their environment, forming the foundation of a true hybrid command center.
Storage becomes an accelerator
At HPE Discover, this premise that all technology companies insist on was once again emphasized. If AI is to work correctly and effectively, the data it is fed with must have value for the business and for the applications that are to be launched. In this sense, Neri has insisted that “success with Artificial Intelligence requires access to distributed data quickly and contextually, which forces us to overcome the fragmented storage models of the past.” On this premise, HPE Alletra Storage MP has presented as a unified data platform, designed for data- and AI-centric workloads, with independent scaling for performance and capacity, improvements of up to 40% in performance and reductions in energy consumption of up to 45%.
The big news seen at HPE Discover has been the introduction of HPE Data Intelligence Nodes, GPU-accelerated systems that act as an intermediate layer between storage and compute to preprocess, extract features and enrich metadata in real time, so that data arrives “AI-ready” to training and inference pipelines. Integrated with NVIDIA, these nodes form the foundation of the new HPE-NVIDIA AI data platform, which unifies compute, storage and data fabric in a single architecture, optimized to power GPUs at maximum speed and reduce latencies and costs by avoiding intermediate CPU steps.
Data sovereignty
Finally, Neri was smart to include in HPE Discover a section focused on data sovereignty, an aspect that is increasingly worrying European companies. The CEO has cited data that indicates that more than half of European organizations already use sovereign architectures or plan to do so in the next year, and has highlighted that orders for HPE sovereign AI solutions grew 250% quarter on quarter.
Neri has presented his company as concerned about data sovereignty. And it did so with examples such as HPE’s collaboration with the European Space Agency (ESA) to deploy a supercomputing system in the Observation Center that the organization has near Rome.
