CEO Antonio Neri On AI Systems Growth, GreenLake Momentum And HPE’s ‘100 Percent’ Liquid Cooling AI Advantage

HPE CEO Antonio Neri has promised an “epic moment” at HPE’s Discover Conference with the first-ever keynote at the Sphere in Las Vegas, where he'll be joined by Nvidia CEO Jensen Huang.

Hewlett Packard Enterprise CEO Antonio Neri says HPE’s “100 percent direct liquid cooling” prowess will be a key AI systems differentiator versus competitors with the release of Nvidia’s Blackwell GPU architecture.

“When you look at the [Nvidia GPU] road map with Nvidia leading the pack and you go to [Nvidia] Blackwell [GPU architecture], Blackwell requires 100 percent liquid cooling,” said Neri in an interview with CRN. “HPE has more than 300 patents in that space. Also, HPE has one of the largest manufacturing footprints when it comes down to water-cooled [systems] including in t he U.S. and Europe. Our leadership continues to be validated in the Top 500 list [of most powerful supercomputers] and the Green 500 list [of most energy-efficient supercomputers].”

The Blackwell GPU was unveiled in March at Nvidia’s GTC AI conference, with systems expected later this year.

HPE has a big advantage in 100 percent direct liquid cooling versus competitors that are relying on hybrid liquid cooling, which still requires fans in AI servers, said Neri.

“When you look at some of our competitors, they announced products that are direct liquid cooling,” he said. “The reality is they are what I call ‘hybrid,’ meaning 70 percent is direct liquid cooling and 30 percent is air-cooled. They still use fans to cool aspects of the system.”

Neri has promised an “epic moment” at HPE’s Discover Conference with the first-ever keynote at the Sphere in Las Vegas on June 18. Neri will be joined by Nvidia CEO Jensen Huang on the Sphere stage to reveal what HPE is calling “breakthrough AI solutions for enterprises of all sizes.”

HPE has a long history of working with Nvidia, including collaborating on the new Nvidia Grace Hopper-based supercomputer.

HPE has four of the top 10 fastest supercomputers on the Top 500 list and seven of the top 10 Greenest supercomputers on the Top 500 list.

Under Neri’s leadership, HPE has staked out a leading position in the supercomputer market with its acquisition five years ago of Cray. HPE also gained key high-performance computing intellectual property when it acquired Silicon Graphics in 2016.

Neri’s comments on liquid cooling came in an interview ahead of HPE’s fiscal second-quarter results, which included a doubling of AI systems revenue to over $900 million. That strong AI systems growth sent HPE shares up 12 percent in midday trading Wednesday to $19.67.

Key to the strong AI systems revenue performance was the improved availability of Nvidia’s H100 GPUs, which were in short supply and now are available in six- to 12-week lead times depending on order size and complexity, said Neri.

“We expect this will provide a lift to our revenues in the second half of the year,” said Neri. “Enterprise customer interest in AI is rapidly growing, and our sellers are seeing a rapid level of engagement.”

Overall, HPE reported sales and earnings above expectations for its second fiscal quarter ended April 30.

Sales for the quarter were $7.2 billion, up 3 percent from $6.75 billion in the year-ago quarter. The Zacks consensus estimate was $6.83 billion.

Non GAAP diluted net earnings per share for the quarter was 42 cents per share, down 19 percent from the year-ago quarter, but above the Zacks consensus estimate of 38 cents per share.

Here is an edited transcript of the discussion with Neri, including his view on liquid cooling, AI systems growth and HPE GreenLake momentum.

How do you feel about the second fiscal quarter results, particularly the increase in AI systems revenue?

I’m really pleased with the quarter performance in Q2. It was a solid quarter. We beat both revenue and non-GAAP earnings per share. That was all on the back of the AI systems revenue conversion, which more than doubled quarter over quarter to above $900 million. That’s No. 1.

No. 2 is that the revenue that we converted has solid profitability because there were two key drivers. One is business discipline that [HPE Executive Vice President and CFO] Marie [Myers] and I bring to this segment.

Second is over time we’re going to see more and more services revenue and profit attached to it. So the way we approach this business is, of course, growth, a profitable growth through business discipline, path to profitability and services attached. And that’s why when you look at our server performance, you see servers up 18 percent year over year in revenue, and our operating profit was 11 percent, which is in the range we guided the street with 11 [percent] to 13 percent.

So great growth, great profit, and then our cumulative AI systems orders reached $4.6 billion. But what is interesting is that more than 15 percent of that $4.6 billion is related to the enterprise and the number of enterprise customers tripled year over year. And that is the key because there are three customer segments: the model builders with hyperscalers and large service providers; there is [No. 2] the sovereign clouds; and there is [No. 3 the] enterprise.

How do you see the AI demand at this point in those three primary AI customer segments?

Eighty-plus percent of the demand today is in the model builders and service providers [segment]. But more than 15 percent of the demand is in the enterprise. Sovereign clouds are now starting to pick up because nobody wants to be left behind. There are two things at play there: generic AI clouds and then supercomputers for country and nation. HPE can serve both aspects really well.

So AI demand continues to accelerate. What I’m super excited about is that HPE can serve all these three segments and all aspects of the [AI] life-cycle [solutions] and all the new requirements that will come with this accelerated silicon because we have one of the most complete portfolios when it comes to direct liquid cooling.’

How does HPE with its liquid cooling capabilities compare versus competitors?

When you look at some of our competitors, they announced products that are direct liquid cooling. The reality is they are what I call ‘hybrid,’ meaning 70 percent is direct liquid cooling and 30 percent is air-cooled. They still use fans to cool aspects of the system.

In that [hybrid cooling] segment of the market, we don’t have only two [systems], we have 10. And we have our own IP with what I call ‘adaptive rack configurations.’

But then when you go 100 percent direct liquid cooling, we already have six systems that have been deployed and three of them are specific for generative AI.

So when you look at the [Nvidia GPU] road map with Nvidia leading the pack and you go to [Nvidia] Blackwell [GPU architecture], Blackwell requires 100 percent liquid cooling. HPE has more than 300 patents in that space. Also, HPE has one of the largest manufacturing footprints when it comes down to water- cooled [systems] including in the U.S. and Europe. Our leadership continues to be validated in the Top 500 list [of most powerful supercomputers] and the Green 500 list [of most energy-efficient supercomputers].

How important is hybrid cloud in the battle for AI supremacy?

In order to do AI, you need data. And hybrid cloud is necessary to AI because that’s where the data is today. Is it on-prem? Is it at the edge? Is it in the public cloud? You need a hybrid cloud capability, and HPE GreenLake is an essential tool to deliver AI because that is how we manage data on behalf of customers.

And when you look at our results, you can see we added 3,000 new customers to the HPE GreenLake platform. Now we have 34,000 customers. We exceeded $15 billion of total contract value. And our measure of success is both customers, [net-new] logos, and then ARR, our annualized revenue run rate, and that ARR grew 39 percent year over year.

How did the networking segment perform during the quarter?

In networking the market is still going through an adjustment period. We delivered a very strong performance with more than 21 percent operating profit despite the fact that revenue obviously declined because of the customer digestion. But we are now seeing signs of recovery in the traditional infrastructure business, which is good for our partners. And in doing so we expect modest demand improvement in networking in Q3 driven by the state and local [education] purchasing season in the U.S. Also, through our great work with our channel partners we have reduced our inventory and they’re super healthy. And of course I continue to be very excited about the pending acquisition of Juniper [Networks], which is on target to close end of this calendar year or beginning of the calendar year 2025.

Talk about recent innovation from HPE.

There is a lot of innovation, including yesterday with [the announcement of] Private 5G, the extension of our block storage with Alletra into AWS, the file storage for AI, Wi-Fi 7.

At HPE Discover there will be an epic moment because you are going to have enormous amounts of new announcements.

How did the channel perform during the quarter?

In Q2 our performance in the channel with products and services grew 6 percent year over year and 8 percent quarter over quarter.

This is the 10th straight quarter of growth in the number of active Green Lake partners, which represents 24 percent growth year over year.

Our indirect revenue in Q2 is at 60 percent, which is an increase of 6 percent year over year and 8 percent quarter over quarter. So our revenue through the channel has grown 6 percent year over year and 8 percent quarter over quarter.

Now, when you look at the server business, which obviously has a component of AI, that represented 54 percent and the growth was 17 percent year over year and actually 17 percent quarter over quarter.

And then when you look at our hybrid cloud [business], which obviously has storage as a component, the growth was 10 percent year over year. And then when you look at storage, specifically Alletra MP, which is the flagship product we are all driving to, sales grew 65 percent year over year.

So clearly our innovation is resonating with the channel. The volume of business through the channel grew year over year and sequentially. The number of active partners grew again 24 percent for HPE GreenLake. And we need to continue to engage customers in this AI journey, which will require a lot of services. This is where the channel has a big opportunity to provide those services, helping customers figure out what the right AI strategy is, and then providing help running these [AI] systems because when it comes down to running them that’s an issue.

I’m very pleased with the quarter, very pleased with the channel performance. There is a lot of opportunity.

My job is to bring the innovation to our partners and our customers. We continue to do that. You will see that extensively at HPE Discover in less than two weeks.

How do you think the channel is doing selling AI servers, and what do they have to do to be successful there?

It’s hard to break out all the numbers because you have to look at the servers by platform. The servers in total grew 18 percent. It’s fair to assume that the vast majority that growth came through AI adoption, not through the traditional CPU space. That’s why I’m very comfortable to say that of the 18 percent growth in servers the vast majority was AI-driven.