On this episode of The Six Five Webcast, hosts Patrick Moorhead and Daniel Newman discuss the tech news stories that made headlines this week. The handpicked topics for this week are:
- NVIDIA Q2FY25 Earnings with Special Guest Dan Ives
- VMware Explore 2024
- Salesforce Q2FY25 Earnings
- Dell Q2FY25 Earnings
- Pure Storage Q2FY25 Earnings
- HP Inc Q3FY24 Earnings
For a deeper dive into each topic, please click on the links above. Be sure to subscribe to The Six Five Webcast so you never miss an episode.
Watch the episode here:
Listen to the episode on your favorite streaming platform:
Disclaimer: The Six Five Webcast is for information and entertainment purposes only. Over the course of this webcast, we may talk about companies that are publicly traded and we may even reference that fact and their equity share price, but please do not take anything that we say as a recommendation about what you should do with your investment dollars. We are not investment advisors and we ask that you do not treat us as such.
Transcript:
Patrick Moorhead: Welcome back. The Six Five is live. It is our Friday show, and there is nothing I like to do more than crank this thing out. Daniel, how are you doing my friend? And Dan Ives, how the heck are you?
Dan Ives: Great to be here with the two goats.
Patrick Moorhead: Got this. I don’t know, man. It looks like a triple goat. Can there only be one goat? Okay. Maybe we could… I don’t know. You’re the goat.
Daniel Newman: I’m happy to be the third wheel in this group, man. What a good-looking group of guys, and nice-
Dan Ives: Look at that hat.
Patrick Moorhead: Daniel, what is that? Can you give me a little backstory? Okay. Maybe you all knew it, but Dan and I took a little short trip into Aspen. I’m here right now. I usually say I’m in an undisclosed location, but Daniel, why the cowboy hat? Aren’t you from Chicago?
Daniel Newman: No. I’m a native Texan now. I don’t know if everybody understands the importance of me committing to my Texas roots, although I did spend my first 40 years in Chicago. I’ve moved here. I got boots about three months ago. My daughters got me boots for my birthday. And then, who the heck has boots and no hat in Texas? Now I’ll admit, the dress, I’m not quite there yet, but piece by piece. By the way, when I heard Dan was coming, Ives, I felt like I had to do something.
Patrick Moorhead: You have to.
Daniel Newman: The best dressed guy in tech. I feel so ordinary, white button down shirts, blue sport coats. Maybe I could pull this off, Dan. What do you think, Dan? Could I do this? Could I do the whole ensemble? I normally do the goofy sneakers, and then pull this off with it and maybe start going on air that way.
Dan Ives: And then, maybe even change the colors of the cowboy hats.
Daniel Newman: Yeah.
Dan Ives: I know Moorhead starts wearing a pink sports jacket, dude, who knows?
Patrick Moorhead: We can’t ape the Dan Goat here. We’ve got to come in with our own special brew here. But it was funny, Dan, we were in Kemo Sabe getting Dan this cowboy hat and literally saw a guy that was aping your outfit, right?
Daniel Newman: Yes.
Patrick Moorhead: Purple shorts. Purple shorts. Man, he was lit up like a Christmas tree. It’s pretty awesome. I almost asked him, are you trying to knock off Dan Ives by chance?
Dan Ives: Strong respect, but strong respect.
Daniel Newman: Yeah.
Patrick Moorhead: Let’s jump into this show here. Obviously, special guest, Dan Ives Wedbush. We’re going to be talking NVIDIA earnings. We’re going to break down VMware Explore, Salesforce earnings, Dell earnings, Pure Storage earnings and HP earnings. But before we dive into that, I do want to make a, what’s a disclaimer for Daniel Newman and Pat Moorhead? We are not equity analysts, we are industry analysts and don’t say anything we say for investment advice. Dan Ives, on the other hand, does do that, but hey, before we dive into these topics, gosh, we got to talk a little Apple, maybe a little Meta. Dan, we saw you. It was funny. It was super Wednesday with NVIDIA, but I saw you talking to Apple. What’s going on here?
Dan Ives: Look, when you’re talking, you guys have spoken about the revolution in AI, it’s now going to start coming through Apple. I think look, from a consumer perspective, they will ultimately really own, in our opinion, the AI. I think 25% of the world, when they access AI, it’s going to be through an Apple device. So, my whole point is we can talk, of course, Godfather of AI, Jensen, Microsoft, but when you think about AI, and you guys do a great job talking about it, it’s really about second, third, fourth derivative. And that’s also why you could OpenAI in terms of these arms raised bidding process. Does Apple get involved? Does NVIDIA get involved? They’re the golden child of AI.
Patrick Moorhead: Yeah, totally. And real quick, got to get this out. It’s so funny. People are like, “Oh, I don’t believe in the super cycles.” Okay. Let’s throw super cycle out. Let’s say we move iPhone refresh from four years to three because we know there’s only one model that does Apple AI and that’s the highest end 15. And everything else does not do the good stuff. If Apple ships something that is super compelling, we saw a little bit of it at WWDC, then I believe that we will be entering a super cycle. So, Daniel Newman, a lot of talk about OpenAI out there, $100 billion valuation, maybe Apple is going to throw in some money, maybe NVIDIA is going to throw in some money. What’s your quick take?
Daniel Newman: Yeah, I appreciate it, Dan Ives, the retweet on the little Apple Android ecosystem comparison. I think that’s where this is heading-
Dan Ives: That got a lot of… So many haters came out there, a lot of dialogue on that.
Daniel Newman: Yeah, they always do.
Dan Ives: But that was great… I thought that was one of the best tweets that I’ve seen this week in tech.
Daniel Newman: Well, thank you. And by the way, it was really interesting because yesterday, I was on with Tyler on Power Lunch. I was supposed to talk NVIDIA and they’re like, “We’re enough NVIDIA. Let’s talk about this Apple news.” And so, I jumped on and talked about that and I said, look, if they make Siri digestible, they integrate OpenAI, they make Siri work, they start building it all natively into the productivity, the calendar suite, the mail suite, and then they start making the right connections with the apps. That’s absolutely going to drive this cycle, people starting to feel it in the front line of use. Siri has been terrible for a long time. It’s been really not useful. But if that usefulness starts to pop up, I think that’s going to drive things really quick, Pat. The only thing I’ll say really quickly on the analogy is look, we’ve got two parallel channels of development going on. You’ve got AllOpen, Grok with Meta, these companies that are going to build and basically make it available to the world developers.
And then, you’ve got this more closed, which the irony of OpenAI being closed. And then, of course you see Anthropic and others. And I think what we’re going to end up with is these two massive ecosystems. It’s not going to be one is going to win and the other is not going to win. And it’s not going to be two even. But Meta and OpenAI are setting the standard to be the OpenAI is the Apple, Meta is the Android. And that is going to build a list of players down the stream that are going to either go open, go close, but it’s going to go fast. And I love all the haters, but I’m just going to say this right now, Pat. They’re all wrong. They’re all wrong. The three of us are going to be right and time will be the proof.
Patrick Moorhead: I agree. Just 30 seconds on Meta. They made a massive disclosure yesterday. Check out my analysis on forbes.com. And they not only gave stats like downloads on Hugging Face, which is 10X year over year, $20 million in the past month. And that’s essentially downloading a model to go and do something with cloud token volume. And this is the measurements of the biggest cloud. CSP, 10X increased January to July, Meta says largest CSPs, I’m just guessing that’s Azure and AWS. Elite partners, and these are early access partners, 5X increase since LLaMA 3.1. This is absolutely taking off. And what I loved was their enterprise take, which was case studies from AT&T, Accenture, Goldman Sachs, DoorDash, Nomura, Shopify and Zoom.
And a few of them had hard numbers like AT&T, 33% improvement in related response. Accenture, 70% productivity and 20% to 30% quality improvement in report creation. Shopify, $40 to $60 million inferences a day. We’ve got closed, we’ve got OpenAI. They’re both rocking. Guys, let’s dive in. It was super Wednesday two days ago, man. Dan Ives, you were pretty much on every TV show all the time. You have coined NVIDIA’s CEO. Jensen Huang is the godfather of AI. Dan Ives, let’s start off with you. Talk to me about earnings. What are your thoughts?
Dan Ives: Look, I get in terms of people who will pick about whisper numbers for the October quarter, and to me, it’s a masterpiece. In other words, there’s no way that you could deny this revolutions happening. And I think it’s not just about the numbers that they put up in terms of $30 billion, $32.5, that’s probably going to be closer to $34 billion. Blackwell, Dewey’s, all the sort of noise. Put that to rest, my stock takes a little breather, it gets hit a bit. I think it goes higher. I think it goes into all-time highs and this thing gains steam because the most important thing, and I think you guys have talked about it, it’s derivative. Second, third, fourth, what does it say for Salesforce, ServiceNow, Palantir, Microsoft, Oracle, hyperscalers? It’s a trillion dollars of CapEx. It’s the fourth industrial revolution. You come off that call, I think it’s more of a, you get the popcorn ready and now it’s waiting for the second, third, fourth derivatives.
Daniel Newman: Yeah, and I’ll double down on that. I’ve talked about network effects. I’ve said we’ve seen net zeros, net one, so all the designers, chip manufacturers, everybody involved in the semiconductor supply chain should be doing well. The ones that are going to do well are doing well. And then, you’ve got this flow out, the OEMs, we’ve seen some boom bust. They’ve got to figure out how to add more value, make more money, but they’re getting the opportunity by being attached. The Dells and HPs, it’s up to them to figure out how to attach the services. The layering of the cake with the devices we just talked about with the applications, talked about ServiceNow, Raptor, the Now Platform, Salesforce, Benioff came out this week really ambitious. But look, the other thing, and I know we only have a few minutes, I want to make sure, Pat, you get to talk before Dan has to leave, but I’m really get frustrated with this whole people looking at a digestion. This is not the internet cycle. We are not going to see this thing implode the way the internet imploded. It’s a different cycle. There is a digestion period.
When you spend $100 to $200 billion in CapEx, you cannot stand that up, offer services instantaneously, show revenue and Satya, Zoc, Sundar, they’re all basically sitting there. They’re saying, “We will not let this cause our demise by not investing enough.” They all understand it. And so, when all the bears are like, “Oh, there’s no one using it.” First of all, Pat, you gave a bunch of great instances who they are, and that’s just BS. They are using it. But we do need to understand there is digestion between the CapEx in and the OpEx out.
Patrick Moorhead: No, this is good stuff. Listen, I net it out like this. There’s nothing stopping this train for 12 to 18 months. Downstream does not matter for 12 to 18 months. You’ve got hyperscaler FOMO. Nobody, whether it’s Meta, Microsoft, Google, Amazon, it’s pedal to the metal on CapEx. And until we see investors, their investors demanding reduced CapEx, the beat really goes on. And I want to hit the growth areas really quickly that I don’t think people are fully comprehending. The first one is the first build out was with the hyperscalers. The next build out is in tier two CSPs. And we heard Dell talk about this last night and let’s put Supermicro aside for a second, but Supermicro and Dell are the two tier two CSP drivers. The third element is the enterprise.
And we saw a little bit of that outlined. And right now, as Dell’s Jeff Clarke said, the enterprise is really in phase zero moving into phase one POCs. Third growth area is content. They’re not just selling GPU chips, they’re selling racks, they’re selling trays, they’re selling GPU chips, they’re selling multiple networking chips. And oh by the way, they’re selling cables. And let’s not forget software. Every new generation, including Blackwell, we have seen NVIDIA sell more software. And that software started off 15 years ago as low level drivers. And here we are with Blackwell generation called NIMs, and these are actually microservices applications and you’re seeing them be spit out or cranked out by multiple ISVs. And this is essentially the easy button for driving this forward.
Yes, there’s competition, but it’s not going to matter because the market is growing so heavily. Biggest competitor are hyperscaler, their own custom ASICs, AMD is the closest direct competitor and they will likely end the year, I believe at 10% market share, not for the whole year, but ending their calendar year. We will see Intel likely making an impact in ’25 and ’26. They’re making a small impact with current Gaudi today, but it’s only at $500 million. And compare that to AMD, who’s at $4.5 billion forecast. All right guys, Dan, we got to get you out here to go on more TV shows, talk to all of your clients.
Dan Ives: I’m going to Asia again this week. So, again, it comes down to that’s where… Like I always say, you can find AI revolutions in the 10th floor of your office building Metro North for Jersey Transit. And that’s why I think many of the bears, they continue to focus on their spreadsheets and try to find AI in their spreadsheets.
Daniel Newman: Hey, Dan, just as you run off, two things. One is I’ll be running to Berlin for… I’ll be at IFA checking out what they’re doing on these devices. But Pat, I think we need to give Pat a shout. I need to commend this nerd. We were GenAIing what differentiates each of us as analysts. We’re having a little fun. And Pat is the deep technical guy. Did you see this guy call out the mask change on fork? He called out the mask change in real time.
Patrick Moorhead: I appreciate that.
Dan Ives: But look, Patrick is like here, hold my rose, let me deal with this.
Patrick Moorhead: Well, it was funny. Now, I felt a little bad because I literally brought this out live on CNBC, nobody was talking about it stock dips and I’m reading CFO, and then it’s like, what’s a mask set? And it was funny. I was like I’ve never really had to explain a mask set. I tripped a little bit. And then, about a minute later-
Dan Ives: You were talking deep diver.
Patrick Moorhead: Listen, I really do try to simplify it, man.
Daniel Newman: You said Bloomberg’s headline said that they had a mass change. They didn’t-
Patrick Moorhead: Mass.
Daniel Newman: Jensen said… I know, but the headline was Mass Change, M-A-S-S Change.
Dan Ives: All right, guys, thank you so much for having me. Awesome. Thank you for everyone watching.
Patrick Moorhead: Thanks. Take care, buddy. There we go, Dan Ives. That was fun. Now it’s back to us. This is good. No, I appreciate, by the way, Bloomberg deleted that tweet.
Daniel Newman: Did they?
Patrick Moorhead: Yeah.
Daniel Newman: Yes, that was a mass change. That was actually not just the tweet though, Pat. That was actually the name of the article. There was an article headline and then it was the whole thing. So, look, I actually will tell you I listened to him. And if you’re not a real techie and you’re not familiar with what a mask change is, and I mean I know more about it now than I ever did. I’d heard the term. That was as far as I’d really ever spent. And you listened to Jensen, you might not have heard that clearly on the first listen.
Patrick Moorhead: Yeah. So, every chip change that you make that I’ll call it a metal spin as opposed to a poly spin, needs a new mask. And a mask is what is created to go into a foundry like TSMC, Samsung Foundry and IFS. So, Dan, 15 minutes on NVIDIA and other stuff. Let’s dive in. I’m going to call my own number, VMware Explore 2024. Dan and I were both in Las Vegas at that show. We got a rare opportunity to talk with Broadcom’s CEO, Hock Tan, right after he got off-stage doing his keynote. And it was pretty awesome. It was one of these post-game interviews just to break stuff down. Key highlights, this was all about the private cloud. VCF, so today, 15 years ago, the public cloud was born. It was created by AWS, and it was essentially running infrastructure and services that were not on-prem and not in a colo that you’re renting and managing.
And developers wanted to go there because of the flexibility that they would get. It used to take four months to spin up the compute resource, six to spin up storage. And my gosh, sometimes nine months to spin up networking. And going to the public cloud, you could spin it up immediately. You could swipe a freaking credit card to get in there and you had almost unlimited resources. Fifteen years later, you have companies like VMware, Nutanix and Red Hat building stacks that run on-prem that mimic the public cloud at a substantially lower cost. As long as you’re keeping those resources busy. Hey, if you don’t want to manage your infrastructure, you can do that at a colo and do something in between. So, VMware Cloud Foundation, otherwise known as VCF or the short name of it, they went to 9.0.
And that was going from 5.2 to 9.0, but we won’t go down that path. It doesn’t even matter. That was aligned with ESX. But anyways, here’s the upshot. We have been waiting, VCF has been around for a while, but my take is this is the first time it’s actually ready with a full stack and what the company is calling advanced services to go along with it. And one of the advanced services is private AI. You have security services. You have essentially good hygiene services. You have all the… I’m not saying it has as many services as AWS, but it’s getting pretty darn close. The only thing I’m trying to figure out, Dan, is how do you get the scalability of the infrastructure? Now, out of the other side of my mouth, if you want to have guaranteed instances with AWS, you’re not going to buy those on the spot market at the lowest, lowest price. You need to make a one-year, two-year commitment. And if you want to share services across regions, that’s going to cost you a lot as much. And there’s a reason why AWS has 20, 30 points of Op-Inc in there.
And the other rule of thumb, public cloud, the profit they make, AWS, again, nothing wrong with profits. Every dollar invested into CPU instances, you can charge 3X. Every dollar you put into GPU instances, you can charge 8X. And one of the most interesting thing with this private AI offering was that VMware upped its game on virtualizing the GPU. GPU’s virtualization started close to 25 years ago, and I had first started it even at AMD, allowed you to virtualize the CPU, and then they virtualized the storage, and then they virtualized the networking with NSX, and now virtualizing the GPU so it’s not just sitting idle. And I think what I heard out there was GPUs are idle 30% of the time because they’re hard to schedule them, especially when you’re bopping between even doing inference. So, super impressed with the progress the company has made there. Dan, I definitely left some oxygen, a lot of oxygen for you.
Daniel Newman: Look, Hock Tan, and we had the opportunity to sit down with Hock on a couple of occasions. We did a deeper dive conversation and we did a postmortem immediately after his keynote. What I loved most about it was he is very clear in his vision on what he wants to do. You hit a lot of the product and capabilities, Pat. But look, he’s using terms now like AWS on-prem. And we’ve heard that and you mentioned it a little bit, but this is a sea change right now of he’s not just saying, I feel like the hybrid cloud and on-prem companies have been apologetic. You’re still going to need on-prem as you make your journey to being an all cloud company, we want to help you.
He’s saying I want to challenge the C-suite, the boards of directors on really making a substantial pivot away from public cloud, and that is different than the attached to public cloud strategy we’ve heard. We have heard GreenLake has been focused on this control plane. Some of the things you can do with Red Hat enables some significant hybrid and multi-cloud capabilities. But he’s of saying VCF becomes your control plane. It becomes your first stop. It becomes the main place to build your enterprise IT stack. And his business model is also being carried out in real life too, Pat, right in front of our eyes. He’s doing what he does. He’s saying we’re going to take, what, 8,000 SKUs down to just a handful. We’re going to make this easier to buy.
But also, you’re going to have to make a bigger commitment. So, there isn’t really price increases, but you have to buy all the stuff. So, there is a price increase if you were only buying vSphere, for instance, and seeing that play out. It was interesting, Pat, in the room, the initial part of the keynote felt like everybody was a little quiet. You heard things like, “Hey, everyone,” and it was like pin drop. But as Hock presented through that opening keynote, you actually seemed like he helped people. He put them into a more comfortable situation, they dropped their guards a little bit, and they opened up and you could feel the energy pick up. By the end of the presentation, people were starting to see.
Here’s the bottom line, Pat, if the thesis is true, and I think you and I should test this in our economics lab is if the thesis is true, then there is a case to be made that workloads remaining on-prem or new workloads being deployed on-prem. And by the way, there’s not a winner takes all here. I just want to be clear. There’s not a winner takes all. There will be public cloud. He thinks it’s just for bursting. It’s almost the inverse of what the cloud talks about with prem. But Pat, in the end, I would be very surprised as numbers start to come out that you don’t see that he’s executing. I just think people underestimated this coming in. They overestimated the power of the ecosystem and Hock doesn’t make these decisions lightly. So, I was encouraged, but I think the numbers over time will tell the whole story.
Patrick Moorhead: What I’m really interested in seeing, Dan, and we need to be very careful, you could just check the growth rates of hyperscaler enterprise cloud providers and compare that to the growth of the on-prem folks. And I think you mentioned this with GreenLake, there are some really good green shoots out there, but we need to see 50%, 60%, 70% growth in on-prem Cloud for those numbers to compare to Azure, and AWS and Google Cloud. AWS is $100 billion company at this point. I’m all over the Barclays thing. 83% of CIOs are going to move a workload back on-prem. They’re not talking about 50% of their workloads. I do expect though, and this is probably the most important piece, these will be strategic workloads that are steady state. But again, the thing that’s stuck in my craw is how do you get instant growth in that?
And that’s why I think the workloads that will be on there will be ones that are steady state, and then the ones that are bursty, those applications you will put in the public cloud. And the biggest fight is going to be now is we’re you going to do the inference, baby? Where are you going to do the inference on that 80% of the enterprise data that’s on-prem? So, hey, any final words, Dan, you want to make or you want to move-
Daniel Newman: No, there’s always a lot to cover, but we’ve got 30 minutes and five more topics. Four more topics, four more topics, right?
Patrick Moorhead: Yeah.
Daniel Newman: So, let’s give them their due coverage.
Patrick Moorhead: Yeah. So, let’s dive into Salesforce here. Dan, we’ve got a, did AI make a difference? We’ve got a CFO that’s leaving. Again, we’ve seen a ton of executives being replaced, and leaving and moving on. What are your thoughts about earnings?
Daniel Newman: Yeah. So, Salesforce had a bit of a banger. And again, I made the comment tough data report on Wednesday, and some people didn’t understand what I meant by that, but what I basically meant is the entire news cycle seemed to be very heavily NVIDIA centric as the entire economy was weighing on that. But companies like Salesforce and some of the others we’ll talk about today actually also reported on that day. Look, they beat, they beat, they beat. Now again, I’m going to give you both sides here. So, the model that Benioff has put out from a standpoint of revenue and EPS growth has been largely on the money. And so, that is a really encouraging thing. He talked a lot about Agentic AI and the idea of deploying these agents as being a big part of the company’s AI strategy.
And so, it was indicative, but what I didn’t really hear in the earnings, there was a lot of the AI narrative and that’s what we mentioned earlier in the conversation about how this plays out in the multiple network effects, but we’re still not entirely clear how much people are paying for additional Salesforce because of AI yet. You got Data Cloud, which I think is a powerful play for the company, but they’re executing. They got pretty lean last year. They made a lot of operational efficiencies in the business that was various overheads and of course personnel related. But the company is also… They’re delivering almost $10 a share annually on earnings. They are a very profitable company. And then, the first day of the year, they have a huge chunk of their revenue already accounted for because so much of it is subscription, the beauty of SaaS businesses.
Pat, the growth has been a little bit modest, if you want me to be like you’re talking about you run down the line, I think it was about 8% revenue growth. They’re growing in sales services platform in that 10% range, a little slower in marketing, a little faster in integration and analytics. But that’s one of those things where you look at it, you go, well, they’re trending to $38 billion. They’ve hit a lot of large numbers. They can only do so many price increases. So, a lot of their variable growth has to come through either net revenue expansion, which means selling more products to their current customers or it has to come from going out and winning net new business, which is difficult in the economy that we are in because it is still tougher out there than most of us want to acknowledge when it comes to this.
And these tools and technologies are expensive. But where I will leave this, Pat, is it’s a really healthy business. It’s very profitable. It’s very growth oriented. It’s very sticky. What do I want to hear and see as we head into Dreamforce this month? I want to hear and see how the company is going to incrementally grow revenue, expand revenue at the customer level by adding AI capabilities. And then, I want to understand who’s paying for it. Are customers willing to pay a dollar more, $10 more? Remember, you got hundreds of thousands of companies running on this technology. If you can get every user to pay a dollar more, that’s massive in terms of impact. So, what is the AI impact? And Pat, this will be the topic du jour every day for me for the next 10 years, at least two, as we try to see how this CapEx turns to OpEx. And Salesforce is an absolute bellwether for whether or not AI is being consumed at the enterprise level.
Patrick Moorhead: Yeah. So, Salesforce is one of the earliest movers on enterprise AI. Even before generative AI was cool, or maybe three months after we had the OpenAI explosion, they rolled customers out, customers like Gucci that actually were… This is more machine learning than generative AI, but they’ve done the best job being able to show some of the potential impacts. And what I think is happening is they’re balancing two worlds. And we saw Benioff talk about this one or two quarters ago when the numbers just weren’t growing like people had expected. And people were expecting NVIDIA growth numbers when this first came out. But what we had during the pandemic is we had customers buying a lot of tools, new tools, and there’s been digestion on that and people are assessing, you know how this goes, Dan, “Hey, before we buy new tools and pay more money, I want to make sure that we’re getting the full value of the stuff that we bought.”
And I do think that’s what’s happening in the customer base of Salesforce, even though in some of their key verticals and applications, they’re the clear industry leader. And we see Marc Benioff tweeting, “Hey, we’re the number one in CRM for the 80th year.” But within that install base, moving that forward is tough. There were a lot of questions on the call about AI and it was interesting. Most of the questions seem defensive like, “Hey, in the era of AI, you’re not going to need as many seats.” And then, Benioff talked about, well, we’re going to consumption as opposed to number of seats. And we think that that is going to grow and it’s going to sound like a broken record, but their Data Cloud, Data Cloud growth is always a precursor of great and big things to come.
And I don’t always nail it, but when GenAI hit, the first thing I said is you got to get the data right. And sure, we’ve heard garbage in garbage out for the past 40 years, but it’s amplified when it comes to GenAI. The difference with generative AI is that you are mixing types of data that you never thought that you would mix before. That’s the one thing. And the second thing is that the ways of extracting and doing prompt engineering, you can get data out of way. Gosh, why are we on Dell earnings here? You can get yourself into a mess and bring private data out into the public and that can be very bad. And the final thing is if you don’t get your data, you can potentially issue completely garbage results that you have to stick with, with a customer. As we saw with Airlines, he was doing a three for one and they had to give the customer that based on that mistake. Looking forward to the Big Tent event in San Francisco coming up, I think it’s going to be all GenAI and it’s going to be all Data Cloud.
Daniel Newman: Hey, are you hearing me?
Patrick Moorhead: I am, I am. I’m not-
Daniel Newman: Yeah. Okay, I’m back. My super killer brand-new laptop, I’m not going to say who it’s from. It just totally froze up on me.
Patrick Moorhead: That was weird. Yeah, that was weird. I’m on my new AI PC laptop.
Daniel Newman: You look great, by the way, on your new laptop.
Patrick Moorhead: Thanks. I’m using the old fake looking camera. We’ll see if that works. So, hey, I’m going to call my own number here on Dell earnings. Dell had quite a banger in their Q2 FY25 earnings. I alluded to a little bit of it, but they had an absolute banger in their data center group. ISG, which is the data center group, up 38%, but what was fire were the AI servers up freaking 80%, amazing. Storage was off by 5%. And you know what, storage usually follow servers. In the call, they talked about how good the profitability was. I believe that that storage being off was about non-Dell technologies storage, maybe VAST Data. So, they pull in the revenue for VAST Data, but it’s at a lower margin compared to their own. And again, VAST Data is not a storage company, it’s a software company, but they did need to likely pull in a third party for that.
Maybe it wasn’t VAST Data, maybe it was Pure, probably Pure, not VAST Data, but they’re not giving those types of details. And the interesting point too, the thing I really loved, and Jeff Clarke just did a great job on bringing this out and guys, I really recommend to go listen to a conference call. You need to go in and read the transcript or look at the conference call, but Jeff just did a masterwork of talking about what is going on here. ISG is targeting two AI server categories, tier two CSPs, these are not the non-hyperscalers, and then enterprises. Tier two used to be like XAI. We saw Michael Dell tweet about that installation, and we see that half that went to Dell and half that went to Supermicro. You have these GPU-focused companies that again, do GPU as a service. Dan, you’ve been talking a lot about these guys, but those are very much on the move.
And then, there’s enterprises. Jeff talked about, okay, we got through experimentation and we’re just going into POCs, so not like these large installations of POCs. I was a little surprised by that. But that would also indicate how HPE and their numbers not moving as much in enterprise AI, big in HPC AI, but not so much in enterprise AI. Backlog, 3.X billion and the pipeline was multiples, didn’t give us a number, 4X, 5X, 3X, but multiples on top of that. So, the pipeline looked strong, a good backlog, which is about twice as much as they did in the last quarter. Final comments, a little surprised. A little bit surprised on CSG, the PC group, particularly with what we saw from HP and Lenovo was off 4%. Huge consumer declines, around 22%, and commercial was flat. HP commercial was up. Not too certain what’s going on. There weren’t a lot of insights on the call. There was a lot of discussion about the Win 10 to Win 11 transition, October 2025. They’re optimistic about AI PCs, but can’t figure out the market share… Sorry, the numbers compared to HP and Lenovo.
Daniel Newman: Yeah, you covered it well. First of all, side note, but probably worthwhile. I think Dell got a real lift this quarter and will more so into the future as some of these question marks rise around Supermicro. We saw Hindenburg publish this short seller report. They were a bit of a Wall Street darling, and then they saw almost a 50% decline off their peaks. I think as people started to piece together that there was a bit of a house of cards there. The value add is going to be so important and Dell is so much better equipped on the value add on these systems.
Patrick Moorhead: Oh my gosh, Jeff talked about… Use terms I’d never heard about, but yeah, wanted to-
Daniel Newman: And I said this when I was talking to some of the folks when we were getting briefed on everything and I said, “Look, I don’t know that the market fully understands all the important contributions that Dell does to make AI optimized.” And like I said, it’s not really servers, it’s racks. It’s build out and it’s really being able to help companies understand the full requirements to integrate. And this goes back to what some of the alarm bells were with the AMD acquisition with ZT is that who owns this system design build out? And I think it is a bit of a village, and especially by the way, I know AMD said they want to compete with Dell… Or sorry, they want to compete with NVIDIA. But on the Dell side, they want to really be able to drive margin, which has been where all the concern has been around the fact, great revenue growth. But it was at a lesser margin than in the past. And that has been one of the push and pulls is where do they keep adding value so they can stack more margin into every one of these deals? And I think that’s what Jeff really talked quite a bit about. They’re really well positioned, Pat, on these tier twos, really, really well positioned.
Supermicro, this isn’t going to be like world ending, but I think there’s going to be some hesitation that Dell could capitalize on immediately as people are wondering what they’re up to over there. I’ll spend a minute longer on CSG just because you really covered ISG well. Pat, look, I know that we, and at times, we were talking back in the beginning of the year at CES about this device-driven super cycle. And I want to just basically say that you said earlier, let’s take that word off the table. I still believe in it. I do think we’re seeing a bit of a push that’s going on right now in terms of the biggest demand. I think I expected a little more Q3, Q4 demand. I think we’re seeing a bit more wait and see. I actually said I think Apple is going to surprisingly drive a wave of Copilot+ and AI PC growth because people are going to start to see this on-device experience and enterprises are still more commercially attuned to Windows based, but it’s just showing up in the numbers. CSG’s numbers were even softer than some of the others.
But Lenovo, HP, we’ll talk about here in a few, none of them had particularly optimistic. We talked to Enrique about it a little bit and we’ll talk again on HP, but I think everybody is starting to admit that there’s a lot of excitement, there’s a lot of enthusiasm, but it may be more Q1, 2 of ’25 and even beyond before this goes to scale. And some of this too, Pat, with the best features like recall really only being reintroduced here in Q4, there’s still a little work to be done by the ecosystem. And Pat, you and I are going to try to be educators to the market on what people can get from a Copilot+ and in AI PC. But I think that’s part of these numbers being a little soft is, but it’s also a reason for optimism that some of the best is still to come. This buy cycle could be huge, but it hasn’t happened quite as fast as I had initially anticipated.
Patrick Moorhead: Yeah. And just to clarify, I didn’t say that I don’t believe in the super cycle. I actually do. It’s just the definition of it. And when I was on stage in January at CES and Intel stage calling this, I said the second half would be the beginning of it. And I do feel like recall not going, and that was the sexiest feature out there. It’s having an unlimited memory of everything and it didn’t get delivered. And I think that’s an issue. Interestingly enough, Lunar Lake could be a real shot out there for volumes. And there’s this rumored eight core Qualcomm that Cristiano, Qualcomm’s CEO alluded to. The expansion of price points will absolutely make a difference. I think Enrique said 2025, 50%… Sorry, ’26 it would be 50%. But again, the reason that number sounds so low, first off, it’s a premium offering and it’s a little muddy what the desktop market is going to be, which is almost half of the entire PC market. And that’s directly related to the Intel and AMD roadmaps. And are they going to put big NPUs in their desktop parts? Right now, the answer is no.
Daniel Newman: Yeah. And by the way, there’s a massive opportunity for each of the silicon providers to differentiate, and each of the OEMs to differentiate in this category. Of course, there’s some benefit to some to see more standardization. There’s benefits for them to actually individualize these products a little bit, Pat. And I think we also heard something like a 7% to 9% increase in ASP tied to, I’ve heard different numbers, but that’s a pretty good range of what I’ve heard. So, you’ll see revenue grow just because these devices are more expensive. And so, as it goes to volume, you’re going to see on a per device basis, more revenue should be more profitable too as we see this turn.
Patrick Moorhead: Hey, I want to end this one. I want to read a quote from Jeff Clarke on the call. And just to be factual here, their AI margins were higher this quarter than they were in previous, that they got raked over the coals. And he says, “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 we’re bringing to our customers and the expansion beyond the specific node to the rack level deployment. So, our ability to add L11, L12 capabilities,” first time I heard those terms, “which are expert deployment, system validation and testing, the ability to help engineer to 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.”
We have talked about all of this on our previous shows, but the biggest reason that training runs do not complete is A, networking breaking down. And secondly, it comes down to just to be brutally honest, GPUs flaming out. So, there we have it. Great segment. I hope everybody out there give us your feedback. We’ve got two more topics. Dan, what happened to Pure Storage? Are they seeing the AI love? What’s going on?
Daniel Newman: Yeah. First of all, you did have a great segment there, so congratulations. As we head in, I don’t think this is just an AI play here with Pure Storage. This one was interesting. It was a narrow beat on the bottom and a narrow miss on the top. So, it was one of those really close company got battered in the wake of its announcement. And let me tell everybody why that happened because I think a lot of times people go, okay, you barely missed. But they pulled back on their TCV estimates as it relates to their SaaS growth. So, they had about a $600 million annual SaaS estimate. They pulled that number back to $500 million. So, I think that was where a lot of the reticent about the name came. They’ve been making this pivot to subscription, to ARR. And so, showing a bit of a conservatism there I think was met with a bit of a hesitation. Look, I put out my note, a really brief note, and I said, here’s what I really like about the company. One is very focused on power. So, right now as the big what’s at stake for this AI build out is going to be power consumption.
Pure is showing meaningful efficiencies gain versus other tier storage. And that’s a huge thing to be watching for against other modalities. And the other thing I said, and I’ll lean into this and I feel like I have to lean into every quarter, is this is a company that’s obsessed with happy customers. And the highest NPS score in the industry has remained that way for years on end. And so, they’re executing by being really customer-centric, having very low attrition rates, high growth rates, which serves them well in this pivot that they’re making from where they are as a storage company to a full on SaaS ARR company across their Evergreen portfolio. They’re adding customers at scale, they’re growing double digits. Pat, it’s a little slower at 11%, but they did see 24% growth of their subscription business.
So, although they did pull back a bit on TCV on cloud, they are still growing mid double digits, solid backlog. The company is in a good position overall. Pat, I think this company wants to make attachments to AI and I think they’re doing some other things pretty interestingly well in terms of different types of storage that can be leveraged more efficiently for scale up, scale out and direct to flash. They play in that sort of cloud native space. People don’t talk about what they’re doing there, but the Portworx acquisition was material. And then, of course they play in the SaaS space. So, you put all these things together, they address, block, file object, and they’re able to basically help customers reduce power. I don’t know, I think the number is like two to five.
They have a TCO. I’d love to do this in our lab because that’s what I like to do, but they’re showing 50% total cost of ownership reductions versus typical configurations. Now, again, that’s why I’d like to do in our lab because I want to know exactly what that is, but lower cost to operate, higher reliability, Pat, that’s the things that they focus on. And I like that they’re so outcome driven. But the bottom line is this, they’ve told a SaaS story, they slowed their SaaS estimates, market reacted to it that way, but I still think there’s a lot of optimism that should be around this company. It’s well positioned. And of course, expanding too even into areas like cyber resiliency.
Patrick Moorhead: Yeah. So, let me add onto that. So, first of all, talking about the TCV target, this is all about Evergreen//One and subscriptions. They did reduce it by about $100 million from $600 to $500. But there was a lot of discussion that these deals just take longer. They’re big, they take longer to close, but they have not shifted to CapEx purchases. On the Hyperscaler piece, I’d love to know which ones they are. Unfortunately, we didn’t get to talk to Charlie this quarter, but when we attended their Big Tent event in Vegas a while back, he actually talked through that opportunity and floated that on the calls. It’s interesting. I’m going to have to drill down. Hey, is a hyperscaler at the top five or are they tier two CSPs? This XAI deal, I believe that they were very much engaged with that, with this split deal between Dell and Supermicro. The company has the right strategy. Whether it’s on-prem, whether it’s in a colo or if you want to have, I like to call them basically, hybrid cloud connectors. You can do that and you can have one management plane across all of your storage. That’s just the right strategy. That’s the future. And if anybody doesn’t have that, you’re not going to be successful in storage.
On the competitive front, there were some questions about QLC and NAND pricing. They were one of the first in with QLC. This direct flash is definitely a differentiator in my opinion, in some of these price-sensitive workloads. So, hopefully, we’ll get to talk to Charlie next quarter, but it was hopefully you appreciated that analysis. So, hey, let’s jump into HP Inc. I’m going to call my own number on that, but I’ve got to open up my notes a bit here. So, mixed results, really appreciated time you and I both had with CEO, Enrique Lores. I’m going to do a highlights, lowlights and what they talk about the future. So, here are some of the highlights. First time back to revenue growth in nine quarters, that is a big deal and it was on the back of PC growth. It was up 5% year-on-year, and 6% on commercial units, 8% on revenue. What’s interesting with the Dell decline, most of Dell’s business by the way, commercial was flat at Dell. And here, we have HP who did really, really well. Consumer printing was up 2%. Print supplies is expected. EPS within guide, and this is not really a surprise, but they do have a fab for some of their printer chips.
They got 50 million bucks through the Chips Act, surprise. Here are some of the lowlights, a lot of highlights, but there’s some lowlights. Printer margins and revenue, it looked like tough China results, aggressive price environment there. Consumer PC declines, we saw Dell at 22%. They were off 6%. Their revenue, interestingly enough, was only off by 1%. And I can attribute that. I don’t have all the data for it, but could that be the AI PC sell-in? Possibly. Commercial printing was the toughest there at off 5%. And EPS, they hit expectations. EPS was well within the guide. It was below the EPS expectations. There was potential that investors thought that printing would do better than it did, but hey, stocks are about the future, not about how they did. They do expect fourth quarter print improvements with cost cuts, even revenue strength. They do expect meaningful AI PC sales growth in the back half, just like what I said. Over a 10% revenue mix in the second half and 50% mix in 2027, a lot faster growth in consumer. Again, that 50% in 2027, let’s not forget that that’s a split between desktop and notebook.
And we don’t really know precisely what AMD and Intel are going to do in desktop and AI. And right now, the definition of an AI PC doesn’t really factor in any, let’s say NVIDIA GPU. Like Dell, they’re expecting to see commercial demand swell with the Win 10 to Win 11 adoption based off Microsoft cutting off non-free support in October of 2025. You could still spend 60 bucks per head if you want to milk that, but I don’t see enterprises doing that because they had a big swell during the pandemic and enterprises buying all new PCs. Those were Win 10 and those are getting old. Done.
Daniel Newman: Yeah, I think you covered it. I don’t think there’s really much I have to add there. I put out a tweet, you can check it out, on this topic. I fortunately have to run to a meeting here at the top of the hour, Pat. I will say that my comments in the Dell about CSG, about the overall cycle is probably my strongest reflection of the opportunity. I think these mixed silicon offerings, these mixed platform differentiation, the commercial approach differentiation, that’s going to be focal point going forward, getting a commercial to buy-in in a big way. Pat, my heart and head really says mid-’25. I know this isn’t going to be the happiest news, but again, having been out and about, done some shopping myself, started to use these things real, I’m super excited about it. I’m super excited about it. I really think we’re going to see this thing ease in. And I like to say, Pat, slow at first, then all at once, and I think it’s going to be that way. It’s really, really a good time, Pat. And by the way, two items we didn’t really get to this week, but I said earlier, Apple and the glow up event is going to drive all device super cycles. I really do say that. I stick with that.
Apple just is a bellwether. As much as you and I like to pick on them, they will make this stuff digestible and people will see the power of it there. And Pat, we have to get on, maybe even do a special, I don’t know, we’ll figure it out. But this Intel rumor swirling about Intel splitting, dude, you hear activist investors, now you’re hearing splitting, that has me really worried about what this next quarter looks like. Tough times ahead. But you know what, even more important, they’ve got to keep their heads down and execute Gaudi. They got to execute on Lunar and Arrow, and then they’ve got to get these wins in an IFS. Those are the three big boxes they better tick. This is going to get wild though.
Patrick Moorhead: Yeah. IFS split right now is nonsense to me. But we can talk about that-
Daniel Newman: I’m just saying the rumor is really interesting, bringing in bankers and I don’t know how true this is, but usually there’s a little fire when there’s smoke.
Patrick Moorhead: Yeah. My guess is that’s to counter off a Lip-Bu Tan coup d’etat that I’m expecting here. Because what you want to do is you want to be ready. You’ve got a coup d’etat potentially led by Lip-Bu Tan. Coming in, you want to be ready to be able to explain to the investors why you’re on the path, do the right thing. And oh, by the way, with all these I-Bankers, do tell us this isn’t the right time or this would be the right time. We’re doing the topic right now, but Dan-
Daniel Newman: Yeah. I didn’t mean to totally dive in, but just saw this stuff breaking as we were talking. And the news is snowballing by Monday, who knows? But Pat, buddy, great seeing you, man. Thanks for taking a vacation and working.
Patrick Moorhead: No, I know. It’s more of remote work. I’ve taken a few hours here and there. Yeah, my vacation, we went to VMware Explore. Thanks for visiting me here. I appreciate that. And everybody, hey, thanks so much for tuning in. Always let us know how we can make this show better. We really appreciate you. This show is live. There are no parachutes. You can see us going back and forth, and clicking up the wrong Chiron, and all that stuff. But hope you appreciate that and we make mistakes, but we hope that you appreciate live. We typically have about 1,000 to 2,000 people who are watching this live. There’s been some Twitter changes in how it’s viewed, and you don’t even see the image when it’s live. So, that has reduced some of our real-time viewership. But hey, we appreciate you. Take care. Have a great weekend.
Author Information
Daniel is the CEO of The Futurum Group. Living his life at the intersection of people and technology, Daniel works with the world’s largest technology brands exploring Digital Transformation and how it is influencing the enterprise.
From the leading edge of AI to global technology policy, Daniel makes the connections between business, people and tech that are required for companies to benefit most from their technology investments. Daniel is a top 5 globally ranked industry analyst and his ideas are regularly cited or shared in television appearances by CNBC, Bloomberg, Wall Street Journal and hundreds of other sites around the world.
A 7x Best-Selling Author including his most recent book “Human/Machine.” Daniel is also a Forbes and MarketWatch (Dow Jones) contributor.
An MBA and Former Graduate Adjunct Faculty, Daniel is an Austin Texas transplant after 40 years in Chicago. His speaking takes him around the world each year as he shares his vision of the role technology will play in our future.
