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:
- CES 2024 Ruminations
- Memory Market: Call it a Comeback?
- Arm’s “Blackhawk” New CPU Core
- HPE Buying Juniper
- Mobileye Reset
- GenAI Trends in 2024
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:
Daniel Newman: Hey everybody. Welcome back to another episode of The Six Five Podcast. This is the first episode of the, no, it’s not. It’s the second. Is it the first? No, this is the first episode.
Patrick Moorhead: Depends on how you count it.
Daniel Newman: This is the first of our normally on Friday, but not this week on Friday, episodes. This is 199, Pat. This is a moment. We should have a little moment of silence for 198 epic episodes, followed by 198 more. We take one down, we pass it around. Anyway.
Patrick Moorhead: That’s so good.
Daniel Newman: So good to be back, buddy. We just got off the docket. It was a weird year this year, Pat. It was like everybody worked right up until the Christmas holiday and then basically took off the week after New Year’s Day. So that week coming back was a little bit slow. But we were rip-roaring into this week. CES had us moving full tilt. Pat, we did our first ever Six Five live four hours, like 25 segments all over CES, 15, 16 vendors. We had startups. We had some of the biggest companies in the world. We talked AI PCs, robotics, automotive, gadgets. We even snuck in a segment about a, I don’t know, $14 billion acquisition that we’ll talk more about this week. But Pat, it’s been a big week, a big year. It’s a big moment for the Six Five, and I’m so happy to have my best, not so big, bestie back with me. He’s not as big as he was last year, people. Back in the seat, Patrick Moorhead, how you doing?
Patrick Moorhead: Doing great. I mean, this is my favorite time on Friday, but it’s also my favorite time on Thursday. And Dan, for the first time in 13 years, I took time off for the holidays. I worked just a little bit, but not a lot. And I’m just so happy to be back in the seat here, back in the office. For some strange reason, and I shared this with you, my stress levels are lower at work than they were over the holidays. And I need to think about that and figure that one out. But I’m glad to be back.
2023 was a rocking year. And 2024 just looks even better. I mean, the amount of changes in the industry, the winners and losers and the pecking order are going to shift likely. But as, I think, you said so eloquently on our last podcast that we recorded today, well, earlier, was that this is the year of AI implementation versus it being the year of the GPU. And I think that’s just spot on. And I think that’s going to continue throughout 2024. And by the way, if you wanted to look at what Daniel and I called for what 2024 is going to be, go check out previous episodes. We talk about what to expect for 2024.
Daniel Newman: Yeah, so 198, I think, was a predictions post. And then about 31% of my tweets thereafter or something like that, same thing. Of course, you and I did our runarounds, did our show, did TV, did contributions in people’s articles. It’s going to be a big year, everybody. So let’s get through the cliff notes of, what’s that, Pat, the disclaimers?
Patrick Moorhead: Sure.
Daniel Newman: This show here is for information and entertainment purposes only. While we’ll be talking about publicly traded companies, please do not take anything we say as investment advice. Pat, would you please stop making stock picks, buy ratings and setting price targets? Good gosh.
Patrick Moorhead: I definitely do not do that because I am not certified or authorized to do that by the SEC. But I hear some analysts do. I don’t know.
Daniel Newman: Well, some analysts should and others are them. So we got a great show today though. We’re going to talk about CES. We’re going to say our ruminations and all the things that we saw, what do we like, what we didn’t, what was hot? What was not? What happened at the show, what didn’t happen at the show, but that was relevant to the CES space. We’re going to talk memory. We’re going to talk about some big updates from Arm. We spent some time with both their former and current CEO actually at the event. HPE went ahead and bought Juniper. That was big news. Mobileye is on a reset after a bit of a crazy earnings ride. We’re going to finish off talking about something nobody else is talking about on the show. We’re going to talk about generative, gen AI, as most of you like to call it. So here we go, Pat. Here we are. Let’s rock. Let’s roll. Call in your number first. CES, what caught your attention?
Patrick Moorhead: So folks, this is going to be kind of a ruminations. We’re going to give you exactly, what did we see? So I think the biggest thing that I saw was this increase in interest for PCs like I haven’t seen in probably five or six years. And I want to separate that from the reality when the business was just hitting big time about three years ago during the shutdowns. But this is about something brand new and it’s about the AI PC and Intel and AMD. And for a little bit, NVIDIA gave effort at CES on this topic. Not a whole lot of stuff from Qualcomm because they had their event in Maui where they came out with a lot of the content. But that was the biggest source of industry discussion for me.
It’s funny, after the day, you walk by a bar in the Palazzo and you see people from Intel, AMD, Qualcomm, Lenovo, maybe some HP and Microsoft folks in there, maybe some Dell folks. But all in there, the talk was all about AI PC. I did a panel Intel, and I had the presidents of Dell, Lenovo and HP on there to discuss this. But macro, it was good to see the industry coming together to be able to invest in this. And the way to look at the AI PC is, don’t just start off from what you see now. Now is the start. You have core ultra. Even AMD had a release and has systems on the shelf that has a lot of AI capabilities. This is the beginning and not the end.
I can relatively, with high precision, talk about what this will look about now till 2025. But the TLDR is, six-month increments you’re going to see new chips, new capabilities, new software and new experiences hitting here, and I think probably crescendoing in the second half of 2025. But I do believe that we are going to see a resurgence in the PC market in likely the second half and fourth quarter, based on all the excitement about it that quite frankly started today. And you’re going to see a big surge in May and June. You’re going to see a big surge in October and September for the holiday selling season. So I would say, yeah, first and foremost, Dan, I’ll just pong it over to you. I have a few other topics we can talk about, but what were your thoughts?
Daniel Newman: Yeah, I’ve said this a lot now, Pat, but I feel like what we saw was this kind of and inflection. Last year, you quoted me so well, but the year of the GPU, it was the training year, it was the infrastructure investment and deployment. It was a ton of infrastructure on AI being bought by a small number of companies in order to deploy solutions that ultimately could be monetized. Of course, coolest device. I was on Liz Claman, and I said, “What is the coolest thing I saw?” I think it’s these LG or Samsung, I can’t remember which one’s making it, but it’s these transparent TVs. They’re going in the sensors, and the chips are right in the glass and they’re basically completely opaque, or clear. Which one’s clear? Not opaque. And you cannot see the sensors at all in the windows, but you can actually see the content everywhere. I just see so many applications for that in glass. And I think it was just cool. Remember when CES used to be about cool, just seeing things that were just different?
Patrick Moorhead: Well, there was part of that, as you talked about, right? Like see-through OLED, right?
Daniel Newman: Yeah, that’s what I’m talking about.
Patrick Moorhead: No, no, no. I know. That’s exactly what you were talking about. And even the Lenovo, I don’t know if you saw that, they’ve got a detachable where it turns into, you take the display off the PC and it converts from a Windows PC to a Chrome tablet, or sorry, an Android tablet. Just stuff like that that’s visually appealing.
Daniel Newman: Yeah, absolutely. So beyond that, I think, Pat, you and I came into this knowing what was going to be pretty hot in our space. There were some really cool sort of latency free personal assistant type devices powered by generative AI. And that stuff’s really cool. I mean, when you have an assistant, some of them roll around with you, some of them are more like theoretically in your pocket, everything from your personal trainer, a dietician, nutritionist. Intel showed a cool demonstration for people that are deaf and are hearing impaired to be able to get more communication through generative AI. And that stuff’s really great.
And of course, there’s going to be this debate about cloud or prem, what’s going to happen in the cloud, what’s going to happen on-prem? And when I say on-prem, in this case, it’s on the device, right? Because hearing a ton about this AI PC. And Pat, we knew AI PC was going to be a super hot topic. I’m going to put a little stress back on the OEMs and say, look, I have no doubt, and I agree with Pat, I agree with your notion on stage, you did the Intel event, you had all the chiefs of the biggest OEMs in the world, did a great job. You would’ve made Mr. Rogers proud with that presentation. But the communication of where value is going to be derived in this AI PC still has some work to do.
Patrick Moorhead: That’s why it’s the beginning, not the end.
Daniel Newman: It’s the beginning. And it’s the beginning because there are those that really do think that with the right silicon design, a lot can be done in the cloud with low latency. And of course, with these robot tools, nobody wants to wait. So if you can do zero latency or near zero latency, and I call it near zero is when you don’t see latency. So when you don’t feel latency, in my opinion, there’s no latency. It’s kind of like how high resolution can a monitor be before your eyes can no longer appreciate the higher resolution? There’s a little bit of that going on. But I do think the push to new experiences, new designs, new software, new applications that are going to be powered by not only an NPU, but just the whole chiplet and multi-semiconductor designs that incorporate camera technologies and vision technologies and GPU and rendering, et cetera. So there’s a ton of that going on, Pat.
And of course, I’ll just make a really quick pass. Automotive’s always cool, but I think it was just more software defined vehicles, more sort of stacking of moving from more blackbox to custom building blocks. But that’s what we’ve seen Qualcomm doing. Intel’s getting their hat back in the ring. NVIDIA’s, I think, been half paying attention to automotive. They kind of slowed their attention to it when the GPU took off. They didn’t really, the automotive people would be offended by that, but it felt that way. So anyways, I digress. Big topic. We could ramble on for quite a long time.
Patrick Moorhead: Listen, Dan, we’re not rushing this one because I’ve got a couple more topics I wanted to hit.
Daniel Newman: All right, let me kick it back to you because I kind of just gave you the whole rundown, and maybe see if you spark any more memories of things to come.
Patrick Moorhead: Yeah, so CES became the Detroit Auto Show about a decade ago, kind of coincidental, when NVIDIA started talking about what it could do inside the car. Actually, it was 12 years ago. And I remember it really well. And everybody’s like, “Well what is this? The auto show?” Well, this year was again about the auto show, an entire building dedicated to automotive. Funny, tractors, we saw John Deere was back again.
And what it did there, and a couple big statements there. So Intel is back in automotive after spinning out, doing a soft spin of Mobileye, which we’re going to talk about later. And essentially what they did is they committed to compute. And this theory that, by the way, is not new, that we agree with, that we’ve talked about, which is instead of having 25 ECUs that are incompatible that have to talk to each other and have a lot of duplication, centralize that compute. That’s Qualcomm’s strategy as well. And it just makes sense. So with all the new IP that Intel has, that’s a lot more competitive, adding that to the distributed architecture where they can slot in IP from, let’s say, a Ford or a Volvo into the chip and call it a day.
A lot of work and a lot of proof that Intel has to show because the view out there is that Intel got out of automotive and seeded it to Mobileye. Given the troubles that the company was having, it made sense overall. The funny part is that Intel actually kept building and selling products to automotive companies, but just really didn’t talk about it a lot, right? They seeded that conversation to Mobileye. If you want more information on what Intel’s doing, we published a white paper, you can find it in the show notes.
For Qualcomm. It was a reaffirmation that we are becoming the dominant player in this market. And while there wasn’t a lot of huge news, it was more about the momentum. And interestingly enough, I put out something on X, it gets 23,000 people to look at it. So 350 Qualcomm vehicles on the road, in it for 20 years, over 75 models out there. And for them, it was really a time to say, “Hey, we are big and we’re bad. We don’t have huge news to break, but don’t forget how big and bad we are.”
Daniel Newman: All right, are you done on CES or you want to hit a little bit more?
Patrick Moorhead: Maybe just one more, maybe just a final note on this is about XR. So even though Apple says that CES is insignificant, they always drop some of their biggest news. And one of the bigger news stories during that was that Apple Vision Pro will be available on February 2nd. It was literally the only thing that you could see and heard of on a Monday. In addition to that, Qualcomm brought out its new 4K per eye XR2 platform. This is a big deal, folks. 4K per eye is when it gets real, the experience. If you can do that at the right power budget, it gets quite compelling. We’ll talk about Rabbit next time, if we even have to.
Daniel Newman: Yeah, so I did think it was interesting. I did have a bit of a thematic VR is back, XR is back. And look, I said Apple is not always as innovative as people give credit it for. People always say, “They’re so innovative.” They were pretty late with XR. But I think what they do successfully accomplish that so many other companies don’t is they make markets.
Patrick Moorhead: They were 10 years late with smartphones. You’re right.
Daniel Newman: But when they have their smartphone, they timed the inflection point. And so if they’re right, as they’ve historically been, they’re timing the inflection of where we go from very small niche, sort of early adopter use cases, fringe use cases, to high volume. Their pricing doesn’t really enable that. But I think they’re also maybe creating this little bit of what I would call buffer zone for the app developers to create enough apps and usable things so that when they do drop a $1,500 headset at some point in the next two years, scale can come up with it.
All right, Pat, let’s talk about something. You and I had a couple of Six Five on the road at CES. We talked to Micron, but we also had some conversations with Samsung Semiconductors and some of the other players. And one of the things, Pat, that all of this AI, it doesn’t really matter if it’s AI on the PC or AI in the data center, is going to spur is a comeback for memory. And so we talk a lot about compute. And I think that’s kind of like in any stack, compute gets a lot of talk and then network and storage are kind like, yeah, we need some of that too.
And so sometimes it doesn’t get talked about. But here’s the thing, Pat, you and I have said this on the record quite a few times, nothing happens without memory. And in fact, whether it’s accelerated computing, whether it’s acceleration on a CPU, in order to do all this inferencing, being able to get enough in-memory performance, it’s going to be critical. And so you and I kind of brought this question up when we were talking to some of the folks at Micron, in fact some of the folks at some of the other PC makers is, have we come back and found ourselves in a moment where memory is going to, A, find the RAM? And this is going to be, of course, D-RAM and multilayered NAN technology, integrated memory is going to be very important. And are the companies in that particular space about to get some pricing power back?
And you haven’t been following the last 8, 10 quarters. Memory has been brutal. You got to look at the losses that have been endured by memory companies. I think Samsung had a quarter where their profits dropped like 90%, and it was tied to basically all their businesses at once sitting the trendline to go the wrong way. I mean, Micron went from making billions to losing billions. And it was really right in line with this kind of boom, bust cycle. But absolutely happened, and it was a very challenging moment for these companies. But right now, are they sitting there kind of doing their best Mr. Burns thing, kind of tapping their fingers and going, “Let’s see what happens without memory?”
Patrick Moorhead: So good, Dan.
Daniel Newman: And have we found this inflection, Pat, right now? If you’re a betting person, and we’re not talking about markets, not suggesting any stocks, but we are saying if you are a betting person, you are looking at markets and you’re kind of saying, what other companies involved in this AI boom maybe haven’t gotten a lot of attention? Is memory where it’s at? And I think the answer to this could be, yeah. What do you think?
Patrick Moorhead: So there are some things you can bet on, like your watch. And that’s that the memory and the storage market will go through booms and bust cycles, and nobody is satisfied with what is going on, ever. Because in the boom cycles, you can’t get enough. And my gosh, why didn’t these memory companies do good enough supply chain planning? When they’re investing $20 billion into something that they need during the down cycles, they need to pause, but they’re still paying on the CapEx for this investment in negative gross margins. So both SK Hynix and Micron had negative gross margins. Not profit, not operating profit, gross margin. And that is always disturbing. But why was Micron up significantly in 2023 with negative gross margins? Because people, thankfully understand history and that there’s boom and bust, and that we are going to get back in there.
What prompted my analysis of this, and check it out in the show notes was, reports coming out, rumors, about both Micron and Samsung increasing prices 15% to 20% in the first quarter of ’24. And I do believe that this is the beginning of the comeback. Now, SK Hynix is gross margin neutral, and Micron guided to that point. So everything to me says we’re about to see the comeback. But we’re going to start seeing the money printing presses are going to be hit probably in the fourth quarter. I mean, if we see a nominal jump in smartphones, nominal jump in PCs or the non-AI enterprise, you’re going to see prices go up, people can’t get memory, prices will go up even more, and then we’re going to start going through this cycle and it’ll probably be an 18-month to a 24-month super cycle in this. But if you want more information on it, I pulled financial information from Hynix, Samsung, and Micron, if you want to take a look for yourself.
Daniel Newman: Yeah, follow the brilliance of none other than Patrick H. Moorhead. Is it H?
Patrick Moorhead: I don’t know. We just need to turn on CNBC whenever, and you’re going to be on there talking about this stuff. So I got to make up for it somehow, my friend.
Daniel Newman: No, I mean look, the memory space has players that have been starved. They’ve absolutely been starved for this period of time. I mean, everything got pre-bought. Supply chains, they literally sold through everything they had. The panic that was created by that supply chain shortage created this lust for every bit of silicon and memory that would be available as if the boom would never end, as if the music would never stop. And then basically, what happened is one day it stopped, then it was like-
Patrick Moorhead: People are paying 25% what they were paying two years ago for the same density of memory.
Daniel Newman: I mean, unbelievable. So the moral of the story, by the way, people are paying 50% on cars, 50% of the price that they were paying. I mean, people were buying trucks for twice retail. That’s inflation, people. Just understand that this is where it happened. It was these exact things. And that’s also why some of the LTAs started taking place with some of these foundries. You saw some of them really picking up their long-term agreement investments, was like, we’re going to be the companies that start to steady the hand of the ship, which means there will be less boom, but there will be less bust because you guarantee capacity. And knowing that, it just is such the human condition. Because we know emphatically what’s going to happen. And it’s like you could have turned the music on. I would say your watch is right twice a day, broken watch is right twice a day. I just never set the time on that watch.
Patrick Moorhead: Forgot about that one.
Daniel Newman: I was laughing when you said that because I’m like, my watch is completely unpredictable. But anyways, we do have a lot left. And I do have to stop at the top of the hour here. So let’s hit the next topic, Pat. This was one, you wrote a great piece on Forbes, and then you and I actually perused the Arm party at CES, in which I was hanging out with you. And I had to hear no less than 10 times, “Hey Pat, great Forbes article.” It’s like, yep, you’re great. And they was like, “What’d you write?” And I was like, “Hey Pat, great Forbes article.”
Patrick Moorhead: Well, here’s the good news. I didn’t write it on Forbes. I wrote it on my own site, which-
Daniel Newman: I’m sorry. I’m sorry.
Patrick Moorhead: No, no, no, no. I just want to make sure people-
Daniel Newman: I heard, “Great article.” It was so assumptive of me.
Patrick Moorhead: No, I wrote a research note. We’re writing more and more on our own website. I mean, we’re putting out more research grade material like we did for Blackhawk. But here’s the skinny, folks. Arm’s business model historically has been, let’s give an architectural license to certain companies, very narrow, to address markets and geographies. So here, in smartphones, with Apple, based out of the United States, they gave Apple an architectural license so they could go off and build their own CPU. Now, designing your own CPU and building it can cost between about half a billion dollars all in. I know that because I used to do this for a living. It’s very expensive. And it is getting easier, don’t get me wrong. But if you’re doing a grounds up design, it is very expensive to take it out to multiple platforms. So the other business model, if it’s not an architectural license, is to take some off the shelf IP from Arm. And there’s varying levels of engagements that Arm has. Arm makes very little on the architectural licenses if there’s a ton of volume, like Apple.
Daniel Newman: It’s the law of diminishing returns.
Patrick Moorhead: No, it totally is. Then you have somebody like Qualcomm that comes along with the NUVIA based architecture, and they’re suing each other, by the way, on this. But let’s put that aside. You have the two largest smartphone platform providers on the planet that have architecture licenses, and the performance is so much better than what you get with Arm off the shelf, it becomes an issue. It becomes a market issue. And quite frankly, it’s a business issue for Arm.
So the big announcement was that Arm’s Blackhawk CPU is designed to be the best smartphone CPU core this year. And these were their words that they communicated to me, not mine, or inferring based on some rumor. I had a conversation with Arm, got relatively some basic information about it, asked some Q&A. But the goal here is literally that this Blackhawk CPU will outperform the leading edge from Qualcomm and Apple. And that is a big, bold statement out there, almost as audacious is saying that Neoverse would be higher performance than Xeon, which, by the way, on certain workloads, it is. And we see that at AWS with graviton, and we’re seeing a lot more people sign up. Like Microsoft brought out their new Arm-based SOC, and it’s likely that Google will too.
So is it audacious? Absolutely. Is this a giant check to write to say that this would add these performance levels? And by the way, I’m assuming performance levels at the right wattage and at a reasonable cost. But it’s a big deal. It’s a big deal for Arm, and I think it’s a big deal for the industry. So you can imagine people like MediaTek, Samsung related to Exynos, even though all the premium stuff today is Qualcomm. But I love the competition. That’s where I’ll end this. More competition is good. Arm has to prove a lot of things. But by the way, this design’s done. If phones are going to be out at the end of this year, then it’s done. The paint is basically dry and it needs to be implemented. So they must be very confident in what they think that both the Qualcomm NUVIA roadmap is, but also Apple.
Daniel Newman: Yeah, great research note, Pat, on Moor Insights & Strategy’s blog. I do appreciate that. Look, you did a really good job, and your assessment hit the high notes of what I was thinking about. And the things I’m thinking about is, Arm has been a licensing company and it has been a licensing company in an era where very few companies could build meaningful chips. And so now they are a licensing company in an era where it’s becoming increasingly en vogue and possible for more companies with resource and capacity to design, and not only design chips, but to take something out of the box, put it into use and create competitive products.
In 1995, people really cared what chip was in their PC. I don’t know that that’s as true anymore. And that’s not really just my interpretation. This is from talking to the leaders of the OEMs about silicon diversity and hearing what they are saying about, well, is it good to have 3, 4, 5, 6 variants of silicon? And of course, you and I will always say it’s great for competition. It’s good for competition, but is it good for the consumer? And I don’t know that the consumer knows.
Someone said to me, Pat, that Apple could take a dump in a box and sell 10 million units. And that doesn’t exactly correlate to Arm. But my point is that if the right company is selling the product, people will buy it. And so Arm is seeing an increasing financial gain and, especially now as a publicly traded company, has to figure out a way to grow its revenue, and that’s a per device growth of revenue, in the areas at which they compete. We know that the work they’ve done, excuse me, the work they’ve done in data center has been successful, profitable, but it’s still very small for Arm. It’ll grow.
But the way they’ve been able to support and enable a whole new group of entrants into the market, they’ve been able to do very well from that, the more they can create a high value IP licensing agreement where people will take and do minimal modification, and in the case of their lowest level licensing, there’s no modification. Am I right, Pat, about that, about no modification to the ISA? But this is a way you sort of hedge. We know there is an ongoing litigation that’s going on with Qualcomm and Arm. I’m sure this has something to do with it.
On the other side of this though, Pat, I’ll say if Arm can build a leading end, high-end processor that you can take right out of the box, it sort of brings some interesting questions to OEMs that have either, A, spent a lot of time and money to modify them. And B, what’s the value of that work, the architectural work to modify them if someone can just take it out of the box and bring it out to market? It also brings the opportunity back to the forefront for RISC-V, I think. But I will say, we know that’s a longer journey.
So if some of the companies feel that Arm is becoming too much of a competitor and they don’t want to keep building the Arm ecosystem, could they go a different direction? I don’t see that happening very quickly. But I think that risk always exists, and I think it’s a calculated risk right now for Arm. But look, they’re moving very quickly, Pat. They’ve got pressures of being a public company. They’ve got a reason to want to get in right now with this AI cycle. And the sooner this can hit the market and be available, the more the company can potentially benefit from it. I’m going to leave that there.
Let’s talk about HPE and HPE buying Juniper. So this one’s with me. But Pat, look, it happened really quickly. It happened really quickly. Now, where it started for me was I got some rumblings from some journalists. I heard a few journalists starting to track the story. I think it broke on the Wall Street Journal or Bloomberg. I heard from one of the Wall Street Journal. By the time I had a chance to call back, the deal had been announced. So you and I were lucky enough to share some time with both CEOs and have the opportunity to ask some questions. So we talked to Antonio Neri and Rami Rahim, the CEOs of both companies. And look, there is a unified belief between these two CEOs that together, from branch to edge, to cloud, they’re going to be able to come up with a comprehensive networking portfolio to support the massive growth of AI. When I say AI on-prem, I’m basically talking about private cloud and AI.
And I believe it has a lot to do too with being able to deliver consumption based compute networking offers for customers in the GreenLake platform. And that’s been the focus of Antonio Neri now for multiple years. This gives excess portfolio competency capabilities, capacity and customers. It definitely adds meaningfully to HPE’s telco portfolio. Juniper has incredibly deep roots in that area, very successful long time. Provides equipment to many of the world’s largest telcos. And of course, there’s a component of what they’re going to be able to provide to cloud providers. And so that is another opportunity for HPE.
Am I immediately confident that this is going to be a creative to growth at the same rate as Antonio Neri and Rami Rahim? My gut tells me that together they can eliminate one source of meaningful competition. And then instead of focusing on having HPE, Juniper and Cisco, now HPE and Juniper can just focus on Cisco. Now, I mean you have Arista, you have others that are out there. But I do think that that’s a big part of the story of what HPE and Juniper is focused on. I think that the synergies, they’re talking about nearly half a billion in three years. I think that will be achievable. I tend to believe that most companies are over their skis on what they’re spending, and you can definitely get synergies. Nobody has proven that more than the CEO of Broadcom, Hock Tan, of what’s achievable.
And I don’t want to make this just about personnel, but just about everything from go-to-market strategies. I mean, look, Pat, you can take one of the logos off one of the cars. I’m pretty sure Juniper and HPE both sponsor F1. They might even be on the same car. Well, there’s a hundred million in savings. But look, I think the engineering capacity and capabilities, I think the opportunity to package it, GreenLake it, make it consumption based. And the opportunity right now to build data centers for AI, especially for enterprises that have substantial AI workloads on-prem, is palpable. It’s a coming together that has all the markings and makings of success if it can be done correctly. But Pat, as someone who’s done a few acquisitions, not nearly at this magnitude, I can tell you the execution oftentimes far outstrips the strategic opportunity.
Patrick Moorhead: So this one’s an interesting one in that it’s hard to have a definitive hot take. There’s a lot of caveats here. I mean, financially, it makes sense for both companies. It’s a creative immediately and Juniper stockholders get something. But that’s not how we look at things as industry analysts. We look at them more strategically. What I’m trying to get through is the, oh, we’re bringing different things, but we have all this overlap.
And I got to tell you, it’s been hard to muscle through because when I look at the AI capabilities with Mist, which by the way, that is a definitive winner here, and then I look at things like Aruba Central that pull things together. How, if there’s duplication, are the regulators going to look at that? Is it that they have such small market share compared to Cisco and it doesn’t matter? Or are there different things to be put together? I mean, Juniper was a telco play without Mist. Mist made it a data center play. But that doesn’t mean that they’re going to go in there and steal the core switch leaf and spine and everything from Cisco. It’s very difficult, unless you’re Cisco, to get anything in there but the edge. And this is exactly where HPE has been so strong with Aruba. Not that HPE does do some data center networking, but it’s not, market share wise, a big player.
So I need to do a lot more work on this. Will, my VP and principal analyst of networking, cranked out a piece on the website. You could find it in the show notes. But this I think is going to take a little bit more research on my side to get it. There are puts and takes. In the end, Daniel, like you said, it’s going to come down to execution and integration to make this work.
Daniel Newman: Absolutely, Pat. Well said. We have about 10 minutes left for the show. Great analysis so far, Mr. Moorhead, I always appreciate you making time for me.
Patrick Moorhead: I wish I was more definitive on this one, Dan, but this one has a lot of caveats. I want to say this is it, but it’s unclear to me right now.
Daniel Newman: It’s a big deal and big deals come with a lot of complexity. Pat, let’s hit Mobileye. They had a really interesting week. It started out the week, or was it the end of last week?
Patrick Moorhead: January 4th, last week to the day.
Daniel Newman: Okay. What happened there?
Patrick Moorhead: Yeah, so first of all, Intel bought Mobileye, brought it in. And like we said at the top of the show when we were talking about Intel getting back to the automotive market, it was really their automotive play. And then based on, quite frankly, valuations that Intel was getting, the best thing for the shareholders was do a soft spin. Soft spin, I mean let’s give it its own tracking number, let’s roll it out. But Intel retains the highest percentage of ownership. And what they did is they had a few rocking quarters. And then they came in basically forecasting being down, I think 50%.
And you and I have gone back and forth on this, really good discussions on this. It’s like, hey, is this kind of like the PC market where everything built up and then demand went away? Or was this something else, maybe not managing your distribution, not having a good pulse on your customers, or maybe you were incented to push product, incentive to push more product in? And for what it’s worth, customers won’t take more product than they need. So pushing this, you can incent them with pricing. But that doesn’t mean they’re going to use them. So their stock went down big time. And I think people then were questioning what happened.
The overall auto market is single digit increases. But when you narrow in on the more modern, let’s say, ones with safety, you’re looking for much bigger growth on that. So that was why there was so much surprising in here because it’s not like Mobileye is just updating what they have in the next version of the cars. It’s adding more ADAS to more cars. So it’s really about a SAM thing rather than, oh, the ebbs and the flow of the auto market. So with that said, I think the Mobileye ADAS offering is strong and it is doing well. I do think in the future it will see a lot more competition from the likes of Qualcomm.
Daniel Newman: So Pat, I’ve had a bunch of theories. And that’s all I have, is theories right now. Look, what’s interesting is Intel’s going a software defined route. It feels like they’re almost hedging Mobileye. They still own 90% or so of the company, but they’re kind of hedging blackbox versus sort of software defined building blocks. Second thing that’s going on was you’re talking about a company that IPOed several quarters ago. Almost every company that comes out of an IPO comes out with an intent of some period, usually four to eight quarters, where they’re looking to stabilize revenue, stabilize EPS, and be sure that they’re able to reward their early investors for getting them listed. Now, that’s not a hard and fast rule. That’s just something that I have sort of witnessed. Companies very rarely come out and bomb in the first. If you bomb in the first few quarters, it’s doom, because people kind of know that generally you can’t.
And so what I’m saying is I think there were some very, very big contracts with big commitments, and then there was some hope that that volume and growth and market strength wouldn’t slow. But I think if you kind of had forecasted this, there was a reasonable expectation that at some point the chickens would come home to roost, and there would likely be a quarter where the sales would fall off a cliff. And I’ve had conversations with some companies in the automotive silicon space, and no one else is indicating such a drastic miss or fall off. So while cars have softened in the market and the demand has shifted, and high interest rates have certainly slowed down some of the purchasing, not to the extent of 25% a quarter or 50% over the course of the year. This was definitely somewhat missed. Pat, you and I, I think both had an eyebrow go up. How did you miss that? People missed-
Patrick Moorhead: How? Exactly.
Daniel Newman: But when you’re forecasting and you’re working on vehicles, remember, vehicles have a minimum of three, four year, and more likely seven or eight year design cycles. So there’s probably more questions than answers about how this got missed this drastically. And it is probably encouraging for most in this space to find out this isn’t an epidemic across all automotive and all automotive silicon. It was seemingly limited to Mobileye. I think they’ll bounce back, but I don’t know if they’ll fully bounce back. This is going to have created some ripples in trust between Mobileye and its investors.
Patrick Moorhead: Yeah, it typically takes five or six quarters to get the trust back. But yeah, this was tough.
Daniel Newman: This was. So let’s finish off with some big picture thinking. I kind of started here, so I might give a few thoughts and, Pat, actually give you the bulk of time if you want it. And by the way-
Patrick Moorhead: Yeah. And I want to also give you a platform to brag about your research too.
Daniel Newman: Oh, yeah, yeah, yeah. Thanks, buddy. Listen, 2024 is the year of implementation. 2024 is the year that all the investments made in ’23. All the launches, every stage that Jensen Huang graced us with to talk about Salesforce, ServiceNow, IBM, AWS, everywhere else, whether it was large language models, applications, CX, DevOps, DevSecOps or DevSecITOps or DevSec, IT, observability ops or so on, we’re going to start to see companies looking to meaningfully implement to gain these productivity advantages and to create these efficiencies in their business and the spending in the middle enterprise, the larger enterprise, these companies are going to turn up the jets.
So future of Intelligence did a large survey. We looked at over a thousand enterprise buyers, larger companies that are making investments. And we found that there’s an expectation, Pat, that there’s going to be a 3X investment in AI implementations at the multimillion dollar spending level. Meaning, companies are going to be spending multiple millions on the implementation of AI at about threefold higher level than they did last year. So that’s a big spending increase. And what are they looking for? Well, they’re looking for just a couple of things amongst their buyers. They’re, one, looking for expertise. Now, expertise is still nascent and new in this area. Everybody says they’ve got it. Most of these applications are still pretty new. Most deployments are still in proof of concept or early innings. And so the implementation is going to be largely done through SIs.
There’s been a show through our research that there’s a fickleness among enterprises, meaning they’re bouncing between different SIs and different vendors because they’re looking for not only that expertise, but that speed of implementation. They want it fast. And it’s also opening the door, I think, to really out of the box. And the early iterations of AI will be heavily out of the box. Because people using Salesforce or Oracle NetSuite, they want to be able to get AI benefits and features without having to build and implement major stacks to their technology. They want their data to be able to use some sort of very easy to plugin IPAs to hit system to system and be able to do forecasting, do churn, do predictability. And so I think it’s going to be a big year for this, Pat. I think we’re not going to be fully implemented by any imagination. But kind of the way I see a wave of training to inference, I see a wave of investment to implementation that’s going to take place in 2024.
Patrick Moorhead: So it is the year of productivity, and I think that’s going to be, first and foremost, the big driver. Same people getting more work done, less people getting the same work done. What kind of work is that going to be? It’s going to be the work that wasn’t transformed by dotcom and that wasn’t transformed by Web 2.0. Dotcom impacted travel agents, stockbrokers, brick and mortar, electronic stores. Web 2.0 is about streaming music and movies and content and free news and stuff like that. And we’ve seen how that transformed.
So this round, it’s about legal, it’s about healthcare, it’s about accounts payable, it’s about accounts receivable, call centers. The sea of cubes when you walk into a large company today, that, first and foremost, is going to be transformed by generative AI, productivity tools, first starting with SaaS, like we’ve seen from Microsoft, Google, and Salesforce. And then building your own applications, with the likes of IBM, Dell, HPE, Lenovo supporting that. So that is. It’s about implementation. And we need to find new ways to measure this. And there will be new winners, right? The folks not getting a lot of credit today for, or they got last year, you’re going to start seeing increased revenue and increased profits driven by this.
Daniel Newman: And those business models will come to life.
Patrick Moorhead: Yeah, absolutely. And there’s charge more for it, Microsoft model. Give it away for ‘free’, which is the Zoom model, right?
Daniel Newman: That is exactly, exactly correct. All right, Mr. Moorhead, we did it. Show one of 2024, or episode 199,002. Oh wait, just 199. Pat, where are you going next week?
Patrick Moorhead: I’m headed to Samsung Unpacked on the West coast to check that out, new devices, new cool stuff.
Daniel Newman: Did you even unpack in the process or you just?
Patrick Moorhead: Not really. Not really. I did have to do my laundry though.
Daniel Newman: Did you do your own?
Patrick Moorhead: Sometimes.
Daniel Newman: Good for you. Good for you.
Patrick Moorhead: Yeah. I’ve been doing my own laundry since I was eight years old. I tried to tell my therapist, a lot of therapy later.
Daniel Newman: I did my own laundry once. I did my laundry once.
Patrick Moorhead: That’s good.
Daniel Newman: All right. Well, I’m off to Switzerland. I’m going to get a new experience. I have snow boots, snow hats, snow pants, snow gloves, and I still don’t have any shirts that I can put a tie on with, but that’s just because I got a fat neck.
Patrick Moorhead: Don’t get that pretty head of yours cold. I want a big, huge fur. Hopefully it’s a fur cap. I hope it is.
Daniel Newman: It’s obnoxious. It’s as bad as you would suspect.
Patrick Moorhead: That’s awesome.
Daniel Newman: Yeah. So thanks everybody for tuning in. Look, we’ll be doing this each and every week, except for the weeks we don’t. Hopefully we’ll be back next week, if we can make it work from remote. But we will be here most weeks because we care and we want to talk about what’s going on in tech with all of you. If you enjoyed what you saw, hit that subscribe button. Join us here for all the episodes of The Six Five. But for this one, from Patrick and myself, time to say goodbye. See y’all.
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.
