HPE’s Discover 2024 event keynote took place at the Sphere in Las Vegas, which seemed like the perfect setting to say we’ve seen the future. As the first corporate speech to be held within the domed wonder, HPE had the unique opportunity and challenge to rethink how big event keynotes can be done. I’m pleased to report that they did a great job of demonstrating how companies can leverage the unique venue to their advantage.
Accompanied by some of the awe-inspiring clips from the Sphere’s “Postcard from Earth” experience, impressive original graphics, rumble-seat enhanced customer videos, and massive images of their own, HPE CEO Antonio Neri managed to deliver a compelling and informative presentation that few attendees are likely to forget.
Part of the impact, of course, is because attending most anything in the Sphere is pretty compelling. It’s also an absolute textbook example of Marshall McLuhan’s infamous comment that “the medium is the message,” and there was certainly concern that the message could have been lost in the new medium that the Sphere represents. Nevertheless, Neri’s keynote managed to kick off HPE’s customer and partner event in a big way and in a manner that allowed the company to announce its news and provide some interesting context around it.
I have no doubt that we’re going to be seeing a lot more Sphere keynotes over the next few years. Like it or not, it is the keynote of the future.
The other aspect of the future that HPE discussed during the keynote is bringing generative AI capabilities into customers’ environments. While most companies are still doing the majority of their AI workloads in the cloud – if they’ve started at all – there’s growing interest in doing more of these efforts internally where many organizations’ most important data still remains.
HPE is moving in this direction with offerings designed to tap into this growing trend. Neri had Nvidia CEO Jensen Huang join him onstage for an extended conversation to help launch the company’s new Nvidia AI Computing by HPE initiative and the HPE Private Cloud AI offering.
As with many other large enterprise hardware companies (Dell, Lenovo, Cisco, etc.), HPE’s latest partnership with Nvidia involves putting together GenAI-focused solutions based around not just Nvidia GPUs, but also Nvidia NeMo, Enterprise AI and NIM (Nvidia Inference Microservices) platform software.
This series of announcements show how far Nvidia’s influence has grown among the enterprise hardware industry, particularly with regards to its software offerings. In fact, its recent ascension to the most valuable company in the world is arguably due to the growing recognition of its software portfolio and the opportunity that future software revenues represent.
On the other hand, it does seem to present a dilemma for its partners as they’re all getting the same set of tools from Nvidia. Given how important – and differentiating – these software tools are, it could make it more difficult for these various companies to distinguish their offerings from those of their competitors.
In HPE’s case, the company emphasized that it is bringing a significant amount of its own IP to the Private Cloud AI offering, both through its GreenLake tools and the custom work it has done to simplify the deployment of GenAI-powered applications. For GreenLake, HPE’s cloud-native offering that spans compute, storage, software, networking, and services, Private Cloud AI is another service within the GreenLake platform. This means existing HPE GreenLake customers can easily integrate it into their current environments.
For companies that still purchase HPE equipment and services traditionally, this could be a good way to experiment with GreenLake, especially given the rapid evolution of GenAI-focused hardware. HPE has also extended the OpsRamp observability tools within GreenLake to work across the new AI offerings, enabling companies to track performance across the entire AI stack, including GPUs, Nvidia InfiniBand and Ethernet switches, and Nvidia software components.
Many of HPE’s other GreenLake offerings focus on specific applications, such as disaster recovery or building a private cloud, and that same outcome-focused approach has been applied here. One key competitive difference for the new Private Cloud AI offering is that HPE has packaged all the hardware and software components necessary to create GenAI applications and reduced the deployment process to three clicks.
Given how complicated the process can otherwise be, this is an impressive step. It’s also a way to distinguish what HPE is doing compared to, say, Dell Technologies, which offers a broad array of different components so companies can create their own “AI Factories.” In real-world terms, these differences may be more subtle than they first appear as the solutions evolve, but at least there’s different positioning from each company regarding their offerings.
HPE also recognizes that even with this simplified setup, ongoing development of these workloads often requires consulting help, so it partnered with system integrators such as Deloitte, HCLTech, Infosys, TCS, and WIPRO to help customers create and maintain the applications they need.
To power these solutions, HPE unveiled four different Proliant computing hardware configurations, all powered by the latest Nvidia GPUs, designed to let companies achieve the performance they need. For example, if they want to do an initial inferencing deployment for something like a GenAI-powered custom chatbot, they can use a basic Proliant system. For customized fine-tuning of the base model and/or leveraging RAG (Retrieval Augmented Generation), they can use one of the two most powerful Proliant options with more GPUs and faster networking.
HPE also promised to bring configurations with the latest Blackwell GPUs to market as soon as they are available, a point the other vendors also made (likely a key reason for these vendors arranging such announcements with Nvidia).
For organizations looking to the future, it’s clear from these and other announcements that HPE is focused on making GenAI a key part of its future plans. The keynote also demonstrated that the company is working to build a stronger presence in GenAI after arguably playing catch-up with its competitors. Now that the playing field is more level, it will be interesting to watch how HPE continues to evolve its differentiation strategies and how its customers react to its latest offerings.
Bob O’Donnell is the president and chief analyst of TECHnalysis Research, LLC, a market research firm that provides strategic consulting and market research services to the technology industry and professional financial community. You can follow Bob on Twitter @bobodtech