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 Does It Again
- Amazon Shames Apple With New AI
- Cisco & NVIDIA Hug Harder
- Qualcomm Doubles Down On Industrial With New Brand
- Pure Storage Q4FY25 Earnings
- HP Q1FY25 Earnings
- Synopsys Q1FY25 Earnings
- Dell Q4FY25 Earnings
- Salesforce Q4FY25 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: The Six Five Weekly Show is back. Surprise! Thursday evening, not Friday morning, but hey, better early than late, and never. Daniel is hopping on a plane to Spain early. I am not too far behind him. On Saturday, Daniel, my day started at 5:00 A.M. grinding the entire day. How about you?
Daniel Newman: Man, I am absolutely spent right now. Big day, long day, pretty good day. I mean, NVIDIA drained you, and me, we were both running. We’ll talk about that, but that was a lot of energy. I had a pretty interesting week. I did some things I had never done before. Went to New York, went to Washington, D.C, came home, slept in my bed one night. Now I’m going to get on a plane. By the way, I’m not going to Spain. I’m going to Miami where I can sit for six hours then go to Spain, because you can’t get anywhere easily, or directly out of Austin anymore. And also, by the way, I hate going to London, just a sidebar. Hate going to London, but all good buddy, man, I’m just glad we got it in. This is how much we love everybody out there, right? Is that instead of canceling it, there was just too much important stuff to talk about, and we’re like, “We love our people.”
Patrick Moorhead: We do love our people. And it’s not the Six Five, it’s the Six 27. We’ve got a lot of topics, a lot of big news came out, but a lot of earnings. We’re going to hit NVIDIA Day. We’re going to talk about Amazon, brought out a new Alexa plus Cisco, and NVIDIA did another group hug, a deeper hug, a more hearty hug. We’re going to talk through that. I don’t think we’ve ever talked about a new brand, but I think it’s important, Qualcomm brought a new brand out for the industrial markets, and we had earnings from Pure Storage, HP, Synopsys, Dell, Salesforce. Daniel, let’s jump in. You, and I were both cooking broadcasts left, and right. I just got off Bloomberg, and we were talking Dell, but guess what we talked about? We talked about NVIDIA. I was on CNBC for 30 minutes talking about it, and I think you were on 16 shows in New York, or something.
Daniel Newman: We don’t need to keep score, we don’t need to compare. So, you, and I could spend the full hour on just NVIDIA, and by the way, everybody out there can read the print. It was really good. I was really proud though, to tweet five minutes before that I’m like, “I just have this feeling that NVIDIA could totally deliver, and they will still go down”, and I’m so tired. It gets exhausting to be right, but it is absolutely what happened. There was just no guide that could have been big enough. I mean you’re talking about, and by the way, price target raises out the galore. All the analysts came back. They’re like, they did it, $11 billion of Blackwell in its first quarter with very minimal margin disruption that are going to come back. They beat the earnings number, they beat the guide number, didn’t beat the guide number enough. I think everybody wants some astronomical forward guide, but I think this is the pain of trying to support a three, and a half trillion dollars valuation. Just so you know when we say three, and a half trillion, 3,500 billion valuation. When I see the number that way, it’s even more striking to me.
But Pat, here’s the thing. The biggest, and most important takeaway for me of this whole NVIDIA thing is that all the FUD, all the nonsense, all the competition, all the stuff that’s supposedly happening, it’s not happening. It’s just FUD, FUD, FUD. It is all FUD. DeepSeek? Yeah, interesting stuff. FUD. And I’m actually shocked, and I just want to say this to everybody out there, I’m shocked at the stupidity of the people out there that are still bringing DeepSeek up as some sort of red herring that AI demand is going down. It actually proved the opposite, and all the smart people said that straight away. It’s not a demand destroyer, it’s a demand for… Anything that creates more efficient, and scaled AI is going to be a demand creator for infrastructure. You’ve got $500 billion plus dollars just for Stargate, not to mention the 350 billion from CapEx, from..Everything is in the favor of NVIDIA, Pat, I’m going to kick this one over to you rather than just talking all the way through it, and then maybe you kick it back to me. The only thing I would say is very quietly because data center is the only thing I was talking about. Very quietly, all of a sudden, though, they’ve taken interest in automotive, and it’s not a huge business for them, but automotive is up like triple digits. Those are just another interesting little side note amidst all the Blackwell, and data center talk.
Patrick Moorhead: Yeah, so before you, and I go on the shows, they ask for notes. Anyways, I said it was going to be a triple beat, and what I wanted to look for was more Blackwell rollout details, supply chain clarity, and a very direct DeepSeek statements, and also color on gross margins. I popped on closing bell on CNBC, and I got the gross margin question. Again, I didn’t know what the company was going to say, but I said, “Listen”, John Ford asked me, “Hey, what’s typical margins right? Is 70% good margins for semiconductor?” I said, “That’s industry leading.” I think average is around 51%, and anything around there is doing pretty awesome. And then I speculated that this is likely some pullaheads supply chain to get everything ready for Blackwell. I was happy to see that that was the explanation that NVIDIA gave. And to me, this is not me looking at a spreadsheet. This is just being in the semiconductor market, and having to do these things.
I think in the end of the day, this is a stock that people are looking for reasons to sell, and reasons to get out, because of its value, because of just how big it is. I don’t think the company was as clear as it could have been on tariffs. I know that ended up being an issue that came up. But then out of the other side of my mouth, I’m going to say my gosh, in the on again, off again Trump Administration, even if somebody says they’re going to put a tariff on it, what is the tariff really? So, yeah, I think the situation we’re in now, Daniel, and then I’ll punt this back over to you, is I think people are pretty much convinced the calendar ’25 what it’s going to look like for NVIDIA. I think people are wondering about calendar ’26, and how far this goes. I am still surprised. I watched a couple shows today, and DeepSeek is being discussed. Jensen pretty much said it’s a 100X requirement to do all of AI like a reasoning engine. On Bloomberg I use the conservative 20 to 30X, but still, I can’t believe we’re still discussing this, and 4.5 just popped out today from OpenAI, and it’s the most expensive model ever. Not just to pre-train, or train, but to do inference on it.
Daniel Newman: Yeah, I mean, look, we’re orders of magnitude more compute is required for reasoning. These reasoning engines, by the way, are still take a lot of time, and are slow. So, you actually prompt them, and they say, come back later, get your answer. And you think about that. That’s not the ultimate experience. I mean the work’s good. It’s a lot better than an analyst, Pat, that you could have do the work. It would take days, or weeks to pull together maybe some of this stuff, but it’s not the level we’ve come to expect with a search engine. And I mean ultimately that’s where we have to, and guess what that means? A whole lot more compute is going to be required to get us there. In fact the scaling laws are still very much in effect. They’re still going to find more efficiency. The demand curve is still very early on. We’re barely seeing consumption at any meaningful level.
And that’s why I also, Pat, can go on air, and say, and I’m not going to name the show as I did so many of them, but the point is that there is so much flex upside coming here for, and there’s so much upside for NVIDIA. NVIDIA can win, and others can win too. I thought it was interesting that Jensen called out that they’re basically multi cycles ahead of the XPUs. He was sort of challenging the whole notion of the homegrown chip, whether it can be disruptive, whether it can be competitive. And it was kind of interesting because at the same time I saw a lot of charts that showed Hopper demand went off the charts after the DeepSeek moment, meaning that a lot of people looked at that, and said, “Oh, my gosh, we can do more. We don’t necessarily have to go to Blackwell.” But either way, again, NVIDIA wins. The only way they wouldn’t win is if one of these other architectures is meaningly able to redirect revenue away from NVIDIA, and it’s just not happening. So, yeah, I know, I saw that. You’re canceled. Get off my show. I don’t even care. But in the end, Pat, the market…
Patrick Moorhead: Should I ban him?
Daniel Newman: Ban. No, I mean the markets are basically looking for zero-sum. They like the DeepSeek story, the TD Cowen story. They like the XPU competition story because they Ultimately want a winner, and a loser. They don’t want there to be anything in the middle. And the thing is, NVIDIA can be a massive winner. And when we talk about their Cisco partnership, you can talk about how Spectrum X can be a winner, and so can InfiniBand. It does not always have to be one thing. And the market’s sort of missing that in a really substantial way at times is that NVIDIA will keep winning, but others will, too. This will get more democratized. There’s a whole lot of other economic supply chain China related issues that probably we can’t even get into here, Pat, because then it’s going to be the 975.
But Pat, I don’t give a darn what the market did in response to this. Every quarter that the market keeps selling NVIDIA on better results, it means it’s a cheaper stock to own. It actually has gotten the point where it’s one of the cheapest on a forward PE basis of all the Mag-Seven companies, all they do is print cash. And if I’m Jensen, I’m chilling. I’m not worried about it. Just keep doing what you’re doing, because you’re revolutionizing an industry, you’re making a ton of money. Yeah, you’re going to piss some people off. There’s going to be some people who’ll try to disintermediate the supply chain, and if they can do it, hats off to them, but right now they’re getting it done.
Patrick Moorhead: Yeah, don’t do anything unnaturally stupid. Remember that conversation we had earlier today? You’re crushing it. Don’t do anything unnatural that gets you off track here. Hey, I can’t believe it only took us 10 minutes.
Daniel Newman: Well, we could have gone a long time.
Patrick Moorhead: We’re going until 10:00 P.M. right?
Daniel Newman: I just felt like we’ve eight more topics, dude. We’ve got
Patrick Moorhead: No, let’s go in here. So, Amazon came out with its new Alexa, Alexa Plus a Generative AI infused system to be used. Essentially, you don’t need new devices every single one of your Alexa devices, I know Panos left himself a little bit, but pretty much all of them can use this new capability. I got the chance to watch this in action before the event, and I have to tell you, it works. It worked well. And what I really liked is the attitude that the company took where it just came out there, and said, “Hey, here’s what current Alexa is not good at.” They didn’t denigrate any of the other generative device solutions like Siri, and from Google, but what they did is they went out there, and I thought they were very self-effacing, and I really like that. That makes me trust somebody a lot better. And if you remember, Amazon has been dogged a little bit from a, “Where are you guys, everybody’s coming out.” Now, they may not be first with Alexa Plus, but they could be the best, and a lot of different examples were given. But it reminds me of a lot of the Generative AI voice magic that you can do today with OpenAI.
But think about Ring doorbells. Think about all of the smart speakers. I have about 20 of these $20 devices that you plug in a wall socket to be able to ask questions, and get results. And this will actually do things. They’re agents that can manage schedules, you can order groceries. And one thing I really liked that Operator failed at on OpenAI was it actually remembers. You can remember that you only want grass-fed beef. You can remember that you only want organic fruit. It has a memory, and I don’t know why Operator doesn’t have that with OpenAI, but it doesn’t. And these are full agentic, full multi-step tasks. And listen, I’m not Babe Ruthing this at all, because we’ve seen a lot of, I like to say I got the postcard, and it looked great, right? You get to the resort, right? It’s falling apart, it doesn’t work. I want to get time on this to give a definitive, but I have to tell you this even shames what Apple is showing. Gosh, I don’t know, six months ago. And I think once again, not only does it show that okay, Amazon could be in a very good position, but to me it clearly delineates that Apple is not.
And as we know, from researching on the AWS side, it gets full access to bedrock models. It’s primarily anthropic for task automation, and a lot of its magic here. Definitely open, right? We’ll probably see Nova models here. That would be my suspicious suspicion at one point. Here’s the cool part in the pricing. Even though this stuff costs a lot, it is free for every prime membership, which I would guess that pretty much everybody who has an Amazon device, like a smart speaker, smart video, or a Ring camera system, I think I’ve got 27 cameras. You are going to get this for free. You can get it for $19.99. But I got to tell you, man, I don’t know who’s going to do that. Maybe somebody will do that. I think this is a draw into a Prime subscription.
Daniel Newman: So, very interesting, the whole thing, it’s been a decade of sort of stop start with the smart speaker industry. A lot of people bought Alexa, or bought these Alexa for their homes. I don’t know, Pat, I had several of them. I’ve got them all over the place. I’ve got Amazon TVs, I’ve had some of the touch screens I’ve demoed, and played with some of them. Frankly, it was always a bit of a, it was like a physically installed version of Siri. It was always a little disappointing. It was like, “Yes, it could occasionally play the song I wanted to play, and we get that right, but most of the things you would want to do with it just didn’t do that well.” This is that moment. This is that inflection where we are seeing what we’ve kind of had in a ChatGPT moment for a way we interact with the web for how we could possibly bring this to a bit more of an ambient type of experience in our lives.
The biggest, and by the way, I love what you said about the pricing. They’re going to make this really available to a huge install base of prime members. Where’s the rub? The rub to me is do people actually… Is there still an interest in another tool? And what do I mean by that? This is why I think Apple, despite its horribleness in the AI space is such an advantage is nobody seems to want to interact with anything else but this thing. This is the center of our universe, and when it’s not, it’s the laptop. My biggest thing is that there’s so much with Alexa is whether, or not we expand these ambient device usage into our lives more pervasively. It kind of came out early. Apple blew it with the HomePod. Google had something, Amazon did the best out of all of them, but I don’t think I have any of mine plugged in anymore.
So, does this change the sort of calculus of do I want to have another thing to interact with? If I’m booking a reservation, Pat, think about this. How you use your devices to book a reservation at The Well, you just kind of grab on, use your app on your phone. Are you going to walk up to a different device, and ask it to do that? So, I think the usability of it on the phone is going to be super important, and I think that’s going to be where the adoption might take place. I’m not sure if we ever get back to that sort of ambient, physical, additional smart devices becoming sort of the natural interface. And then the thing is, how many different systems do we have to get to know us? I’m just not sure.
But here’s the hats off part, and I’ll leave it here, because I’ve already rambled too long. The hats off part is they did a great job with the mixture of experts using the different tools, creating a very natural, very human feeling. And that’s what I would expect at this point. And to your point, Apple sucks at this. I don’t know how they suck so bad, but they’re just terrible at it. They actually continue to fail, and open doors. So, I think it’s an interesting inflection. It looks really cool. Works pretty darn well. I just wonder about that sort of device physical utilization thing, if people will ever move, if there’s any opportunity to kind of reinstate that as a volume opportunity, or if our phone slash something like that is going to always be the center of our universe.
Patrick Moorhead: You know what? I’m going to go back, and get the data about how many of these devices are still active. These cameras are active for sure, a ton of them, and so are a lot of these speakers. But I want to go get a hard part of those.
Daniel Newman: And by the way, I’m just wondering are people using them, and how that multi-turn well-understood interaction I mean, look, I think it’s awesome. I don’t know, I just don’t know. Who said that? Who said the world’s cruel to you?
Patrick Moorhead: Daniel did.
Daniel Newman: Oh, hey Daniel.
Patrick Moorhead: I didn’t know people from could come in there. That’s pretty cool. I’ve seen it come in on YouTube, but, does that bug you when I bring people?
Daniel Newman: No, it takes me a second to come back, but no, I’m good. I’m good.
Patrick Moorhead: Hey, let’s go on the next topic here. So, Cisco, and NVIDIA already had a partnership for AI. They one-upped it, where they announced that they were creating a unified architecture for AI ready data center networks. Integrate Silicon One technology with NVIDIA Spectrum X Ethernet platform. Dan, is this important? Is this just another press release? Is this significant? What do you think?
Daniel Newman: I mean first of all, the timing was great.
Patrick Moorhead: I know. Timing was great.
Daniel Newman: Everyone’s paying attention. I think right now this is kind of an area where I think people kind of saw again ZeroSum was like, “Oh, NVIDIA is doing networking. This is bad for Cisco.” Well, maybe not. I mean, first of all, there’s a massive opportunity in ethernet. This is not a ZeroSum game. It’s not all going to be InfiniBand. And then of course a lot of people I think don’t realize Cisco has a Silicon business. And so that’s another big opportunity here is that actually Cisco is integrating its Silicon into the Spectrum X platform. I thought that was pretty good. I mean, this is true kind of pick, and shovels. I mean all these GPUs have to network. All these systems have to talk to each other. And so the idea that they’re going to have this, they’re going to do reference designs, they’re going to work cross portfolio, they’re going to be focused on open, and interoperable standards.
And Pat, I think a lot of this is enterprise AI centric. It’s not so much focused on large hyperscalers. This is about the fact that the diffusion of this technology outside of the hyperscalers, beyond the tier ones, and tier twos to having enterprise data centers that are going to be stood up just like private cloud data centers now. There’s got to be technology, and there’s got to be the economy of scale around this. And so I think it solves a lot of problems, Pat, and it’s a nice win for Cisco, and frankly, NVIDIA needs distribution outside of the big handful, too. So, Cisco can help bring more revenue to NVIDIA in some of these enterprise environments where right now it’s still pretty nascent.
Patrick Moorhead: Yeah, I’m pretty sure this makes Cisco the first third-party ethernet Silicon provider with Silicon One. I think switches are coming out in mid 2025. Plain, and simple, Cisco is the enterprise leader in this type of network, and you have NVIDIA with kind of a three tier. You’ve got Mesh with NVLink, you’ve got Backend, you’ve got Frontend. The Backend end could be Ethernet, typically it’s not Ethernet. Then you’ve got the Frontend which is Ethernet, and Cisco is pretty much the company of Ethernet, and plopping this into an environment where whether it’s a tier two CSP, like a CoreWeave, whether it’s a comm service provider, or whether it’s a very sophisticated enterprise, they’re likely banking on Cisco. So, I think this is a win for both companies. I think this increases adoption of AI in non-CSPs, and I think that this is a win for NVIDIA as well. So, I am interested, I think you, and I are going to meet with Chuck at Mobile World Congress, and I hope we get even more skinny on this.
All right, let’s dive in here. Qualcomm brings out a new brand called Dragonwing, like why on earth are we talking about a brand on the show? Isn’t branding just naming? No, branding is not just naming, and I think Qualcomm has been pretty savvy with Snapdragon that’s really focused on consumer, and auto compute, XR, sound, gaming, but one of Qualcomm’s huge biggest strategies out there is diversification, and they have a decent size business in the industrial embedded IoT, even networking infrastructure, think FWA, and then obviously cellular infrastructure, and Qualcomm is driving a bunch of revenue out there, and now they’re making a commitment to a brand, which I think is super important.
A couple things this signal to me, so first of all, I think this is a signal of future commercial growth. This is a, as I said in the lead up, it’s strategic brand play, not just some tactical naming exercise. And given the investment in how they’ve treated Snapdragon, I think we can expect a similar treatment here I am trying to get under kind of the attributes they want this brand to stand for that are different from Snapdragon, and I’m hopefully I’ll nail that down at MWC. The funny part about this, if I look at competing brands in this space, it’s just like, it’s either brand lists like a model number, like the FSM 1000, or something like that. Or if it’s like this bag of brands, companies, it just brand proliferation, and you’re just like, “Why on earth are there this many brands? Who cares?” Well what happens is you get said product manager who wants a brand, and they’re like, “I want a brand. It makes me special.” Okay, I get it. I used to do products, I used to think that I was special, and I wanted my own brand.
Daniel Newman: I think you’re special.
Patrick Moorhead: Yeah, especially fat, and ugly now.
Daniel Newman: Boo.
Patrick Moorhead: I got my inbox.
Daniel Newman: You’re beautiful.
Patrick Moorhead: I’m putting on… I’m getting fat again, dude.
Daniel Newman: Congratulations. Welcome to life.
Patrick Moorhead: No, I appreciate that. But anyways, big commitment here. Hope to learn more about it at Mobile World Congress.
Daniel Newman: Yeah, it looks like that’s where it’s going to happen. I think this is all about that sort of 50/50 strategy that Qualcomm has been focused on. I think by the end of the decade they want to diversify half the revenue away from handsets. They’ve had this boondoggle in automotive. A lot of the rest of the business has been pretty, revenue-wise has handset centric. PCs, interesting space entering with the AI PCs, but they’ve had this really kind of boisterous business in IoT, and industrial applications, and it probably needs something to help it stand out. I mean when you’re so sort of known in a certain area. The interesting thing is could they have played off a Snapdragon with something in a way to stay consistent like they’ve done with auto, and handsets, and PCs? Maybe, but I at the same time think maybe they’re spreading their wings, and trying to define…
Patrick Moorhead: I saw what you did there.
Daniel Newman: Thank you. And define this as sort of a very different part of the business part. Functions, operates, executes, goes to market, and as a guy that has a few brands himself, I think sometimes people need that clarity, and sometimes they don’t. Qualcomm has been very brand-centric. Their leadership has been very focused on trying to build the brand. I think one of its historic risks has been, it’s been so important, but sometimes nameless in many of the places where it’s appeared, and I think they’re very focused on changing that. The only thing about industrial is I just don’t know if brands stand out the same way in that market. I get it, like Manchester United, I get it, being on the Mercedes car. If you want to be a choice when someone walks into Best Buy, and picks a PC, or picks a handset, I think it’ll be interesting to see how brand plays out when someone’s thinking about sensors for smart cities. But if they’re going to do it, they’re going to do it, and I think there’s a very good possibility it could pay off. So, we’ll see.
Patrick Moorhead: Yeah, it’s good examples. And this is not Qualcomm’s brand for what it’s doing in the data center, but look at Xeon, right? Look at Epic, look at some of these industrial brands out there. They seem to be pretty well known. And I know these aren’t device brands, but you remember the Sun Microsystems brand, and how fricking cool that was, or maybe you were still in high school, or something, maybe you don’t remember that, but it was one of those really cool industrial brands out there. Hey, we’re enter in our speed round for earnings here where we might not do the same five, 10 minutes per company, but we’re just going to give a down, and dirty highlights. Dan, let’s start with you on Pure Storage.
Daniel Newman: Yeah, so Pure had a good quarter. I’m pulling up a few notes here. Give me a second. I had the wrong one up, buddy. One second here. All right, here we go. 12% growth, 3.2 billion for the full year. $879 million double-digit growth year-on-year. One of the things I really liked about this quarter’s results was the subscription growth. It grew well. Could it grow more? Of course, but they basically hit the numbers, and guided in line a little bit below, but I felt like it was a conservative talk to Charlie Giancarlo, CEO. Look, the company’s been very focused on that pivot to subscription. It’s been very focused on having a story for AI. It’s been very focused on diversifying into cyber resilience, and I think they’re doing all the things, Pat, and so always been very centric to his business has been about great customer support success. It’s got a story evolving for moving to cloud, but at the time they got a story revolving for cloud moving to prem.
So, I really like kind of the way it’s set up. It’s diversified. We know that in some ways storage is sometimes seen as a commodity. So, one of the big challenges for them is to be continuously innovative, to continuously show that the way storage plays a role in the, like I said, in data, and AI, which has been very focused on, and of course at its size, and scale, it’s also been focused on trying to bring value to its shareholders, and it did that as well this quarter. It even announced a $250 million stock re-purchase program. So, in my conversation with Charlie, he really beat the drum on the company’s transformation though is that I don’t think they want to be just storage. And I think one of the big takeaways for the investors out there, they really do continue to deliver. They’ve got really happy customers, and they are evolving, and that gives them more SAM, and TAM to address. And so it was a good quarter. I think if they execute it has the opportunity for some great quarters ahead.
Patrick Moorhead: Yeah, I want to thank Charlie for the conversation as well. A couple of things, I’ll fill in the gaps. The guide wasn’t viewed as exceptionally good.
Daniel Newman: It was a little below, right?
Patrick Moorhead: Yeah, it was 11%, which was below expectation, and it’s very clear. I like when there’s a definitive answer there, which you think there is. I think it will bode a lot better in the long term, right? It’s not like we had some execution issues. This is all about the e-family growth. And don’t forget, e-family is designed to replace spinning disks, and you had Nan pricing that rose, you had disk costs that stayed the same, and you’re selling a lot higher percentage of the mix. This is exactly what you are going to get. And also when, and I read the transcript, looked at the Q&A, and Charlie was very definitive that this is a temporary situation, so we’ll have to come back, and see how temporary it actually was.
One thing I really appreciated was the discussion of fusion. Enterprises have a hard time pulling together data from in disparate places in the enterprise, and fusion pulls that all together, and I’m glad that Charlie went through, and explained a lot of that. Some more questions that came out slower than expected growth, and subs compared to prior quarters. I think we could be getting into the law of large numbers here. It used to be that when the company was, this is kind of one of these problems that AWS runs into where they’re growing, growing so big, and you get into just these massively large numbers. It is a challenge to hit the percentage growth that people are expecting.
So, let’s move to the next one, and that is HP. You, and I got the chance to talk with Enrique. The company pulled out a beat, meet, and meet. I was down 4% after hours. A couple highlights here, third straight quarter of revenue growth, strong commercial business in the PC division. And they even cited areas like AI PCs, and I was on Bloomberg a couple hours ago, and they’re like, “Hey, is this company just going to be focused on cost cutting?” And my response was, “Well, when you have a market that’s not doing great related to PCs, of course you’re going to go with the cost reduction lever.” And they have a program called Future Ready, and their target’s between 1.6, and 1.9 billion for 2025. So, this is an interesting company. If you can imagine they could take some of that AI magic that they picked up from Humane, built a on device, and hybrid arbitrator, and one that goes between printers, PCs, room devices, room video devices, and peripherals. You have something that is super, super compelling.
Daniel Newman: And no on-prem corporate Alexa?
Patrick Moorhead: Something like maybe possibly can arbitrate between all these devices, and be hybrid. I think that’s pretty interesting. You’re a product manager, dude, I want to hire you. I love product manager.
Daniel Newman: I’m just saying this is where I’m trying to, how all this stuff starts to get really super conflated. That’s all. I totally get you. You know what I mean? Yes, for five years I’ve been saying, don’t you want to just walk around the office, and have this pervasive, and connected experience? That’s what I mean. I hope someone’s going to build that, but it feels like there’s consumer things that do it, but they got to meet all the corporate sovereignty, and governance, and compliance things that makes it hard. But sorry, I didn’t mean to cut you off. I just got excited about Alexa.
Patrick Moorhead: Yeah, a final comment, and this is just an overall commercial markets. You’ve got Windows 10 that’s going out of support. Actually you can pay extra for a while to get extra support in October of ’25. And the thesis here is this is going to boost the enterprise purchasing cycle. For consumer PCs, Microsoft has to deliver an operating system that does a lot more AI magic than it does today. Still waiting on GA of recall. So, still waiting.
Daniel Newman: Dying on this whole PC space. It’s like so much potential. Well where’s that killer app that’s going to send units flying? By the way, the handset business is no different. Nothing’s quite sparked the imagination just yet.
Patrick Moorhead: Yeah, Apple’s getting hammered as it should be. I mean anyways…
Daniel Newman: On the HP front you got most of it. It’s an operate, and execute. Those are both tough markets right now. Print, and PC. They’re not super vibrant markets compared to data center GPUs right now. So, this is about managing the market, trying to gain more share, trying to manage silicon diversity right now, and getting the new SKUs in. Of course, winning in commercial seems to overwhelmingly, you’ll see that with HP, and Dell both are doing better there. The consumers just seems to be lagging, moving very slowly here to upgrade, and make new purchases. And so it’s a challenge for Enrique, and the team at HP, because they don’t have that boom business like Dell has AI server right now. Having said that, they seem to be managing it well, and the numbers are coming in about where they’ve been expected to be.
Patrick Moorhead: They did beat on revenue, missed on EPS, but yeah.
Daniel Newman: I use the analyst word, dude. I said about.
Patrick Moorhead: No, no, no, you’re right. You were-
Daniel Newman: I’m kidding. It wasn’t far. It wasn’t like a, “Oh, my God”, it’s just unfortunately right now a beat is a guy, sorry, what’s my joke? A beat is a meet. A meet is a miss, and a miss is that you’re dead. I mean that’s what’s going on right now in the marketplace. You just don’t want to miss.
Patrick Moorhead: It’s tough, man. Hey, let’s move to the next one, Synopsys. How they do, do they get the big AI behind them? China?
Daniel Newman: I mean they did well compared to their guide. So, they actually had a good quarter. They were actually down from a year ago on revenue, but they were above the midpoint of their guidance, and of course they beat both guide, and estimates on the EPS side. Had a little interaction, a little exchange with the CEO Sassine Ghazi, very optimistic. There’s so much activity related to AI there, and they’re getting a lot of tailwind on that. The going forward guide looks pretty good. This is an EDA company, but I guess for those out there that are not following like a Synopsys, this is so important to companies like NVIDIA right now that are building. They are a very critical, and of course they’re an IP partner, they’re the second largest. Everyone knows Arm. Nobody realizes that right behind Arm is synoptic.
So, I think overall solid start, the one thing I’ll say is the Ansys deal, it feels very at risk to me right now. This is a China complexity thing. China has very little reason to be helpful right now with everything that’s going on, and anything that makes our semi-industry stronger doesn’t really suit China to be cooperative. I just get the feeling if nothing else, they’re going to kind of sit on this thing a while. It’s starting to remind me of the NXP Qualcomm stuff that was going on. I don’t know if it gets done. They’ve had some good wins they’ve been able to get through Europe, and the UK, and some other places. I just don’t know what kind concessions going to be made here. And this isn’t really anything that Synopsys can do. It’s all what Trump’s doing, and what our overall posture is about China.
And you heard Jassy, a few others have come out today talking about how we have risks of falling behind on AI if we don’t change, and the curbs on the chip restrictions to China. There’s so much macroeconomic, and me is now a commentator on broader political, and macroeconomic issues. I also consider myself an expert on China. Joking. But in all seriousness, there’s a lot there. That’s probably the biggest thing, I think, if there’s some hesitation is there was a lot of expectation of this deal, and I would genuinely say, well, Synopsys has done all the right things. This one might end up being harder to get done than I originally thought.
Patrick Moorhead: Yeah, listen, the company did what it said it was going to do. It was way ahead of the China challenges back in 2024, probably one of the most pragmatic companies, and it’s reflected in their guide. The only thing I can’t explain after doing some research I need to check in with the Synopsys folks is on design IP segment decline. I don’t know if that’s a China thing, or not. Didn’t understand the answer on the conference call. They reaffirmed their guide, which has all of China baked in. But again, I think a meat is a miss here. The company did exactly what they said they were going to do. It’s funny.
I’ll take the glass is half full on Ansys, right? EU’s a go. US waiting period is over, meaning it’s not going to be challenged, and the CMA is working on it in the UK. So, those are the pluses on this. I hope this is not going to be another waiting game like NXP. I’d love for a company, I know it’s always easier for me to say something crazy like this when I don’t have the P&L, and I’m not the CEO, and on the board, but just pull out of China, cut them off. Completely cut them off on this if they won’t go through, or keep on any combined products they might come up with, they don’t get access to it. You don’t want the end to end chip to physical design system to make your cars, okay, go build your own, or keep using these disparate systems. I really want somebody to say that. I want somebody to pull out of a country.
Daniel Newman: I thought PoC might do that with VMware. Because I mean that actually could have worked, right? It was just not that big of a deal.
Patrick Moorhead: I think 10% of VMware usage was in there, but 1% of the paid licenses, or something.
Daniel Newman: Yeah, I mean if that one had gotten hung up, but that was also probably why it didn’t, because it just wasn’t, I don’t know.
Patrick Moorhead: Well, Figma, and Adobe, I would’ve loved for Adobe to say, okay, we’re pulling out. Sorry.
Daniel Newman: Yeah.
Patrick Moorhead: Okay. Final comment on Synopsys. This just bugs me. This whole notion. I heard it on some of the questions in the conference call. Why can’t you speed up Generative AI? Let me be very clear here. The technology we use today to do searches where, yeah, it hallucinates a little. On an EDA tool, if it hallucinates like 0.01%, and 0.01% of the layout, or the test is off, you’re pretty much screwed. Okay, let me just put that out there, and I urge Synopsys to come up with a certain set of analogies, or something to try to get that across. I know that could sound defensive, but literally you’re like, “Hey, why aren’t you using Generative AI to put the space shuttle on Mars yet?” Right? Why isn’t it? Well, because it’s really exceptionally hard, and you can’t make mistakes.
Daniel Newman: Yeah, it’s the whole STEMI supply chain. I call it the cartel. It’s got so many gates, routes, limits, and it’s very hard to work around. And accuracy is so important. I mean you can’t have a 98% working chip.
Patrick Moorhead: I know. It hallucinates a little, right? Oh, when we burned it in two nanometer silicon, your net dye per wafer is like two. Anyways, such nonsense. I don’t even know where to start.
Daniel Newman: Nerd.
Patrick Moorhead: Hey, let’s get into Dell. We are going to talk Dell, and I think this is my topic. So, first of all, you had a miss. This is interesting, a miss, and a beat, and then a meet, and a slight gross margin miss, and a meet on the non-gap EPS. Is that confusing enough for you? What’s interesting is I expected based on that for the stock to be down a little more, but here’s the deal. I think investors are recognizing that AI is super lumpy, particularly as it goes to ISG. I ISG grew by 22% on the back of AI, and it’s no secret that a lot of the Dell can claim some of the largest cloud service providers. Well, what do you mean? Well, XAI, right? 100,000 GPUs out there. One of the largest, I don’t know if it’s a, I guess it’s a cloud service provider. It delivers services of AI. So, yeah, take that.
And I also think the nine billion AI server backlog was as well, but NetNet the company, good revenue growth, and I think this is law of large numbers, 8% on $96 billion for the year. EPS growth, 39% in 2025, forecasted 19% in 2026. Company looks like they have AI momentum here. And I think XAI is a good example of that. They’re kind of caught in the slog of the AI PC conundrum here waiting for applications. I know they are working with ISVs on applications, so they’re not just waiting around, but they need a little help there. And like HP, they’re going to be able to take advantage of the Windows 10 to Windows 11 transition, which is starting, which is why the commercial PC market was up for them. But the consumer PC market is just absolute trash. By the way, capital allocation, they increased the dividend. And while I don’t think that, I did talk to some of the folks at Dell, and it’s very strategic, but not related to anything on future EPS to let’s say not juice the numbers, but I mean Apple does this all the time game. They come out with something, and they increase the dividend, or they do buybacks.
Daniel Newman: Financial engineering, buddy. It’s like you make your stock Teflon by making it really a lucrative sort of safe harbor for investors. Good thing about Dell is you kind of know they know how to operate. When the revenue’s not there, they find other ways to get it done. That’s kind of their thing. Their supply chains are really well managed. They know how to create cash flow. They still got kind of that. IBM does that too. It’s like the growth wasn’t there, but they still pump in cash flow. I mean they did book that..
Patrick Moorhead: Cash machine.
Daniel Newman: They booked that five billion dollar deal right after the quarter. So, their backlog got huge on AI. The dependence, and the ability to differentiate why they’re the AI server factory versus the now non-defunct Supermicro, or HPE, or Lenovo. The differentiation is going to be continuing to be hard because it seems like a lot of the values within video, and that’s a problem they all have to solve. That’s not unique to Dell, but Morgan Stanley came out with a note today that said 1.07 trillion in GPUs by 2028, Pat. It was 15 billion in ’23. Someone’s going to have to sell all those. Dell is going to probably do pretty well.
The net effect of that all is you’re going to have margin pressure, competitive pressure, but I actually think the ISG business is good. They’re differentiating around storage more, and more. They’re trying to networking business. But yeah, I don’t know what else to say, Pat, when it comes to the PC business. That’s why I said that about HP. It’s like that business sucks. I do not know. I will say when I’m wrong, and it’s not, like you said, self-inflicted, non-self-inflicted. Not all of it was our fault. I thought by ’25, I didn’t think necessarily in ’24, but I thought the full year a lot would happen. If you think about how much the advancements in data center cloud LLMs, and enterprise type of related AI took place, and how un-noteworthy that same has been when it’s come to the device side, it’s kind of shocking.
Patrick Moorhead: Yeah, it is shocking. And I just want to clarify. I don’t see anything wrong with doing buybacks, and dividend work to keep investors happy, except if you’ve got $150 billion in cash sitting, and you’re not doing anything with it.
Daniel Newman: I think my biggest beef is just that Apple’s been so non-innovated, but they’ve just bought their float down, and kept their price artificially high for so long that somehow they just get away with never innovating. It’s weird.
Patrick Moorhead: …look at us, look at us, the lemmings that we are.
Daniel Newman: I’ve got Apple everywhere. Like I said, I actually tweeted something. I’m like, “I don’t know how they do it. They don’t really innovate, but I keep buying the new phones, and I keep buying the AirPods”, it’s like maybe they don’t have to. Maybe we’re stupid. I mean, I don’t know. Maybe we’re smart enough to know we’re stupid, but we’re still too stupid to be smart.
Patrick Moorhead: Wow, this podcast is different this week.
Daniel Newman: Is it?
Patrick Moorhead: Isn’t it?
Daniel Newman: I don’t know.
Patrick Moorhead: Hey, let’s get to the next topic here. Salesforce did announced their Q4 ’25 earnings. Daniel, how did you..
Daniel Newman: …is this nine?
Patrick Moorhead: Yeah.
Daniel Newman: If they’re nine, geeze, good for us. Yeah. So, yeah, Salesforce. So, let’s just say this. The numbers did not make people happy. The revenue growth, the guide, Mark, I was actually cited on CNBC Money Movers today asking. I got asked by a CNBC, Seema Mody asked me, she said, “Is this Mark’s obsession with Microsoft?” Asked me in some way, or another if it was healthy. And I basically said, “Look, Mark, he’s a force of nature. His personality is he wants to slow down. People go to Microsoft, they never leave. So, anything to get people to pause is probably in his best interest. Do some shareholders probably not love the go negative strategy? I mean, yeah, of course some people don’t like that, but I don’t think Mark’s ever truly cared. I think Mark wants to build the best business he can, and he believes right now that that’s his fiduciary. Not making an excuse, just that’s my take on it.”
The good thing though is that they did break out some of the growth around Data Cloud, and the agent force. You’re talking about thousands of customers. You’re talking about real strength there. This is a situation now where Legacy Salesforce platform, and no one likes the word legacy. Let’s just say that the core user base of Salesforce, it’s all about conversion here, Pat. It’s all about moving them to Data Cloud, moving them to agent force, moving them their hyper force. They had announced a big partnership with Google this quarter, but their Data Cloud business was up to $900 million, was 120% year over year. I mean, so the point is, and I think they said they got 5,000 closed agent force deals, and 3000 contracts. So, revenue’s growing slower. But this is also part of that sort of infrastructure to service consumption gap that I opine on just constantly.
The fact that we’re building these cities of the future with data centers has not really converted to meaningful revenue on the software side. There’s been a lot of consulting contracts, and service contracts to do the build out. But if you look at Oracle, you look at SAP, you look at Salesforce, you look at a lot of the software companies, their ability to really articulate the AI impact on revenue has been somewhat difficult to do. Is it making them more sticky? Does it make the software more usable? But here is what’s going to happen is companies are going to move to more efficient agentic tools that are going to make them faster. They’re going to make them smarter, they’re going to give them better outputs, more productivity with less investment, but it is going to take time.
And so I think there’s a lot of good leading indicators in these numbers, but I think the problem has been is Salesforce was a company that grew mid, double, like 20, 25, 30% for a long time, and they’ve matured, and now they’re growing at high single, and 10%. And I think that starts to be are we a growth story, or are we a safety stock? What are we? And so this is a multi quarter journey. It’s a multi quarter pivot. I think they’re moving the right direction, but I think until they start having those big outsize beats, Pat, I think people are going to wonder, is this transformation happening, and is Salesforce going to be a leader at the end of this AI transformation?
Patrick Moorhead: Yeah, a couple things going on here. I think when the first agent force numbers came out, they were incredibly high. And I think the expectation was set that it was going to continue. And this quarter it didn’t. I was on CNBC, I was asked, it was interesting in the context of Snowflake, what was going on, and I talked Salesforce Data Cloud is always a precursor to AI goodness. Because you have to get your data house in order before you can get the best results, particularly if you’re trying to pull different types of data to activate.
So, it was the expectations, and I think that got it here. I think there was a little bit of large numbers going here, but I just think this stuff is tough. It’s tough to get maybe those initial sales were the trailblazers that will sign up, and everybody knows each other, and they’ve known each other for years. And implementation was just easier. And I think where we are now is Salesforce can find ways to make it simpler to get onboard. And I know it’s easy. It’s harder than it is, than I’m making it sound, but that’s exactly what they need to do there. I think Data Cloud was like $900 million.
Daniel Newman: That’s what I said. Yeah, 120% year over year.
Patrick Moorhead: Yeah, I mean, it’s crazy. So, Service Cloud, and Sales Cloud are 8%, I think you hit it. Is this the safe cash stock at 8%, or something different? I know Mark Benioff is not going to be happy being the safe stock. So, I’m really interested to see tactically where he is going to take this. Dan, we did it, buddy.
Daniel Newman: Oh, yeah. We got five more topics, everyone.
Patrick Moorhead: Yep. Hang on. Put on your helmets. No, just kidding, folks. I would appreciate everybody for tuning in here. We will not have our morning show. I know, how sad. But if you prefer watching it in the morning, just press the play button at 9:00 A.M. Central.
Daniel Newman: Didn’t change the topics we talked about, unless something really crazy breaks, right?
Patrick Moorhead: Exactly. But hey, Dan, and I are going to be at Mobile World Congress next week over in Spain. I don’t know if we’re going to do the Friday show. Maybe we will. I think we’re on the same plane coming back. We get back Thursday late. But I’m looking forward, The Six Five is going to be at Mobile World Congress. We have some great interviews set up with Dan, and myself, and a shout-out to Will Townsend. I want to thank everybody for tuning in. Make sure to hit that subscribe button, tell all your friends, your children, your dogs, your broker to tune in for the best show with the best moderator out there on the internet. 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.
