Earnings Call Transcript - HPE Q2 2024

Operator, Operator

Good afternoon and welcome to the Second Quarter Fiscal 2024 Hewlett Packard Enterprise Earnings Conference Call. My name is Gary, and I'll be your conference moderator for today's call. At this time, all participants will be in listen-only mode. We will be facilitating a question-and-answer session towards the end of the conference. As a reminder, this conference is being recorded for replay purposes. I would now like to turn the presentation over to your host for today's call, Jeff Kvaal, Head of Investor Relations. Please proceed.

Jeff Kvaal, Head of Investor Relations

Good afternoon. Welcome to our Second Quarter Fiscal 2024 Earnings Conference Call with Antonio Neri, HPE's President and CEO, and Marie Myers, HPE's CFO. Let me remind you that this call is being webcast. A replay of the webcast will be available shortly after the call concludes. We have posted the press release and the slide presentation accompanying the release on our investor relations webpage. Elements of the financial information referenced on this call are forward-looking and are based on our best view of the world and our businesses as we see them today. HPE assumes no obligation and does not intend to update any such forward-looking statements. We also note that the financial information discussed on this call reflects estimates based on the information available at this time and could differ materially from the amounts ultimately reported in HPE's Quarterly Report on Form 10-Q for the fiscal quarter ended April 30, 2024. For more detailed information, please see the disclaimers on the earnings materials relating to forward-looking statements that involve risks, uncertainties, and assumptions. Please refer to HPE's SEC filings for a discussion of these risks. For financial information, we have expressed on a non-GAAP basis; we have provided reconciliations to the comparable GAAP information on our website. Please refer to the tables and slide presentations accompanying today's earnings release on our website for details. Throughout the call, all revenue growth rates are presented on a year-over-year basis and adjusted to exclude the impact of currency unless otherwise noted. Finally, Antonio and Marie will reference our earnings presentation in their prepared remarks. With that, Antonio.

Antonio Neri, President and CEO

Good afternoon, and thank you for joining us today. HPE delivered a very solid performance in the second quarter, with revenue and non-GAAP diluted net earnings per share exceeding our outlook range, driven by AI systems revenue more than doubling from our first quarter. I am very optimistic about where we're headed. AI demand continues to accelerate with cumulative AI systems orders reaching $4.6 billion this quarter. We have a robust pipeline in this business, though large AI orders can cause fluctuations during the quarter. We anticipate continued revenue growth driven by increased AI systems demand, continued adoption of HPE GreenLake, and ongoing improvement in the traditional infrastructure market, including servers, storage, and networking. Due to our confidence in the second half of fiscal year 2024, we are raising our full-year revenue and non-GAAP earnings per share guidance and reiterating free cash flow guidance. Marie will provide more specifics in her remarks. While we focus on translating strong AI customer demand to revenue growth, we continue to drive cost discipline to operate more efficiently and to preserve the ability to make targeted investments that will sustain our growth into the future. We are being prudent with our spending and reduce operating expenses in the first half compared to the prior year period. We're also driving business process simplification across the company, including through digitization and automation with AI. Demand for HPE's AI systems is accelerating at a faster pace, and our solid execution enabled us to more than double our AI systems' revenue sequentially to over $900 million, helped by supply chain conversion through improved GPU availability. Our lead time to deliver NVIDIA H100 solutions is now between six and 12 weeks, depending on order size and complexity. We expect this will provide a lift to our revenues in the second half of the year. Enterprise customer interest in AI is rapidly growing, and our sellers are seeing a higher level of engagement. Enterprise orders now comprise more than 15% of our cumulative AI systems orders, with the number of enterprise AI customers nearly tripling year-over-year. As these engagements continue to progress from exploration and discovery phase, we anticipate additional acceleration in enterprise AI systems orders through the end of the fiscal year. HPE has decades of experience in the design, manufacture, and management of air and liquid cooling systems, including the data center infrastructure to reliably deliver the highest levels of computing performance. Customers appreciate our AI at-scale expertise and intellectual property, as well as unique liquid cooling manufacturing and services capabilities. As accelerated computing and silicon innovation advance, higher power density demands direct liquid cooling technologies. Building direct liquid cooling AI systems is complex and requires manufacturing expertise and infrastructure, including power, cooling, and water. With more than 300 HPE patents in direct liquid cooling, proven expertise, and significant manufacturing capacity for these types of systems, HPE is well-positioned to help customers meet the power demands for current and future accelerated compute silicon designs. Our leadership in AI was once again validated in May when the latest top 500 list of the world's most powerful supercomputers was released. HPE now has four of the world's ten fastest supercomputers, all of which are direct liquid cooled. Two of these systems are exascale supercomputers, with Frontier still the world's fastest and Aruba now breaking the exascale barrier. I'm also proud that we have built seven of the world's top ten energy-efficient systems according to the latest Green 500 list. This experience makes us an attractive partner to sovereign states and governments pursuing AI strategies. HPE benefits from a strong ecosystem of AI partners, including NVIDIA. We introduced co-engineered enterprise solutions with NVIDIA last year to streamline the model development process, as well as to enable enterprises to fine-tune large language models with their private data to accelerate inferencing. I'm excited that NVIDIA CEO, Jensen Huang will join me at HPE Discover Las Vegas in just two weeks. Together, we will unveil new exciting and differentiated innovations that will simplify and accelerate enterprise AI adoption and deployment. Enterprise customers are already responding to our unique AI portfolio. For example, QBox, a facial and image recognition company in Korea, is developing new generative AI models using HPE AI systems to enhance identity verification at locations such as Incheon International Airport in South Korea. JT Group, based in Japan, operates a pharmaceutical business and is planning to use our HPE AI systems to support AI model training and simulations to accelerate drug discovery. And we just announced that we will power Scaleway’s AI cloud service offering using our HPE AI systems. The new service will make powerful computing accessible to companies to support their various AI workloads and use cases. As we capitalize on the AI growth opportunity, we also see indications of market recovery in traditional and cloud infrastructure markets. Orders for traditional services grew sequentially and year-over-year, driven by enterprise, public sector, and SMB customers in North America and Europe. We are seeing no indication of cannibalization from accelerated computing demand. Revenue grew sequentially as customers transitioned to higher AUP HPE ProLiant Gen11 servers. Differentiated customer-centric innovation positions us well to capture this market recovery. For example, our leading HPE GreenLake hybrid cloud is attracting new customers. In the second quarter of fiscal 2024, the number of customer organizations using HPE GreenLake increased sequentially by almost 9% to 34,000. Our as-a-service lifetime total contract value grew to more than $15 billion in Q2, with our annualized revenue run rate or ARR growing 39% year-over-year. Demand is increasing for our HPE GreenLake on-premise private cloud solutions. The Defense Information Systems Agency, a combat support agency of the United States Department of Defense, selected HPE to develop a distributed hybrid multi-cloud platform prototype on HPE GreenLake, as part of an effort to simplify the organization's management of disparate IT infrastructure and resources across public and private clouds. In storage, we accelerated the transition of our portfolio in Q2 to meet the needs of hybrid cloud and AI. Several weeks ago, HPE introduced significant new functionality to our HPE Alletra storage offerings. We rounded out our block offering by extending the hybrid capabilities of HPE Alletra block storage to AWS, doubling its capacity to address more customers and enhancing its automation capabilities with generative AI. Early this quarter, we introduced new HPE GreenLake for file storage capabilities with options specifically targeting the unstructured data demands of AI. We have also added significant specialist sales capacity in recent months. While it takes time to activate new sellers and bring them to full productivity in a market with long sell cycles, we expect increased order-to-revenue conversion in the future. The enhancement to our winning portfolio complemented by a more focused sales force positions HPE to strengthen the already robust customer adoption of HPE Alletra. More than 1,000 new HPE Alletra MP systems have been deployed to date, which is the fastest product ramp in the history of our company. In networking, the market remained in transition during the quarter, as customers continued to work through their current inventory. As demand in this segment gradually returns, we believe our broad portfolio positions us well. We expect modest sequential networking demand improvement driven by the state and local purchasing season in the United States. We announced significant new innovations during the quarter to align with HPE's broader AI strategy. These solutions include generated AI capabilities to improve AI Ops and Wi-Fi 7 access points that capture edge data for AI inferencing. In addition, last month we launched new security and AI observability tools to help fight AI cyber risks. Just yesterday, we expanded the most complete private 5G and Wi-Fi portfolio in the market with the launch of HPE Aruba Networking Enterprise Private 5G. All of this will be delivered through our HPE GreenLake cloud platform. Customers are responding to our networking innovation. In the last few months, customers ranging from Houston Airport to Baptist Health Group to Mercedes-Benz Stadium to the University of Maryland have turned to HPE to enhance the experience for visitors, residents, and employees. As we look to our future in networking, we continue to be very enthusiastic about the proposed acquisition of Juniper Networks. We're currently in the regulatory process for this transaction and expect to close by the end of 2024 or early 2025. As I mentioned earlier, we continue to invest in innovation while we drive operational and cost discipline to continue to improve our cost structure. We are focused on reducing complexity in our business processes, as well as implementing automation and AI across the company to enhance customer service, R&D productivity, and overall team member experience. I also want to note that we recently announced we have restructured the sale of our stake in H3C. Since the transaction is large and complex, the required regulatory approvals were taking longer than previously anticipated. So we agreed to restructure the sale with units. The updated agreement provides HPE with the opportunity to sell a significant portion of the shares in the coming months. The new payment structure has no impact on the pending Juniper Networks transaction, as the structure of the deal financing does not rely on any of the H3C proceeds. Finally, you likely saw that we announced in May that we have agreed to divest our communications technology group to enhance our strategic focus in high growth, high margin parts of the market, including the service provider and enterprise markets. In closing, I want to reiterate that I'm proud of the very solid performance in Q2. It shows the alignment of our strategy and innovation to major market opportunities. We have greater optimism about the second half of the year, leading us to raise our full-year revenue and non-GAAP earnings per share guidance. AI is creating growing demand across our portfolio, and we see significant opportunities across customer and business segments. Our competitive advantages, from deep expertise in standing up AI systems to our differentiated HPE GreenLake Cloud to our networking storage offerings position us well. We have an excellent team, and I'm confident in our ability to continue executing with discipline to take advantage of the incredible opportunities presented by this era of innovation. I am looking forward to sharing with you our latest breakthrough innovation and partnerships across AI, hybrid cloud, and networking later this month at HPE Discover Las Vegas. We are very excited to be the first corporate keynote at Sphere, and I hope to see many of you there. I would like now to hand it over to Marie, who will talk about the details of our segments and our outlook. Marie, over to you.

Marie Myers, CFO

Thank you, Antonio, and good afternoon, everyone. It's a pleasure to be here with all of you after my first full quarter as HPE's CFO. Over the past three months, I have become even more excited about our opportunities across AI, hybrid cloud, and networking. We remain in the very early days of AI, yet it is already driving strong interest, pipeline, orders, and revenue across our portfolio from services to storage to financing. Our AI system revenue inflected positively in Q2. We are winning deals in the AI market now and are well positioned for additional demand from enterprises and software into the future. Our differentiation includes decades of liquid cooling expertise, which we expect to become even more in demand with future iterations of chips, including NVIDIA's. In short, we see AI as a long-term driver of our financial results and as a pillar of our strategy to pursue higher growth, higher margin revenues. We are very pleased that we have exceeded our expectations in Q2 across key metrics. We exceeded the midpoint of our revenue guidance by $400 million. Non-GAAP diluted net EPS was above the high end of our range, and free cash flow exceeded $600 million. Improving enterprise demand for traditional servers on top of the expected sharp ramp in AI servers drove the outperformance. Our AI orders and a healthy intelligent edge are set to grow sequentially beyond Q2 as expected, and AI emerged as a driver of healthy HPE GreenLake momentum. We are seeing rapid growth in AI system revenue. Overall, I am very pleased with our performance in Q2 and am excited about our continued progress through fiscal 2024. Let's take a closer look at the details of the quarter. Revenue grew 4% year-over-year and 7% quarter-over-quarter in constant currency to $7.2 billion. This exceeded the midpoint of our prior guidance by approximately $400 million. We have strong momentum in HPE GreenLake. The number of customers that have adopted HPE GreenLake rose 9% sequentially. ARR grew 39% year-over-year to above $1.5 billion in Q2. Storage and networking are typically the fastest growth elements of ARR and both retain robust growth rates. This quarter, AI was the fastest growth component of ARR. Our software and services mix rose approximately 200 basis points year-over-year to 67%. ARR is the best indicator of our model transformation to our as-a-service offerings. This growth validates what our customers are telling us, that HPE GreenLake is a key differentiator. We expect HPE GreenLake's value proposition to key customers, including enterprises and sovereigns, to sharpen with the advent of AI. Our Q2 non-GAAP gross margin was 33.1%, which was down 310 basis points sequentially and year-over-year, driven by a mix shift from our higher margin Intelligent Edge revenue to server revenue, plus an unfavorable mix within hybrid cloud. At Q2, non-GAAP operating expenses fell 1.6% year-over-year, despite our revenue growth of 4%. Our OpEx discipline partially offset lower non-GAAP gross margins and held the non-GAAP operating margin decline to 200 basis points sequentially and year-over-year to 9.5%. The OpEx discipline plus higher revenue drove GAAP diluted net EPS of $0.24 and non-GAAP diluted net EPS of $0.42. The latter exceeded the high end of the guidance range on strong revenue and cost discipline. And non-GAAP diluted net earnings per share excludes $247 million in net costs, primarily from stock-based compensation expense, amortization of intangibles and acquisitions, and other related challenges. We are managing the business with focus and discipline and evolving into a simpler, more agile company. We are also investing to capitalize on growth from the interrelated inflection points of AI, hybrid cloud, and networking and to drive structurally higher profitability over time. Let's turn to our segment results. Server revenues were $3.9 billion in the quarter. This was up 16% sequentially and up 18% year-over-year. Strengths in both AI systems and traditional servers drove the healthy revenue growth. Our cumulative AI system product and service orders since Q1 2023 rose approximately $600 million sequentially to $4.6 billion. I am very pleased with our AI system product revenue more than doubled sequentially to over $900 million. This strong revenue growth allowed us to make progress against our backlog, which is now $3.1 billion. Given the growing importance of our services business, we have updated our AI disclosures for this quarter to include services. Services is a small portion of our AI systems metrics at present, though we expect it to become more meaningful over time. Our differentiation with liquid cooling, software, HPE GreenLake, and increasingly, services is resonating in the market. We have seen a threefold increase in our enterprise AI customer base in the past year. Revenue from our traditional server business increased sequentially. We expect this trend to continue. Demand is improving, as enterprises digest prior purchases and gain more comfort with the macro outlook. Structural mix shift to higher AUP Gen11 servers is ahead of our expectations, and we are able to pass through rising input costs. We are encouraged that our Gen11 pipeline is starting to include AI inferencing activity and enterprise applications, and we see more evidence of adoption in the enterprise in Q2. Our Q2 operating margin was 11%. This was down 40 basis points sequentially and was in line with the expectations we laid out last quarter for our operating margins near the lower end of our long-term 11% to 13% range. While pricing remains aggressive in the server market, particularly in AI systems, we remain disciplined in cost and price, as we pursue profitable growth. Hybrid cloud revenues of $1.3 billion were up 1% sequentially and down 9% year-over-year. We are already seeing some cross-selling benefits of integrating the majority of our HPE GreenLake offering into a single business unit. I mentioned a 39% growth in ARR this quarter. Our traditional storage business was down year-over-year. The business is managing two long-term transitions at once. We talked about our migration to the more software-intensive Alletra platform. This is reducing current period revenue growth, so locking in future recurring revenue. Storage ARR growth of over 50% year-over-year offers early confidence into the migration. The second transition is from block storage to file storage driven by AI. While early, this is also on the right trajectory. Our new file offerings plus the sales force investment Antonio mentioned tripled our pipeline of file storage deals sequentially in Q2. Our operating margin was 0.8%, which was down 300 basis points sequentially and 110 basis points year-over-year. Reduced revenue scale and an unfavorable mix of third-party products and traditional storage was the largest driver of the sequential change. Intelligent Edge revenues were $1.1 billion. Revenues fell 9% sequentially and 19% year-over-year. Backlog consumption created difficult compares with both prior periods. Our backlog is now at normal levels. The demand environment remains soft, and large enterprises have yet to return to the market in force. However, we do see some green shoots that give us confidence in networking will transition to modest sequential growth beyond Q2, as we had expected. Our channel inventory remains within the normal range. Wi-Fi has grown sequentially for two consecutive quarters. Growth remains strong in software and services. Attach rates and renewals for Aruba Central, SASE and our AIOps software remain strong. The Intelligent Edge portfolio of subscription revenue grew above 50% year-over-year. The segment operating margin of 21.8% was down 760 basis points sequentially and 290 basis points year-over-year. As expected, the lower revenues, reduced mix of switching business, and the less revenue from backlog were the primary drivers. As we indicated last quarter, we have reset our OpEx plan for the year to account for lower revenue and expect the Intelligent Edge, operating margin to be back in the mid-20% range by Q4. Our HPE Financial Services revenue was up 1% year-over-year, and financing volume was $1.7 billion. Our operating margin of 9.3% was up 80 basis points sequentially and 40 basis points year-over-year. Our Q2 loss ratio remained steady below 0.5%. These results are what we have come to expect from this high-quality, predictable business. However, underneath these steady results, FS is already adapting to drive AI growth across the business. Year-to-date, nearly $0.5 billion of our $3 billion in financing volume went to AI wins with both cloud and enterprise customers. This illustrates our prior point that AI is driving demand to every one of our businesses. Turning now to cash flow and capital allocation. We generated $1.1 billion in cash flow from operations and $610 million in free cash flow this quarter. HPE typically consumes a significant amount of cash in the first half of the year and then generates cash in the second half. We are ahead of traditional free cash flow patterns thus far in fiscal 2024, given higher than expected net income in Q2, prepayments for AI systems, and the timing of working capital payments. Our cash conversion cycle was negative four days, which is a reduction of 28 days from Q2 '23. Our days of inventory and days payable were both higher to support our expected growth in AI system revenue in the second half. We returned $240 million in capital to shareholders in Q2, including $169 million in dividends and $45 million in share repurchases. Our year-to-date capital return is $386 million. Let's turn now to our forward view. We expect a materially stronger second half led by AI systems, traditional servers, storage, networking, and HPE GreenLake. Let me recap the key drivers that factor into our expectations for Q3 and the full year. For server, we expect improving GPU supply for AI systems and improving demand for traditional servers to drive sequential revenue increases through fiscal year 2024. While the rising AI systems mix is a gross margin headwind, we are balancing this with higher-margin services revenue, improving scale and cost discipline. We expect the segment operating margin to be approximately 11% for the fiscal year. For hybrid cloud, we expect slight sequential revenue increases throughout the year. HPE GreenLake growth should continue and traditional storage should improve slightly. We expect operating margin to improve modestly to the mid-single-digit range through the year as HPE GreenLake deals mature, new products ramp, and our sales force optimization gathers momentum. For Intelligent Edge, we anticipate slight sequential growth in Q3 and Q4, driven primarily by seasonal education spending rather than improving markets. We continue to expect our cost reduction efforts to materialize in the second half and our full-year operating margin to be in the mid-20% range. With that context, let me now turn to our outlook. For Q3, we expect revenues in the range of $7.4 billion to $7.8 billion. We expect GAAP diluted net EPS to be between $0.29 and $0.34, and non-GAAP diluted net EPS between $0.43 and $0.48. For fiscal year 2024, we now expect constant currency revenue growth of 1% to 3%, which is up from our prior 0% to 2% range. We reiterate our non-GAAP operating profit growth guidance of 0% to 2%. We are reducing our GAAP diluted net EPS guidance by $0.20 to $1.61 to $1.71 to incorporate the recent updates to our H3C proceeds. We are raising our non-GAAP diluted net EPS guidance up $0.03 to $1.85 to $1.95. This incremental $0.03 or $0.06 annualized reflects the contribution from the retained portion of our H3C stake. We are also increasingly comfortable with the high-end of the non-GAAP diluted net EPS range, given our OI&E and operational improvement. We are excluding from our non-GAAP results the gain on sale from our H3C and CTG divestments. This year's mix shift from networking to AI systems should weigh on our gross margins. We expect the fiscal '24 non-GAAP gross margin to be below our full-year expectation of 35% from our Analyst Day. To balance the mix shift, we are driving further simplicity and efficiency across the business. We are accelerating our generative AI capabilities such as implementing HPE-specific large language models and chatbots for our sales and service representatives. As I mentioned last quarter, prudent cost management, simplified processes, and disciplined execution across cycles are key tenets of our long-term journey towards higher margins. These cost actions will be evident in financial results in the second half of fiscal '24. We now expect fiscal year '24 OpEx to be down modestly from fiscal '23 OpEx. Our prior view was flat to down. This includes a sequential increase in Q3 for marketing before a sequentially lower Q4, which will serve us well heading into fiscal '25. We expect our fiscal '24 operating margin to be flattish year-over-year. We now expect OI&E to be less of a headwind this year. We anticipate a $150 million headwind versus our prior expectation of a $200 million to $250 million headwind, given a one-time benefit in Q2 and the retained portion of our H3C stake. We expect the effect of currency to be immaterial. Our strong first-half free cash flow increases our confidence that we will deliver at least $1.9 billion in fiscal year '24. We expect significantly stronger free cash flow in the second half of the year led by higher earnings, given our ramp in AI systems. This does not include the $2.1 billion that we expect to receive from Unisplendour this fiscal year, as a result of our recently restructured agreement to sell our stake in H3C. We expect working capital to be neutral to free cash flow as we expect declines in inventory to balance declines in accounts payable. We remain committed in the long term to our balanced capital allocation framework, including our target of returning 65% to 75% of free cash flow to shareholders. In the near term, we expect to continue share repurchases at a pace in line with Q2 as we prudently manage our balance sheet, ahead of the anticipated receipt of the H3C proceeds and the Juniper transaction closing. The proposed Juniper deal remains on track to close in late 2024 or early '25 as planned. We remain committed to our dividend and to our investment-grade rating. To conclude, our solid Q2 results illustrate how comprehensively AI is affecting our portfolio. We are capturing profitable growth opportunities in the AI market. We are excited to discover and look forward to seeing many of you at our IR Summit. I'll open it up now for your questions.

Operator, Operator

We will now begin the question-and-answer session. The first question is from Amit Daryanani with Evercore. Please go ahead.

Amit Daryanani, Analyst

Thanks a lot. And congrats on a nice sprint. I guess my question is really on the AI system side. You folks obviously had some really good revenue performance here. I think business pretty much doubled sequentially. I'm hoping you could spend some time talking about how should we think about margins on the AI side versus maybe segment averages because your margins appear to be holding up a lot better versus they have done at Dell or Super Micro, for example. So I'd love to just understand how do we think about AI system margins versus corporate averages? And if you just touch on what do you think differentiates your AI solutions versus companies like Super Micro, Dell, that'll be super helpful. Thank you.

Marie Myers, CFO

Hi Amit, good afternoon. It's Marie, and happy to take the question. And look, obviously, we shipped $900 million of AI revenue in the quarter. And I think as you saw, our margin rate on our service segment from an operating profit perspective was 11%. So look, honestly as the CFO, I'm very pleased that's what we guided and that's what we achieved, and that's what we've guided for the second half of the year. In terms of what's the puts and takes and the headwinds and tailwinds, let me give you some color around it. So from a headwind perspective, obviously we're dealing with a tougher inflationary commodity market, particularly even in the second half. And we also see, as you saw in the quarter, a greater mix of AI servers actually in our overall server mix. And as you know and you've heard, it's a pretty competitive market out there I might add too. But clearly, we're managing it very well. And I'd say what's benefiting us is a couple of really important tailwinds. One is what we are seeing in terms of the Gen11 transition. That part of the business brings a higher AUP. And then also frankly, Amit, it's the work we're doing around cost. So as a result, I think this really illustrates that we are very disciplined around both price and cost and frankly, it's part of our strategy to drive profitable growth. So pleased with the results, and that's how we are guiding the back half of the year, Amit.

Antonio Neri, President and CEO

Amit, regarding the differentiation, I’ll summarize it in four key areas. First, we have the ability to deliver and manage systems at scale, particularly with AI systems, which is a unique skill set built over decades. Second, we possess proprietary technology for infrastructure cooling, with capabilities to cool systems in three different ways. Our manufacturing footprint is the third element, as we have one of the largest water-cooled manufacturing operations globally, located strategically in the US and Europe near our customers. Lastly, the services aspect is critical; running systems at scale demands specialized services capabilities. This is why we’ve started highlighting the services pull-through with Marie, which can enhance gross margins over time. We handle every stage, from initial consulting to advisory and professional services during the design and build stages, and finally, operational services for ongoing management, including expertise in direct liquid cooling. All four of these elements significantly set us apart.

Jeff Kvaal, Head of Investor Relations

Thank you, Amit. We move to the next question.

Aaron Rakers, Analyst

Yes. Thanks for taking my question. I guess sticking on the AI topic, if I could first ask, when you referenced the AI enterprise customers starting to show up, and I think the comment on the conference call was, it's now north of 15% of your AI orders. Can you give a little bit of more context of that? What has that been over the last couple of quarters? I'm just trying to think about the trajectory of that. And then Antonio, on the liquid cooling side, as we and investors think about the Blackwell product cycle from NVIDIA, I'm curious of, can you be a little bit more specific of exactly where, from a technology perspective, you differentiate at liquid cooling? Is there something unique that HPE does within the 300 patents that you would want to highlight for us, as sustainably differentiated? Thank you.

Antonio Neri, President and CEO

Thank you, Aaron. It's accurate to say that enterprise demand began to rise in Q1 and significantly increased in Q2. In early 2023, the majority of demand was driven by model builders, service providers, and hyperscalers, as evidenced by our partnership with Microsoft. When I consider the market, I see three segments. The first is the dark segment, which is generating substantial demand for GPUs and accelerated computing. The second segment involves sovereign clouds, which are gaining traction with many customers worldwide. Lastly, we have enterprise customers, which will expand into the thousands over time. Currently, we are very satisfied with the momentum in enterprise AI. Numerous discussions are underway. During my recent trip to Europe, I visited London, Paris, and Madrid, where AI adoption is a priority. Customers are drawn to HPE for our trustworthiness, quality, and our ability to provide a simplified experience through the HPE GreenLake platform. In terms of differentiation, HPE utilizes three distinct cooling methods. The first is a traditional liquid-to-air cooling system, using a water supply in chilled locations to cool the air around the systems. This has been a standard method for some time. The second method, common in the industry today, is predominantly direct liquid cooling, or a hybrid of both liquid and air cooling, where competitors also rely on fans for cooling. While some competitors refer to direct liquid cooling, they are actually employing hybrid systems. HPE, on the other hand, has already deployed ten systems in the market that we are currently shipping and configuring for customers. We also have a unique method known as 100% direct liquid cooling, which we have been implementing for quite some time. At present, six systems are operational, three of which are for generative AI. As we look ahead to the next generation of silicon, like Blackwell, it will necessitate 100% direct liquid cooling. This presents a unique opportunity for us, as we possess both the intellectual property and the cooling infrastructure capabilities, along with significant manufacturing resources. That's why I emphasized in a previous call that HPE has one of the largest water-cooled manufacturing capabilities. This is what truly sets us apart. Even in hybrid cooling configurations, HPE does not utilize off-the-shelf solutions. Instead, we employ an adaptive rack cooling system, which is a one-of-a-kind design. Importantly, we can place direct liquid cooling and air-cooled systems within the same data center aisle, which is advantageous because customers prefer not to retrofit their data centers. This represents a significant opportunity for us.

Jeff Kvaal, Head of Investor Relations

Aaron, thank you very much. Gary.

Meta Marshall, Analyst

Hi, this is Mary on for Meta. I just had a question for you on Intelligent Edge. Where are you in the Intelligent Edge inventory digestion and when do you expect to emerge from it?

Marie Myers, CFO

Yes. Good afternoon. In terms of where we are at Intelligent Edge, I think as I commented in my prepared remarks, we saw Q2 as being the trough period, and we've been transitioning through that throughout the quarter. If you look at the market and where it's at right now, I'd say, our channel inventory right now is in really good shape. We did mention in the guide that we expect a modest sequential improvement in networking in the back half of the year.

Antonio Neri, President and CEO

Yes. And I will say in addition to what Marie said, what I'm really excited about is that we see interesting areas of growth happening now. If you go back to Mobile World Congress and you see even announcements like we made yesterday, the enterprise private 5G is picking up significant momentum. Of my almost 40 meetings I had in Barcelona, more than half were about enterprise private 5G.

Jeff Kvaal, Head of Investor Relations

Thanks very much, Mary. Gary.

Toni Sacconaghi, Analyst

Yes, good afternoon, and thank you for taking my question. I just wanted to follow up on the guidance. You talked about enthusiasm for the second half, but you beat revenues this quarter relative to your expectations by $400 million, and by guiding up an additional percent, you're actually only guiding up the full year by $300 million. So I'm wondering, are you just being conservative, given the commentary around enthusiasm and forces at work in the second half, or how do we reconcile that discrepancy? And then also just on AI servers for the second half, I think you talked about six-week to 12-week lead times. So if you have $3 billion in backlog and lead times for six weeks to twelve weeks, why can't you deliver $3 billion in AI systems like next quarter or certainly in the second half? Thank you.

Marie Myers, CFO

Yes, hi Toni, good afternoon. This is Marie. I'll address the first question regarding the guidance. We have adjusted our guidance to a range of $1.85 to $1.95, which reflects the impact of the 19% stake in H3C. Specifically, we've incorporated an increase of $0.03 related to that stake. I want to emphasize that I'm pointing to the higher end of this range due to our confidence stemming from the rise in revenue. We're observing growth at the top line, and I am also confident in our cost management efforts, which have been under strong scrutiny in recent quarters. Additionally, we experienced some favorable one-time items in OI&E this quarter. Overall, we're maintaining our guidance at $1.85 to $1.95, leaning towards the higher end based on the confidence you've expressed. Now, I'll hand it over to Antonio to address the second question.

Antonio Neri, President and CEO

Yes, Toni, I believe there is an opportunity to potentially exceed that. The main limitation is not supply, but the availability of data center space. I mentioned this in Q1; it involves data center space, power, and cooling. We are coordinating with our customers to time everything correctly, which could take six to twelve weeks, possibly less than a quarter, but installation is also required. A significant portion of our generative AI deals are actually under GreenLake. While we can recognize the revenue upfront, we are deferring all the service components. Ultimately, it will depend on the timing of the data center, power, and cooling. If everything aligns well, we may have the chance to do even better. However, we felt it was wise to maintain our current outlook and raise it by 1%.

Jeff Kvaal, Head of Investor Relations

Toni, thanks very much. Gary.

Joe Cardoso, Analyst

Hi, thanks for the question. This is Joe Cardoso on for Samik. So maybe just following up on the AI questions. You're seeing a sequential decline in AI backlog this quarter, admittedly off of a very robust revenue quarter. But maybe you can just discuss the pipeline you're seeing. Sitting here, like what are some of the puts and takes that investors should be considering when seeing the sequential decline in backlog despite what appears to be a strong underlying demand environment overall? Thanks.

Antonio Neri, President and CEO

Sure. Thank you. Yes, our backlog was slightly down quarter-over-quarter, also based on the demand we took in and also the fact that we converted more. I think you have to realize some of these deals, particularly larger deals take time and a little bit lump. Some of them have to go through the financing side. We feel good about where we are today. But in terms of the pipeline, think about multiples of the current backlog, multiples of the current backlog. As we go through the next weeks and quarters here, we feel very confident in our ability to capture that AI for all the reasons I described before, and also the ability to close these deals as some of them may include, by the way, the need to provide data center space as well because we are not just building the system. These are all generative AI systems, by the way, all of them that we also need to run it for customers.

Jeff Kvaal, Head of Investor Relations

Thanks very much, Joe. Gary.

Victor Chiu, Analyst

This is Victor Chiu in for Simon. The doubling of the AI system's revenue is a pretty sharp inflection this quarter. Can you just tell us how much of the increase was driven by allocation improvement versus the ability to ship orders? And kind of just help us understand the dynamic there a little bit.

Antonio Neri, President and CEO

No, I don't think it has to do anything with allocation. I read all this thing about allocation. We have a fantastic partnership with NVIDIA is about lead times in the general H100, which obviously NVIDIA made significant improvements on their own. As we start ramping our manufacturing processes, we actually become better and better at the revenue conversion. I would say, I want to thank my team because they did a fantastic job and we feel good about as we go forward. So it's a combination of multiple things, better lead times on supply. We feel pretty good about the ability to deliver systems six to twelve weeks to Toni's question, and some of them are straightforward is just shipments and some are with our HPE GreenLake offering wrapped around.

Jeff Kvaal, Head of Investor Relations

Thanks very much, Victor. Gary.

Ruplu Bhattacharya, Analyst

Hi, thanks for taking questions. It's Ruplu filling in for Wamsi today. Can you talk about how much of the GreenLake revenue and ARR growth came from AI? And then, Antonio, do you think that having GreenLake is helping you sell AI systems? And I also wanted to clarify, Antonio, you talked about sovereign AI. Can you talk about what innings you're in? And are there specific requirements of sovereign AI where you think HPE is well positioned to satisfy?

Marie Myers, CFO

Hi Ruplu, we look forward to seeing you tomorrow. So in terms of your question around just ARR and AI, actually, it was the fastest growth element of ARR in Q2, followed by storage and networking. So I think, as I said in my prepared remarks, we are starting to see AI to sort of move through our entire portfolio. So pleased with the progress that we had this quarter, and I'll turn it over to Antonio to add some more context.

Antonio Neri, President and CEO

Yes. On the second part of the question, I think we are early at this point in time. I think there is more to be seen. I want to make sure I captured exactly what you said. Can you repeat the last part because you were a little bit breaking.

Ruplu Bhattacharya, Analyst

Yes, I just wanted to ask about sovereign AI. What innings are you in, and are there specific requirements that you can satisfy?

Antonio Neri, President and CEO

Yes. So on sovereign AI, I said it's early, early on. I think there is a lot of engagement right now happening at the country level. The good news is we have a very good reach across the board. It's a combination of both. There is a combination of providing what I call generative AI locations where customers, enterprise customers can get access to a sovereign cloud that the government may be helping fund at the same time. The other one is what I call supercomputing power, right, itself. The two are very well aligned to the sovereign AI opportunities. I'm excited because HPE already provides a lot of sovereign governments supercomputing. Now we can extend into generative AI. If you think about the example of the UK Bristol as a generative AI system that the UK is funding as part of the Bristol University, that the venture is going to be open to start-ups and enterprises in the UK to either train models or do other type of research. That's what we see.

Jeff Kvaal, Head of Investor Relations

Okay. Thanks very much, Ruplu. Gary, we have two more questions, please.

Ananda Baruah, Analyst

Hi, good afternoon guys. Really appreciate, taking the question. Just a quick clarification and a quick question, Antonio. Did you say that in the last 12 months, AI growth was driven by service providers? And was it mobile, you said? And then the question is, what's a good way to think about cloud, your gen AI cloud scale opportunity, hyperscalers, and Tier 2s going forward? Thanks a lot.

Antonio Neri, President and CEO

Yes. No, thank you. I said that the segment where we have seen obviously the vast majority of action and demand is the model builders. Those are the big companies that build large and small language models. Obviously, you saw our comments about our partner with Microsoft and the extension of their capacity to OpenAI. That's an example of a model builder. But there are also other service providers. In fact, in my remarks, I mentioned Scaleway, which is a French service provider that provides the capacity for the local French model builders. There is a very vibrant ecosystem in France about building AI models that will use that capacity to train the models. That's what I referred to at this point in time. We are very interested not just in the model builders but Tier 2 and Tier 3, which will also be a big driver. However, understanding that enterprise, ultimately, the action is going to happen in terms of fine-tuning and deploying these AI models over time.

Jeff Kvaal, Head of Investor Relations

Okay. Thank you very much, Ananda. And last question, Gary.

Louis Miscioscia, Analyst

So thanks for getting me in. My question is really about GPU or accelerated diversification. Obviously, AMD and Intel and others are starting to come out. So as you think about calendar 2024, when do you think that your system will be for this? And what do you think demand will be and then maybe continue that into 2025?

Antonio Neri, President and CEO

I'm very practical about these matters. Currently, NVIDIA is the leader in generative AI, which is where we have focused our strategy and offerings. As I mentioned earlier, we already have 10 systems operating in a mixed cooling environment and six systems in direct liquid cooling. We are collaborating closely with NVIDIA, and that’s why Jensen will join me on stage at Discover to discuss our partnership. In the sovereign space, where supercomputing may play a role in the future, they are creating their systems with our assistance, and that will involve a mixed environment. Recently, we launched a new system at Los Alamos, which is an NVIDIA system combined with HPE and features direct liquid cooling. We're also preparing for systems to be introduced in 2025 that might utilize different kinds of accelerators. Over time, we will be aligned for timely market entry. Right now, our efforts are closely tied to NVIDIA. I invite you to join us at HPE Discover, where you'll have the chance to see everything we've discussed today on the exhibit floor. All the systems mentioned in my slides and the responses to your questions will be on display. These are systems and IP that we are currently shipping, and you'll witness our quick market entry with this silicon alongside all our services in HPE GreenLake. It will be a fantastic opportunity since we'll also be hosting the keynote at Sphere, where we expect over 17,000 customers and partners to be present. I appreciate your understanding that we have limited time for questions, but thank you for being here today. To summarize, the quarter was strong, with our AI system revenue more than doubling to $900 million, which allowed us to surpass our revenue and non-GAAP earnings per share guidance. With robust demand for AI, we have a significant backlog, a promising pipeline, and are seeing recovery in some traditional infrastructure areas. With our disciplined approach to pricing and costs, we are increasing both our revenue and EPS outlooks. There might be further revenue adjustments depending on timing. Regarding EPS, we're comfortable at the higher end of the range, as Marie noted. Overall, we are optimistic about the second half of the year and look forward to seeing many of you at HPE Discover in less than two weeks. Thank you for your attention today.

Operator, Operator

Ladies and gentlemen, this concludes our call for today. Thank you.