HPE Discover 2024 live: All the news and updates as they happened
HPE Discover 2024 is a wrap – here's everything we learned in Las Vegas this year
Jane McCallion is ITPro's Managing Editor, specializing in data centers and enterprise IT infrastructure. Before becoming Managing Editor, she held the role of Deputy Editor and, prior to that, Features Editor, managing a pool of freelance and internal writers, while continuing to specialize in enterprise IT infrastructure, and business strategy.
What you need to know ahead of the opening keynote
With just a couple of hours to go before the opening keynote, delegates are already starting to file into the Venetian Conference Center, from where they will make their way to the Sphere for a 9.00am start.
HPE is very keen to highlight that this is the first keynote ever given from the venue, which was opened less than a year ago.
While audiences have been consistently marveling at the technology involved in the performances put on here since then, they have to date all been musical performances. So an enterprise IT conference is quite the change of speed.
While we don’t yet know what Neri will be announcing from the keynote stage, we have been able to make some predictions purely based on the fact Nvidia CEO Jensen Huang will be joining the keynote.
Huang has been a frequent guest on the tech conference circuit so far this year, having made appearances at Dell Technologies World in May, the recent Databricks Data + AI Summit, and an appearance via video link at the Snowflake Data Cloud Summit earlier this month.
It’s expected that AI will be a key talking point at HPE Discover 2024. The invitation of Jensen Huang for Neri’s keynote is a clear nod in that direction, given Nvidia’s prowess in AI and the fact that every other appearance during this year’s conference season has been about AI.
Of course, the big announcements from last year’s keynotes were the change in strategy (or slogan) from “the edge to cloud company” to “the edge to cloud to AI company” and the launch of HPE GreenLake for Large Language Models (LLMs).
GreenLake for LLMs leant heavily into the company’s high performance computing (HPC) heritage, but for now it still only has a handful of pilot users. Consequently, it’s unlikely that whatever today’s presumed AI announcement will be much related to this previous launch.
HPC and AI aren’t the only game in town for HPE, though, and it’s been a busy few months for the company in terms of product launches.
In May, the firm unveiled a host of new features for its flagship HPE GreenLake platform.
This included block storage for AWS as well as new private cloud offerings.
Elsewhere, new Aruba Wi-Fi 7 access points were unveiled by the firm in April as the company looks to boost enterprise IoT capabilities at the edge.
HPE’s reasoning for this move is clear; with the arrival - and rapid uptake - of IoT in recent years, these devices need to do more than simply route internet traffic.
The company said the new access points can provide up to 30% more capacity for wireless traffic than traditional methods.
You can read more about both these product updates below.
Although the show floor hasn’t opened yet this morning – this will likely come after Neri’s completed his keynote – delegates from the company’s Partner Growth Summit and other early arrivers were able to have a bit of a look around yesterday afternoon.
The AI zone has pride of place by the entrance to the main hall, once again doubtless hinting at announcements to come today. Tucked away among the stands is a replica of Venado, a supercomputer built by HPE and Nvidia for Los Alamos National Laboratory in Los Alamos, New Mexico, in April this year. The installation, which features Nvidia’s Grace Hopper superships, is designed to handle AI workloads to enable modelling and simulations across various areas of scientific research.
The trickle of attendees flocking into the Venetian is now very much becoming a flood as we make our way to the Sphere for the opening session. Big queues, but we're on the move and it's nearly time for Antonio Neri's keynote.
As ever, there are some gimmicks to keep delegates entertained as they file into the atrium, such as this robot in the atrium of the Sphere. 'She' (it has a female voice and form) seems to be some kind of wise-cracking generative AI. Attendees are invited to ask questions via a microphone.
"Can you sing?" asks one delegate. "Unfortunately I'm not strong enough to [heavy pause] carry a tune, haha," comes the response.
A giant HPE logo looms over the empty stage where HPE CEO Antonio Neri will appear in about 30 minutes. Meanwhile, the auditorium is steadily filling up with an animated crowd.
Lights, camera, action. We've started with a quote from Aristotle before getting the full 4D experience of the Sphere.
After a small amount of talk about dawn this morning we're onto some of the key themes for HPE: AI (naturally), sustainability, and using knowledge gleaned from experiments in space to improve life on earth (this latter part doubtless being a reference to the company's Spaceborn Computer project).
Antonio Neri is now on stage. The venue is so enormous, he's barely visible, but once again the company's making full use of the Sphere's technology to have two virtual high-resolution images of him on the walls.
Neri starts by talking up AI and the future of the technology, and how HPE's high performance computing (HPC) is foundational for the company's AI efforts. "I stand before you full of inspirations as we arrive at the AI solutions that will accelerate the generative AI Industrial Revolution through breakthrough organic innovation, targeted acquisistions, [and] key visionary partners," he says.
"From renewable energy to helping solve foods food scarcity to drug discovery HPE enables our customers to harness the power of AI as a driving force for good," he continues. "For instance, HPE proudly offered its patented innovations in AI systems to accelerate the world's response to the COVID pandemic. We provide a supercomputing software and application expertise as well as our systems to help research ports run and optimise is such an applications for the research. "
Neri also touches on the company's sustainability credentials. "As we look forward the next generation [of HPC] silicon will require more power density. AI systems generate heat in waste," he says. "HPE is at the forefront of applying the latest liquid technology in thermal cooling, and has more than two decades of experience."
He also points out that seven of the top 10 supercomputers on the Green 500 list, which he puts down to the company's "liquid cooling leadership".
Neri's onto the company's acquisition of Juniper Networks, which is expected to close either late this year or early next. "We expect to harness the combined capabilities in both [Aruba Networks and Juniper Networks] to create an industry leader [with our networking portfolio]," he says.
"This will complement our business strategy to [accelerate] deployment and adoption for both hybrid and what that means for enterprises, managing [and] simplifying complex network architectures and extending into new areas such as private 5G for environments where you need persistence," he adds.
Neri is now talking up the company's existing partnership with Nvidia, in particular their partnership on AI products such as Venado at Los Alamos. I feel we're about to get an announcement...
ANNOUNCEMENT: HPE is partnering with Nvidia to launch Nvidia AI Computing by HPE
"We are not just witnessing the future. We are shipping it with AI for more than a decade. Nvidia has been a visionary partner who shared our purpose and commitment to innovation," says Neri. "HPE and Nvidia have a proven track record for delivering innovation. Today we're taking a partnership further.
"It is with great excitement that we announced Nvidia AI computing by HPE to accelerate the generative AI industrial revolution."
Neri also welcomes Nvidia CEO Jensen Huang to the stage to talk more about this partnership.
Nvidia AI Computing by HPE isn’t one single product but a range of initiatives. The flagship of these, however, is HPE Private Cloud AI. You can read our full write up of the announcement here.
According to the CEOs, it takes just three clicks to set up HPE Private Cloud AI, with small, medium, large and extra large options.
"Remember," says Huang, "you're not buying a computer you can use, you're buying a service."
The architecture of all four options, Huang says, is completely interoperable. So if a customer wants to start small and grow, they can.
Neri says that while reticence over AI was understandable in the early days, organizations need to move forward.
Huang adds: "The era of generative AI is here you must engage in the single most consequential technology in history. You must translate ... your company's intelligence into digital intelligence and you must find a way as quickly as possible to turn that flywheel of your company's experience into your company's AI."
A final word for HPE GreenLake from Neri. "Today, HPE GreenLake connects more than four million ... devices and [provides] storage and compute systems support in more than 34,000 unique customer organisations to transform their business."
He's going over some of the big name GreenLake users, such as Harry Reid International Airport in Las Vegas, English Premiership football team Tottenham Hotspur FC, and the recently announced partnership with Spanish football team FC Barcelona at its Camp Nou stadium.
Neri thanks partners, customers and the "more than 3,000 team members around the world" and with that the day one keynote is finished. Make sure to come back tomorrow for live coverage of the day two keynote with CTO Fidelma Russo and click here for all the latest news from the conference.
It's a busy morning on the show floor as people start to file into the main theatre for CTO Fidelma Russo's second day keynote. There's the burble of conversation as we move towards the venue, including some people who have stopped to talk to an avatar of HPE's CEO "Antonio Nearly".
Russo has taken to the stage and is quickly joined by a representative of one of HPE's customers, Alexi Varanko, VP of cloud, data, SRE, and product engineering the Walt Disney Company.
Varanko explains how because the company's operations span various locations, managing it requires a "flexible, scalable pricing structure that can adapt to diverse markets... performance and reliability".
Previously, the company was doing this manually, which Varanko says made it hard to standardise. This led to the adoption of cloud technology, however the company moved to HPE GreenLake as so much of its computation is done at the edge. "GreenLake helps us manage across multiple different locations, deploy computing closer to the edge where the data is, as well as ensure security compliance."
Varanko leaves the stage and Russo moves on to HPE Private Cloud AI.
"HPE GreenLake is the hybrid cloud that enables enterprises to manage costs, while increasing flexibility and optimising operations [and] we'll continue to invest in the Green Lake hybrid cloud portfolio," she says.
"Yesterday we made two big announcements," she continues. "We announced our HPE private cloud AI codeveloped turnkey privatised in partnership with Nvidia. [It] is designed to accelerate accelerate enterprise AI adoption, with a click to deploy engineering system, preloaded with a complete AI software stack code developed by Nvidia and HPE."
The second announcement was a new virtualization capability for HPE Private Cloud.
"As a trusted partner, customers have reached out to us as they are reevaluating how they modernize," she says. "They're looking for new ways to support their legacy virtualized workloads and accelerate innovation through cloud age and AI workloads. Our customers want choice and HPE remains committed to offering choice as we continue to support all virtualization technologies in the industry."
This new virtualization capability is based on an open source KVM hypervisor, which has been 'hardened' for enterprise use.
"With the addition of our new HPE virtualization capability for private cloud, we now have a complete set of runtime capabilities across VMs containers and bare metal, which further supports our ambition to provide fully integrated hybrid cloud solutions," she says.
We're now supposed to be watching a demo of HPE Private Cloud being deployed from Denis Vilfort, director of business development, however it's quickly gone awry.
Unfortunately nothing is loading, although we do briefly get a look at the famed "three clicks to deployment" – to cheers from a very sympathetic audience – before the gremlins once again nibble at his presentation and now his microphone as well.
Russo finally returns to stage to a likely relieved Vilfort and the conversation turns to yesterday's biggest news, HPE Private Cloud AI.
Rather than writing out all of the features of HPE Private Cloud AI again here, we encourage you to read our write up from Antonio Neri's keynote yesterday.
She restates the four configurations of Private Cloud AI that will be available at launch: Small, medium, large, and extra large.
"Each one of these is modular. That allows you to expand and add capacity over time and you get to maintain a consistent cloud managed experience with HPE," she says. "We are committed, with our partner Nvidia, to introduce the latest models [so you can] can you start off with inferencing pilots when you scale to multiple use cases to either run inferencing RAG or LLM fine tuning in one solution."
Russo: "Now the moment you've all been waiting for and I'm sure Dennis is really thrilled to come back to stage and show you how all of this comes together."
Luckily for Vilfort, the self-described hardware guy, this is going much more smoothly than last time he was on stage. He has opened the door on a piece of the new HPE Private Cloud AI hardware that has been lurking meaningfully on stage and we're first complementing it on its looks.
"Look at this nice hardware," he says, "we're responsible for that! And here's what I keep saying to my software brothers, remember your software – as cool as it is – i runs out of something somewhere."
Now we have a rundown of its components.
"We've got the foundation down here, these are servers – what a surprise, we have servers in the mix!" he jokes. These are, he says, a new kind of servers, "AI servers".
What makes it an AI server, exactly? "That's here Nvidia GPUs come in."
While customers may know Nvidia GPUs from gaming PCs, these are not quite the same thing. "These are 700 watts apiece. They have 80GB-worth of VMI, and thousands and thousands of CUDA cores."
Vilfort also points out the high speed networking components, storage "the heart of the whole experience", which is located poetically in the middle, and a relay control plane. "We actually have to have a cluster, highly available, managing everything in the machine," Vilfort says. "And if you go to just two or three racks, this little brain manages the entire system.
Vilfort is moving back to the software demo now to show that this "fully integrated system" can be deployed in just a handful of clicks.
"Tour de force, next 10 minutes, you ready?" he asks the audience "I hope so!"
And indeed he has succeeded in his task – from the first click to the last takes a little under 10 minutes. The audience seems impressed or at least supportive of his success.
Vilfort now runs us through the importance of guardrails in a generative AI chatbot.
"Here we have the Nvidia guardrail and we have compiled it into the code and no one can turn it off," he says. "So now we have a safer chatbot – let's see how safe it is."
Vilfort asks the chatbot to hand over HPE CEO Antonio Neri's Social Security number. It responds with a simple "I cannot assist you with that" message.
"So now it's fully safe! Pretty cool, right?" he says.
Fidelma Russo is now back on stage and is joined by Manuvir Das, VP of enterprise computing at Nvidia.
"We've been through a lot of shifts in the industry, but how is this shift compared to other ones?" she asks.
"This shift is actually fundamentally different," he replies. "The reason is that if you think about other shifts – virtualization, databases, even the cloud, which was a massive shift –if you think about the average company and the average employee and the average worker level [this has a greater direct impact]."
"I think there's a new kind of enterprise app coming and I hope in two or three years from now that's what you're going to see in all companies ... where it's no longer 'I've got these databases and I'm running little applications on top of databases', [but] 'I've got the data lake where I've got all my enterprise data. I've got this infrastructure farm to host models, and I've got 1000s of apps that are assistants of different kinds that I used in production every day'," says Das.
Das continues: "If you look at the journey of AI so far, it's been a little oriented towards the bigger companies, the companies that have the wherewithal to stand up huge amounts of infrastructure and all of that.
"But who needs gen AI? It's actually the smaller companies that have fewer resources that will benefit more from gen AI because if I have only 20 people in my company, and I can be twice as productive, I have 40 people in my company. So what I hope to see two three years from now is every company out there has doubled in size because of all these assistants."
With that Das leaves the stage and it's time for Russo to wrap things up.
"By partnering with HPE and adopting HPE Greenlake, you can fully harness the power of AI," Russo says. "Let's applaud the future possibilities and success we will achieve together. Thank you for your time. Thank you for your patience and enjoy the rest of the conference."
That's it for this live blog, but if you want more content from HPE Discover, why not check out one of these articles from the show:
