DELL - NYSE

Good afternoon, and welcome to the Fiscal Year 2025 Second Quarter Financial Results Conference Call for Dell Technologies, Inc. I'd like to inform all participants this call is being recorded at the request of Dell Technologies. This broadcast is the copyrighted property of Dell Technologies, Inc. Any rebroadcasting of this information in whole or part without the prior written permission of Dell Technologies is prohibited. Following prepared remarks, we will conduct a question-and-answer session. [Operator Instructions] I'd like to turn the call over to Rob Williams, Head of Investor Relations. Mr. Williams, you may begin.

Thanks, everyone, for joining us. With me today are Jeff Clarke, Yvonne McGill and Tyler Johnson. Our earnings materials are available on our IR website, and I encourage you to review these materials and the presentation, which includes additional content to complement our discussion this afternoon. Guidance will be covered on today's call. During this call, unless otherwise indicated, all references to financial measures refer to non-GAAP financial measures, including non-GAAP gross margin, operating expenses, operating income, net income, diluted earnings per share and adjusted free cash flow. A reconciliation of these measures to their most directly comparable GAAP measures can be found in our web deck and our press release. Growth percentages refer to year-over-year change unless otherwise specified. Statements made during this call that relate to future results and events are forward-looking statements based on current expectations. Actual results and events could differ materially from those projected due to a number of risks and uncertainties, which are discussed in our web deck and our SEC filings. We assume no obligation to update our forward-looking statements. Now, I'll turn it over to Jeff.

Thanks, Rob, and thanks everyone for joining us. We executed well in Q2, and I'm really proud of our team and our performance. Revenue was $25 billion, up 9%, with another record for our servers and networking business. Diluted EPS was $1.89, up 9%, and cash-flow from operations was $1.3 billion. Our AI momentum accelerated in Q2 and our results and outlook demonstrate that we are uniquely positioned to help customers leverage the benefits of artificial intelligence. In ISG, our AI server orders and shipments increased again sequentially. Our unique capability to deliver leading-edge air and liquid cooled AI servers, networking and storage tuned and optimized for maximum performance at the node and rack level combined with leading ecosystem partners and world-class services and support continues to resonate with customers. Orders demand was $3.2 billion, primarily driven by Tier-2 cloud service providers. Encouragingly, we continue to see an increase in the number of enterprise customers buying AI solutions each quarter. Enterprise remains a significant opportunity for us as many are still in the early stages of AI adoption. We are also excited about the emerging sovereign AI opportunity, which plays to our strengths given our position with governments around the world. We shipped $3.1 billion of AI servers in Q2. As we exited the quarter, our AI server backlog remains healthy at $3.8 billion. Most exciting, our AI server pipeline expanded across both Tier-2 CSPs and enterprise customers again in Q2, and now has grown to several multiples of our backlog. As we begin the second-half of the year, we have optimized our sales coverage to better focus on AI opportunities across CSPs, and both large and small customer segments and geographies. In addition, we have added substantial engineering capabilities, including data center networking and design to support these AI pursuits. Traditional server demand continues to improve and we saw strong demand again in Q2, our third consecutive quarter of growth and our fifth consecutive quarter sequentially as customers invest in both traditional and AI infrastructure. Dell IP core storage demand, including PowerMax, PowerScale, PowerStore and PowerProtect Data Domain grew double-digits in the quarter, a positive sign as we move into the second-half of the year. In CSG, we saw modest commercial PC demand growth in the quarter with healthy operating profitability and we expect growth in the second half of the year. In CSG, we continue to pursue profitable share focusing on commercial PCs, high-end consumer and gaming with our strong attach motion. We are optimistic about the coming PC refresh cycle, as the installed base continues to age, Windows 10 reaches end-of-life later next year and the significant advancements in AI-enabled architectures and applications continue. In closing, we are leveraging our strengths to extend our leadership positions and lean into new opportunities like AI. We are offering customers choice, flexibility and control of how and where they build, train and run artificial intelligence. We are still in the early innings, and our AI opportunity with Tier-2 CSPs, enterprise and emerging sovereign customers is immense. Our view is supported by an AI hardware and services TAM of $174 billion, up from $152 billion, growing at a 22% CAGR over the next few years. We are competing in all of the big AI deals and are winning significant deployments at scale. Progress will not always be linear in the early stages, but we are winning in the market with strong feedback from repeat customers while acquiring new customers every quarter. We have the right AI portfolio with more to come, the right services capabilities, and we are optimizing our sales coverage to capture this once-in-a-generation opportunity. With the coming IT hardware recovery cycle and our positioning in AI, I really like our hand. Now over to Yvonne for the financials.

Thanks, Jeff. Let's start with an overview of our performance. Then I'll dive into specifics of ISG, CSG, DFS and guidance going forward. In the second quarter, we delivered strong operating results and solid cash flow, both of which position us well for the second-half of the year and beyond. I'm encouraged by the great momentum we're generating in ISG with AI leading the way. Our total revenue was up 9% to $25 billion, including the headwinds from the exit of our VMware resale business. And our combined CSG and ISG business grew 12%. Gross margin was $5.5 billion, or 21.8% of revenue. This is down 230 basis points due to an increase in our AI optimized server mix and a more competitive pricing environment. Operating expense was down 4% to $3.4 billion, or 13.7% of revenue. Let me emphasize, we expect solid top line growth in the second-half of the year even as we continue to optimize our cost structure to enhance our competitiveness over the long-term. A big part of this optimization effort is leveraging AI to reimagine our business processes and drive higher productivity. To that end, in Q2, we took a $328 million charge for workforce reduction as we continue to position our business for the long-term. Now, let's look at operating income. We delivered a 3% increase to $2 billion or 8.1% of revenue. This was driven by higher revenue and lower operating expenses, partially offset by a decline in our gross margin rate. Finally, Q2 net income was up 7% to $1.37 billion, primarily driven by stronger operating income. And our diluted EPS was up 9% to $1.89. So, that's a look at the whole. Now let's talk about the specifics, starting with ISG, where we delivered strong performance. ISG revenue was a record $11.6 billion, up 38%. Server and networking revenue was another record of $7.7 billion, up 80%. Server demand continues to outpace shipments with strong growth across traditional and AI servers and the mix of our AI optimized server demand grew sequentially again. Storage revenue was down 5% at $4 billion. We were pleased to see double-digit demand growth across our core storage portfolio, including PowerMax, PowerScale, PowerStore and PowerProtect Data Domain. This strong storage performance was offset by headwinds in the partner IP portion of our HCI portfolio. ISG operating income was up 22% to $1.3 billion, due in large part to AI-optimized server revenue growth and associated gross margin dollars. Our ISG operating income rate was up 300 basis points sequentially to 11% of revenue. This rate improvement was the result of operating expense scaling, driven by higher server revenue and storage profitability. We're pleased with the sequential improvement in storage profitability. We gained scale with 6% sequential revenue growth. We were more disciplined in our pricing. We had a higher mix of Dell IP storage solutions and a better geographic mix with more North America activity. As we mentioned last quarter, we expect our ISG operating margin rate to continue to improve in the second half of the year. In CSG, revenue was down 4% to $12.4 billion. Commercial revenue was flat at $10.6 billion, while consumer revenue was down 22% to $1.9 billion. CSG operating income was $767 million or 6.2% of revenue due to a more competitive pricing environment. We expect growth in the second-half of the year, particularly in the fourth quarter. The coming PC refresh cycle and the longer-term impact of AI will create tailwinds for the PC market. Dell Financial Services originations were up 5% to $2.4 billion in Q2. Strength in both client and ISG financing offset the exit from our VMware resale business and the sale of our consumer revolving portfolio in Q3 of last year. Normalizing for these two impacts, DFS originations were up more than 30%, proof of increasing interest in our customer payment solutions. Now let's move to cash flow and the balance sheet. Q2 cash flow from operations was $1.3 billion. This was primarily driven by sequential revenue growth and profitability, offset partially by working capital. Our cash conversion cycle was negative 43 days, up four days sequentially, driven primarily by the increase in our AI server business and AI order linearity. We ended the quarter with $6 billion in cash and investments, down $1.3 billion sequentially. This is the result of capital returns of $1 billion and net debt paydown of $1 billion during the quarter. We repurchased 5.5 million shares of stock at an average price of $130.03 and paid a $0.45 per share dividend. Since our capital return program began at the beginning of FY '23, we've returned $9 billion to shareholders through stock repurchases and dividends. With our additional debt reduction and increased profitability this quarter, our core leverage ratio was 1.4x. Turning to guidance. Indicators continue to point towards growth in the second-half of the year. Against that backdrop, we expect Dell Technologies FY '25 revenue to be in the range of $95.5 billion and $98.5 billion with a midpoint of $97 billion or 10% growth. We expect ISG revenue to grow roughly 30%, driven primarily by AI and ongoing momentum in our traditional server business. We expect CSG revenue to be flat to low single digits for the year. We expect the combined ISG and CSG business to grow 13% at the midpoint. We expect our gross margin rate to decline roughly 180 basis-points due to inflationary input costs, the competitive environment and a higher mix of AI optimized servers. We will continue to drive efficiencies in the business and expect operating expense to be down low single digits for the year. We expect both ISG and CSG operating margin rates to be within our long-term financial framework for the full year, 11% to 14% and 5% to 7%, respectively. We expect interest and others to be roughly $1.4 billion and an annual non-GAAP tax-rate of 18%. Diluted non-GAAP EPS is expected to be $7.80 plus or minus $0.25, and up 9% at the midpoint. For Q3 of fiscal year '25, we expect revenue to be in the range of $24 billion and $25 billion at the midpoint of $24.5 billion, up 10%. We expect the combined ISG and CSG businesses to grow 14% at the midpoint with ISG up in the low-30s and CSG flat-to-up low-single digits. We anticipate operating expenses to be down low-single digits sequentially. Our operating income rate is expected to improve as we continue to drive profitability in ISG. Q3 diluted share count should be between 714 million and 718 million shares. Diluted non-GAAP EPS is expected to be $2 plus or minus $0.10. To close this out, I'll echo what Jeff said. We are very optimistic about FY '25 and beyond. AI and the coming IT hardware refresh cycle will be tailwinds for our business and no one in the industry is better positioned than Dell. We are applying artificial intelligence and beginning to realize the benefits across our own business. We're using it to improve customer and team member experiences in sales, software development, services, content management and our supply chain. And in turn, we're using our experiences to help our customers realize benefits of AI for themselves. Before I turn it back to Rob for Q&A, I'd like to share that after 32 years at Dell, Rob has decided to retire. For those who may not know, Rob began his career at Dell in 1992 in Corporate Treasury while attending graduate school. He's held many roles since then and has seen Dell continue to transform and grow over the last three decades. This will be Rob's final earnings cycle. We wanted to personally share the news with you all today. Paul Frantz will lead IR going forward. Paul is a longtime finance leader at Dell and a familiar name to many of you. Paul and Rob have been working closely over the last few years to ensure a smooth transition. Rob, I'd like to thank you for all you've contributed to Dell's success over so many years. You will be greatly missed. I'll save my extended thank you and goodbyes for later. For now, I'd like to turn it back over to Rob for Q&A.

Thank you, Yvonne, and thanks for those kind words and congrats to Paul. Let's get to Q&A. We ask that each participant ask one question to allow us to get to as many of you as possible. Let's go to the first question.

We'll take our first question from Amit Daryanani from Evercore.

Best of luck with the retirement. I guess my question is really around ISG margins. This could really step up from 8% in Q1 to 11% in this quarter, despite the 82% sequential growth you had in AI servers. So can you just talk about what's enabling this kind of margin expansion because really a lot of your peers on the AI server side are struggling with -- trying to -- are struggling with their margins, I feel right now. So I'd love to understand kind of what's enabling this margin expansion. And then critically, as we think about this 11% to 14% target for the full-year, what are the key inputs of building blocks together in the back-half of the year? Thank you.

Thanks, Amit. And I will get started on that one. So we were very pleased with the operating income rate we saw in the second quarter, though 11%, up 300 basis points quarter-over-quarter. That was really driven by improvement across the entire portfolio. I mean, first, revenue was up quarter-over-quarter 26%, which helped drive scale within the P&L. And as expected, the headwinds we saw in Q1 did not persist into Q2. In storage, we had scale. We're price disciplined. We mixed more towards our own Dell IP storage offerings, which was very helpful and saw strength in North America enterprise. In the traditional server space, the demand environment continued to improve, although we're still seeing some competitive pressures. And in AI servers, we had strong shipments with improved profitability and growing enterprise customers in that portfolio mix. I'd say, we do expect ISG operating income to finish FY '25 within our long-term framework that 11% to 14% and do expect as we move through the second-half of the year that we'll continue to see that as we mix more towards our storage portfolio as we do each year.

Yes, Amit, maybe two additional data points that I think are worth noting in the improvement. One as Yvonne said, our Dell IP core storage. On a demand basis, it was up double-digits. We improved the margins in each of our categories, PowerMax, PowerStore, our PowerScale and PowerProtect, data protection quarter-over-quarter through the price discipline that Yvonne mentioned. And the second one, which I think is very important, we improved margins of our AI portfolio. And we did that with the same sort of price discipline, but more importantly, the engineering value-add and the technical value-add that we're bringing to our customers and the expansion from beyond the specific node to the rack level deployment. So our ability to add L11, L12 capabilities, expert deployment, system validation and testing, the ability to help engineer the solution at our customer site, the extension into networking, the ability to cable these things up and deploy them at scale, that is allowing us to extract additional value in our AI deployments.

All right. Hi, thanks for the question, Amit. Let's go to the next question.

And our next question comes from Ben Reitzes with Melius Research.

Hi, can you hear me okay?

Yes.

Yes.

Okay, great. Hi, how you doing? And congrats, Rob. I can say I knew you back when, but I won't. Congrats and so listen, I'm just trying to reconcile your guidance, Yvonne. And then I have a follow-up. It looks like there -- it's more 4Q loaded than what the Street was thinking and there was the talk of the layoffs into the print here. And I was just wondering if you can reconcile that for us, the $2 and then it looks like maybe the Street has to raise the 4Q vis-a-vis where they were. And I was wondering if that has to do with timing of savings, and I was wondering also if you can quantify the savings, what the $328 million kind of equates to in savings and how we should think about that? And then, Jeff, if you don't mind if you could just mention how the AI server margins are progressing and what we should think about with the flat backlog Q-to-Q? I know you had great sales, but should the flat backlog be a concern? I know there was a lot in there, but since I congratulated Rob, maybe you'll give me a pass and answer them all. Thanks.

That was the greatest one question at a time of five that I've seen.

That was a classic, Ben.

We may have Rob answer...

We'll see if we can get through at least the first part of those questions, Ben, will let Yvonne start.

All right. So for -- I could talk to the second-half, but let me start with Q3. We were talking about CSG and ISG, combined are going to grow at 14%, right? Total revenue at the midpoint of $24.5 billion, so up 10% year-over-year. So 14% for CSG, ISG, 10% in total. ISG up 30% year-over-year. So very, very impressive continued performance there, driven primarily by AI servers as well as continued momentum in traditional servers. So we're expecting to see that. CSG, we're expecting to be flat-to-up low-single digits in the third quarter. I'm expecting gross margin rates to be up slightly quarter-over-quarter. And I'd say, remember that we're expecting inflationary environment in the third-quarter, in the second-half actually holistically as well as a continued competitive environment. OpEx, which I think you have a -- somewhat a question on, we've already guided to that OpEx holistically and that will be down, we did that last quarter. So we expect that to be down about 2% quarter-over-quarter. And then from an OpEx standpoint, expecting continued improvement there driven by ISG and storage and traditional server mix as well as performing within our long-term framework holistically. I don't know if there's anything to add-on the third-quarter as we going to the full-year. So I think your question on, are we waiting more towards the fourth-quarter than the third-quarter? We're sequentially higher in the fourth-quarter from a historical perspective and expect to continue that tradition. And we are expecting a CSG recovery to begin in the second-half, more weighted towards the fourth-quarter. And then you'd also have your natural largest quarter from a storage perspective in that fourth-quarter. And so you'll see that would be reflected both the momentum in CSG as well as the storage momentum in the profitability and performance in that period.

And then on the two AI questions, the AI server margins, I'll start there summarizing what we discussed with Amit is, margins were up sequentially. We were able to extract value out of our offer. We were able to extract the value out of our AI offer by expanding our -- again, our value proposition to our customers beyond the box to the rack level. So our ability to sell services around the box, the ability to sell networking and network integration, the ability to sell what we call L11 and L12 capabilities, which think of it as factory integration, solution testing and on-site deployment. The ability to configure these at-scale for our largest customers is allowing us to extract differentiated value of our competition. We're going to continue to focus on that and continue to find more ways to help our customers deploy AI at-scale. In terms of the backlog, I mean, the first thing I would point to is shipments nearly doubled quarter-over-quarter. And I think that's an important thing to think through. For us over the last 12 months, we shipped $6.5 billion of AI infrastructure to our customers on the demand line of just under $9.5 billion. What you're seeing is improved supply. Still working with customers on their delivery dates, still working with customers on their ability to accept the equipment and integrate it into their data centers. I'm not worried about the backlog as well. I think we've made reference before of our five-quarter pipeline. Our five-quarter pipeline remains healthy. I think I made reference that it's several multiples of our backlog. We have visibility to a pretty significant pipeline, and it's our job to convert that over the coming quarters.

All right. Thanks.

Thanks, Ben. Clearly exciting. Exciting opportunity for us. All right. Next question.

We'll move to our next question from Erik Woodring with Morgan Stanley.

Great, guys. Thanks so much for taking my question. And Rob, just echo what everyone was saying, you'll be missed. Thanks for all your help over the years. Jeff, I appreciate all the detail that you provided in your prepared remarks on the makeup of your AI optimized server customer base, increasing number of enterprises, expanding pipeline in Tier-2 CSPs and enterprise. Is there any way you can just maybe expand on those comments a bit? And what I'm really trying to get a better understanding on is just that the concentration of spending within your backlog, within your pipeline for AI optimized servers and how that's changing for you? Thanks so much.

Sure. Let's see where do I start with that. Obviously, the pipeline I talked about is growing, growing at the size of multiples of our backlog. The composition of that is continuing to see the number of enterprise customers grow and the amount of revenue in that pipeline is growing, which to me is encouraging and signals enterprise customers specifically are moving from experimenting to now piloting the technology to do a vast number of things, some of the use cases that Yvonne mentioned earlier. So we're encouraged by the number of enterprise customers growing quarter-over-quarter. We're encouraged by the number of enterprise customers in our pipeline and the revenue that represents. Clearly, some of these large Tier-2 CSP deployments, training foundational models are still a large percentage of the backlog, large percentage of the pipeline. And I think that's just the scale, but we are seeing the continued growth of enterprise, which again, I think is very, very encouraging, whether that be in life sciences, higher education, financial services, some of the national labs, et cetera, we're seeing expansion across the enterprise. When I look at the composition of the backlog and technology, it ranges from parts that are available today to parts that are available in the future and everything in between, our backlog represents things that we can ship in Q3, things that we can ship in Q4 into next year. And obviously, a five-quarter pipeline represents the same thing, which extends into next year. Eric, was that enough color? Was that helpful?

Yes, no, that's perfect. I appreciate that. Thank you. Thanks for the detail.

You're welcome.

And we'll move to our next question from Toni Sacconaghi, Bernstein.

Yes, thank you. And Rob, congratulations on your retirement. We will miss you. I just wanted to -- I have one confirmation and a question. In terms of improving profitability in AI servers from Q1 to Q2, did gross profits on AI server gross profit percentage on AI servers improved materially between Q1 and Q2 or just operating profit dollars because there was clearly significant OpEx improvement?

The short answer to that is both improved.

Okay. And then I was hoping maybe you could just provide a little more color on the storage dynamics because you talked about Dell IP businesses growing double-digits and yet the storage business was down 5% year-over-year. So maybe you can help dimension what else is in that equation that suggests that hyperconverged is either an enormous part of your business or was down by a huge amount or there's some other piece that we need to be able to make those numbers pull together. So perhaps you can provide a little more color on that. Thank you.

Sure. Yvonne and I'll try to tag team this one. But I'll start with, as I mentioned previously, if you look at our Dell IP portfolio and the success we had in Q2, it was good to see. We had double-digit growth as we referenced on a demand basis across the high-end products, the mid-range products, the unstructured products and the data protection products. Those products from a margin contribution point-of-view are significantly greater than the partner IP references that Yvonne mentioned earlier. So when those businesses are growing and I think I've just mentioned in the previous questions, and their margin rates improved with better price discipline in the quarter, we got, if you will, leverage from both growth as well as improvement in the gross margins, which will ultimately make its way to the operating margin for Storage. And that was significant. The thing that was driven by some mix of geography, North America that Yvonne mentioned earlier, but primarily driven by the core products themselves. And then obviously, your conclusion is correct. The partner IP business was down quarter-over-quarter. That's less profitable for us and the mix shifted increasingly more towards our Dell IP products.

I mean, sorry about that. Yes, go ahead.

I just -- so the double-digit growth was sequential growth in the...

Yes.

That you were seeing in those businesses.

The double-digit...

ISG revenue growth was up 26% quarter-over-quarter.

So the storage growth.

Yes, that was year-over-year, Toni. The core comment about double-digit growth was up year-over-year.

Correct. So - but the whole year I mean if you kind of running year-over-year.

Yes, so you've got a decline in conversion, hyperconverged infrastructure and it is -- we haven't broken it out specifically, but it is obviously a large part of the business. And data protection would be the other area where there was a bit of decline year-over-year. So those two offset the increases in the other four businesses we mentioned.

Thank you.

And our next question comes Asiya Merchant with Citi.

Great. If I could just you know, talk a little bit about CSG. And Rob, again, thank you for all the help and support over the last few years. We'll miss you as well. On CSG, if I may, you sound very optimistic about growth here in the fourth-quarter. Maybe if you could just double-click on that a little bit and give us some -- a guidance on what you're seeing in the pipeline or the demand indicators that gives you confidence in that growth as we enter calendar '25? Thank you.

Sure. I'll take a run at and then Yvonne can help come over-the-top. I mean, we remain optimistic about the refresh. I mean, I think it's reflected in our guidance that we think the refresh is shifting and more towards the end-of-the year than we thought maybe at the middle of the year. I know all of you have done your supply base checks, it would indicate the same thing that refresh is heading towards end of '24 into '25. And what's important about that is, as the refresh takes longer to start, history suggests it snaps back faster because the Windows 10 end-of-life date is not moving. So we have a Windows 10 end-of-life date. We have an aging installed base of machines bought during the COVID era, all mounting to be refreshed with exciting new products built around AI and more AI applications are coming. And we remain optimistic about that recovery. Calling the timing has been difficult. But the end-of-life is around the corner or a quarter closer. The installed base is bigger and older, exciting new products are coming, applications to help productivity with end users is around the corner. And if you think about the extension of AI out to the Edge in inferencing and what inferencing will be done on the Edge on PCs, that opportunity is immense as well. When you think about running these small language models with larger memory footprints on the Edge on your PC to do amazing things and having a personal agent on your, if you will, screen helping you along the way, that's all in front of us. So we remain very optimistic about that. Closely from our performance, you can see that we had called a recovery a little earlier and recalling today, it's a little further out. Yvonne, anything you would add?

No, I think you hit that, Jeff.

Thanks, Asiya. I appreciate it. Next question.

And we'll move to our next question from Samik Chatterjee with JPMorgan.

Hi, thanks for taking my question. And Rob, congrats and thank you for the help. I guess, Jeff, if I just go back to your comments about what you're seeing from enterprise customers in terms of AI server demand. I'm just curious to understand, clearly you're excited about that building pipeline that you're seeing from the enterprise side, but what are you seeing when enterprise customers come in? What is that giving you in terms of opportunity to either do like attach of more services or storage for that matter? And how should we think about how much of a differential on the gross margins does that lead to when you compare to sort of what you're doing, the work you're doing with the Tier 2s?

Sure. Maybe in order that at least I think about this. First of all, we still believe we're in the very, very early innings in deployment enterprise. And we know that as we've surveyed thousands of our enterprise customers. And we have a pretty good understanding of where they are in their AI journey. And if you think of this as in stages, they are truly in what we call stage 0 and 1 trying to determine what they're going to do, what their strategy is. And a large percentage of our customers are in that state. And I think we have a pretty good represent -- pretty good understanding of that, given our broad market coverage with the largest sales force and technology reaching down into small businesses and everything up to large multinationals. We have a pretty good pulse of where enterprises, business or commercial customers are on that journey. And that gets us really excited that they all see the opportunity. Many are experimenting, moving to piloting the technology, so they can actually take advantage of the opportunities to drive productivity or an advantage with customers or customer service or what have you. And in doing so, it really brings us tremendous opportunity. We've built professional services to help customers do exactly the four things they need to make progress here. Build a strategy, help them implement it, ultimately think about how they deploy the model and then adopt it and scale it to get the advantages of it. We've built services around that. We continue to scale that capability. And then because these are very complex systems, the ability to engineer and to help our customers integrate these. You just can't put this into a data center. There has to be space, it requires power. In many cases, it requires new forms of cooling. The network interconnect is very complex. The storage subsystems have to be high-speed systems, three-tier architecture systems with the flexibility and efficiency and scale to really run at high-performance. That's kind of what we do. We get excited about that opportunity and to help customers really extend from a strategy to implementation to be able to run these at-scale and to take advantage of this enormous capability is exciting. And then if you think about the use cases, the use cases are generally around large language models, small language models are coming. We're already talking about technologies like agents and agent is a buzzword of our industry today, and we've just scratched the surface of how agents are going to help each individual's productivity, help with specific skills, which provide an opportunity to take the millions of processes that are in businesses today and help automate them. That's what's in front of us. And I think we've said each of the last several calls and we'll say it this call again, the margins selling to enterprises are better than the margins selling to our largest customers. And that's the value we continue to provide and we'll continue to build upon. I think I made reference to this as well. I think this is a very important thing. We've tuned our go-to-market model where we have a dedicated pursuits team for the largest CSPs and we have a overlay specialty organization that covers our largest enterprise customers down to many thousands of them with AI specialized skills and then the engineering capability behind it to help the sales force win each and every engagement. I know that was a lot. I think I covered the questions.

All right. Thanks, Jeff, and thanks. Thanks, Rob. I appreciate it.

And we'll move to our next question from Michael Ng with Goldman Sachs.

Hi, good afternoon. Thanks for the question. And I also wanted to extend my congratulations to Rob. I just have two questions if I could. First for Yvonne, I wanted to ask about the low 30% ISG revenue guidance for the upcoming quarter, which implies ISG revenue is down sequentially. Could you just talk a little bit more on the drivers of that? What are you assuming for AI servers to happen quarter-on-quarter in terms of revenue? And then second, for Jeff, I was just wondering if you could talk a little bit about some of the partnerships announced in the last couple of months, the new Nutanix Hyperconverged partnership, the NVIDIA SuperPOD certification, certification for the Dell PowerScale. Should that help drive an inflection in storage revenues as we head into the latter part of the year and into next? How material are these partnerships? How much does it change your strategy in storage, if at all? Thank you.

Yes, I'll start off, Mike, your question. For Q3, we do expect ISG revenue to grow in the low-30s year-over-year, and for the full-year in 30s also. Traditional servers and storage are roughly in-line with historical sequential business that we see. Servers will grow in the low-single digits and storage will be down in the low-single digits. So that's relatively normal sequentially. So that would imply that AI servers are down quarter-over-quarter. Now that said, shipments are or were up in the second-quarter to $3.1 billion. The backlog is sitting at $3.8 billion. The pipeline, as Jeff has already talked about, is a multiple of that backlog. And enterprise orders, customers and pipeline are all growing. So with all that said, I would say if we have the GPUs and the customers are ready, we're fully motivated and ready to ship more AI servers in the third-quarter, but that is what's embedded in our guide.

Awesome.

Let me see if I can work my way through the partners question and what's happening. I think there are two distinct things. One, if you think of HCI as a category, at least from our view, it really is a category of products with a single hypervisor. It's really meant to drive simplicity and ease-of-use for that use case. And if you think about what customers, particularly enterprise customers, in fact, almost solely enterprise customers, large and small are, as they think about AI, they don't have space. They need to make space. They need to create room for bringing AI gear in. So there's a consolidation of foot in business. And HCI is a wonderful tool to be able to consolidate traditional applications. We have a good product with VxRail and we clearly added to that to give customers choice with our Red Hat product, with our Azure product and the Nutanix product that you described. And we'll continue to do that. Probably what I think is really interesting is these new modern workloads and certainly all those around accelerated computing are demanding more performance, more scale, more flexibility, more efficiency. I think I've said this before, you have to feed the beast. These new architectures are demanding some data, they gobble data up and you need to feed it in a high-performance way. We think the best way to do that is through a three-tier architecture. We think it meets the demand -- the high demands of these new workloads. We can drive high performance, we can scale-up, we can scale-out. We can drive the flexibility and efficiency needed and getting SuperPOD certification with our PowerScale products, the F910 and F710 is essential to be able to work in that environment, bringing the very best of our company going-forward and it's why we made the announcement since we're talking about announcements. In May, earlier this past quarter about Project Lightning bringing a high-speed parallel file system to the forefront as well because that's another opportunity to expand our position in storage and quite frankly, expand our margin opportunity around storage, around GPUs and around what's happening in AI.

All right. Good. Thanks, Mike.

Thank you.

Next question. You're welcome.

Our next question comes from Wamsi Mohan with Bank of America.

Yes, thank you. And Rob, congrats as well. Yvonne, you just said quarter-on-quarter decline in AI server revenues, your backlog is flat quarter-on-quarter in AI servers. Is that really reflective of just Blackwell delays? And does your backlog include the liquid-cooled offerings like XE9680L? And should we expect an acceleration in that backlog then over the next several quarters? And Jeff, you noted higher margins in enterprise consistently. Can you also comment on where sovereign might fit in that? And can you also help us with which models do you think will drive Enterprise and Sovereign versus maybe 9680 primarily at Tier-2 CSPs? Thank you so much.

All right. You all are killing me on the multiple questions.

Okay. You need one question, one question. Rob, you need one question is three or four. Let me see if I can work my way through that. So when I think of the backlog Wamsi, I think, adding opportunity of deliveries from customers next quarter and then in Q4. And then clearly into next year and that implies Blackwell. We have sold our most advanced architecture aligned to Blackwell to a number of customers. We have sold H100s and H200s and availability, more importantly, customer availability to take the product, which is what Yvonne is reflecting in our guidance. She didn't make a demand statement, she made a shipment statement. So demand with that five-quarter pipeline that I described that is now multiples of our backlog is converting the backlog or converting that into orders as quickly as we can. That opportunity is in all sorts of architectures, the vast majority with NVIDIA, H100s, H200s and Blackwell, as well as a couple of others opportunities around AMD and Intel, but the vast majority is NVIDIA. So our work as a pursuits team selling to enterprise is to get the order, can obviously convert that pipeline into the opportunity into the order and get into backlog, and then our supply chain team needs to get the parts, which isn't -- availability has improved, it continues to improve. I think Jensen made reference to that yesterday that availability will continue to improve. That will help customers deploy and then customers have to be ready to deploy. So that's what we're working through. That's the best reflection of that of interpreting those signals into the guidance that Yvonne talked about is what we know today. And obviously, we're trying to improve that. And you asked about sovereign. Our opportunities in sovereign, I think are large and immense as I described in our prepared remarks. As it stands now, none of the five-quarter pipeline reflects the sovereign opportunity of any great size.

All right. Hi, thanks for the question, Wamsi. Let's go to the next one.

Our next question comes from Aaron Rakers with Wells Fargo.

Yes, thanks for taking the question. Rob, on your last call, I'll stick to the one question and congrats to you and Paul. I guess my question is not AI. It's actually on the traditional server side. Looking at the numbers, it looks like you've seen a pretty good uptick in this last quarter. I'm just curious of how you would characterize, I'll call it, the pipeline of opportunity that you're seeing in the traditional server market, the duration of demand, just kind of thinking about that potential kind of recovery cycle that you're seeing in outside of these AI servers for traditional servers? Thank you.

Sure, great. Great question, Aaron. Thanks. So maybe a couple of things to reinforce the momentum we're seeing is now five quarters in a row of sequential growth, three quarters in a row of year-over-year growth. We are seeing unit growth. We are seeing TRU expansion, i.e., more cores, more memory, more storage. And I think three things are driving the demands in course servers today. The first is we're coming out-of-the longest digestion period ever. It was eight quarters, two years. The installed base is old. And as it's aged, what I'll link to a conversation I just said made moments ago is that customers are looking for room to put AI into their infrastructure. And to do that, we think there's a consolidation that is beginning to occur to make room. And that consolidation is important because the new technologies today are incredibly more efficient than what you might be replacing that's four or five years-old. For example, if I was to look at a 14G product we shipped four-plus years ago to a 16G that we ship today, our product today has 2.5 times to 3 times more cores in it. It's 25% to 35% more power-efficient and a single 16G server can replace 3 to 5 14G servers in a rack. So consolidation is going to occur because that space and power is needed. And then lastly, as I think about it, we've made some references to this in the past. Workloads are coming back from the cloud, they're repatriating. And as they repatriate, they got to go somewhere and they're coming back on-prem on servers and storage. Those would be the three things I point to.

All right. Hi, thanks, Aaron. I appreciate you playing with us here.

Thank you, Aaron.

We appreciate that. Let's move to the next question.

And our next question comes from David Vogt with UBS.

Great guys. Thanks for taking my question. And Rob, congrats again and Paul, congrats. And I'm going to stick to one question as well. Obviously, you were clear about the GPU and the delivery constraints on the AI server side. But how are you thinking about how that gets reconciled as we move through your fiscal year into next year. Is it more GPU availability? What are the gating factors that you're thinking about? And does that mean as we exit this year, you're still going to see some constraints in terms of the ability to convert backlog and orders into revenue and we should expect backlog to continue to grow at a faster clip until that's resolved at some point, I would imagine at some point next year. So any thought or help there would be...?

What I hope we conveyed is, supply is improving and supply is improving. Our deployments and scheduled deliveries to customers is what we have to manage through. Again, much of this is very complex deployments, readiness of a data center, readiness of power, readiness of cooling, if it's direct liquid cooling, the ability to have water in the data center. All of that infrastructure has to be put in-place and coordinated with the delivery of a GPU and a server. And that's what we're working through. There's clearly opportunities that we're working across our customer set of what technologies they want to deploy, urgent need, if it's a need that can go into next year, they're clearly picking technology that will be available next year. So it's that composition of demand that we're trying to translate into shipments that were -- again, we're reflecting in our guidance and this is our best understanding of it today. Again, try to reinforce that our five-quarter pipeline is multiples of the backlog. There's demand out there. Our Pursuit teams are out working to convert that into orders. And as it converts into orders, an improving supply situation that we have to go work on.

Yes. I mean, we do expect to grow again next year, right? And we expect AI momentum to continue and we expect, as we talked about a little bit previously, the CSG business to grow also. And so we're really excited about next year. And as we get closer to it, we'll update you on our expectations.

Great. Thanks, Jeff. Thanks Yvonne.

I appreciate the question.

Congrats again, Rob.

Thanks, David. You bet. Next question?

And our next question comes from Simon Leopold with Raymond James.

Thank you. And Rob, congratulations. Congratulations to Paul as well. I'm going to stick to the one question as well. I wanted to see if you could unpack a little bit your thoughts on the storage trend as it relates to AI. I think in the past, you've talked about a lagged effect behind the compute side of the business. I'm wondering if you could talk about how you expect storage to benefit in terms of degree and timing. Thanks.

Well, I think consistent with our previous discussions and what I've tried to highlight today is the opportunity for storage is vast, it's large. Again, these large language models consume lots of data, and they need it fast. So the opportunity for us is rearchitecting, improving performance, more IOPS in the data categories that are most interesting to our customers here. The fastest-growing category is unstructured, right. When you look at what Dell is doing in the structured storage space, we are investing more on in unstructured storage apart from our own proprietary IP. Our scale, as an example, building on the F710, F910, Project Lightning, our parallel file system, the opportunity to help customers provide that data to these GPU engines in the form of moving from text to rich media types like video, audio, other forms of 3D images and complex datasets like that, different modalities. That is what we're building high-performance storage subsystems for. And it's a huge opportunity for us both from our largest customers that buy today and enterprise customers that buy into the future. I hope that helps.

Yes. Hi thanks, Simon. We're going to try and get one last question in here real quick. So please one question only.

And our final question comes from Steven Fox with Fox Advisors.

Thanks for squeezing me in. I assume the only reason Rob is retiring because he has a big book deal coming or something. But congrats. So I guess just for my one question, I mean, there's been -- there was a lot of talk in different metrics on pricing and also a lot of talk in terms of mix, et cetera. I guess from a big-picture standpoint, Jeff, like how do you feel about just available pricing power versus your ability to offset new market entrants or just more competitive entrants on any of your areas of expertise going-forward into next year? It seems like that's going to come up a couple of times. Just wondering if you can summarize it. Thanks.

Hi I'm making an assumption, so I'm going to clarify it. In AI, in GPU?

Actually, I mean, yes.

Infrastructure broadly across the portfolio.

Broadly across the portfolio. Yes, just strategically. Thank you.

Well, if you -- our industry has always been competitive. I do not expect that to change, whether that's with new entrants or traditional competitors that exist in each of our categories. So if I was to look at commercial PCs, clearly as a refresh is slower and starting than all of us would like, it's a competitive environment today, slightly different than it's been over the past 40 years. The same would be true in consumer, a larger number of competitors there. The consumer market is lagging relative to the commercial market in terms of its performance or its growth and it's a competitive environment there. If I work my way down through servers, when a large percentage of traditional servers are consumed by a smaller number of customers, the large bids are very competitive. That hasn't changed. I don't expect that to change. And we're seeing the same thing in GPUs. There's a lot more in it, a lot more participants. And I'd go back to what makes Dell special and why I think we win and continue to win. One is our coverage model. We have the broadest coverage model in technology, a combination of our direct sales force, and our partner programs. We're going to continue to invest in coverage. We're going to continue to broaden our channel programs to cover all customers who want to buy DellGear. And those who don't know they want to buy DellGear, it's our job to train them to want to buy DellGear. I look at the advantages of our supply-chain. I think it's unmatched in this industry. It provides us a source of competitive advantage and we're continuing to invest in that. It's one of the areas where internally we're applying AI. I think about what we're doing in R&D, the particular product categories that I described. The investments that we are making in all of these categories keep us competitive and in many cases, we're a leader. I think we're the market leader in commercial PC revenue. We're the market leader in displays. We're the market leader in workstations. We're the market leader in traditional servers. We're the market leader in storage. We're the market leader in data protection. I could go on. We have a number of number ones and we're going to continue to invest to differentiate and win in those categories. And then lastly, the fourth thing that we think makes us special and different is we service our own products. We take care of our customers and we're expanding those services. In AI, as I mentioned before, it helped with strategy to help define the infrastructure to help them deploy and implement and then ultimately adopt and scale. We do that broadly across the smallest companies in the world to the largest companies in the world. That's who we are. That's what we're going to invest in. As we think about the efficiencies that we're driving in our organization, it's to allow us to do those better.

Well, that was about as good a close as I could ask for, Jeff. Appreciate that. We look forward to seeing everyone out on the road over the next couple of weeks. Michael and Yvonne will be at Citi next week, and Jeff will be at Goldman the week after that. Have a restful Labor Day weekend. Thank you.