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:
- Dell Tech Q2 2024 Earnings
- MongoDB Q2 2024 Earnings
- Broadcom Q3 2023 Earnings
- Elastic Q1 2024 Earnings
- Huawei’s Mystery P60 Smartphone
- Qualcomm IAA Announcements; AWS-BMW Tie up
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.
Daniel Newman: Hey everybody. Welcome back to another episode of The Six Five Podcast, episode 183, Pat. We’re charging towards 200 from our Friday pod. While that doesn’t sound like a lot when you’re only doing it once a week, this has actually been quite a commitment and it’s one of the best days of my week every single week, and it’s why you and I tough it out when we’re on vacation, when we’re overseas or whatever else we’ve got going on. I’m back from Munich. You’re on a, I’m putting vacation in quotes because you’ve done a horrible job of it, Pat, but it’s Friday, and by the way, this is the one part of the week where even when we are off, better to do it than not to do it. How are you doing today, buddy?
Patrick Moorhead: Man, I’m doing great and I do love the pod for a lot of reasons. First of all, when we’re off doing our stuff, I get to see my bestie, but it’s just a different way to get out content. While it’s easy to say that, oh, everything is research grade in content that we pump out, the reality is we have various levels and various ways of educating markets, educating enterprise and users, the tech beltway. I think our audience appreciates our ability just to crank something out, not only in a different medium, but it’s almost all opinion. It’s good to be here and I am transmitting from my bunker in Colorado here doing a really poor job of taking time off. I did take Monday off. I’ll admit, I didn’t work a whole lot on Monday, but I’m back and yeah, ready to crank it out. What do we have in store here, Dan?
Daniel Newman: Look, we’ve got a bunch of earnings from last week because you and I know that while it’s not our favorite thing in the world to do, a lot of our big tech companies clients, you had OEMs, you had software companies that reported last week and we just weren’t able to get to them. And then we had a big week this week and I was over in Munich. I was at the IAA Mobility Show and we’ll talk a little bit about that, but there’s also a bunch of other tech news that we’re not going to talk about today, but I just want to flash it across the screen. There’s a huge Digital Markets Act move with all these new gatekeepers, Pat. We’re probably going to have to come back and talk about that at some point.
There were some other interesting breaking stories that came out throughout the week with Apple and Huawei. We’ll talk about that a little bit on this show, but it’s been an interesting week for tech and it’s fascinating how fast things are moving. Of course, this morning, I don’t know about you, but I got up early. I was walking on the treadmill because I don’t run, I’m too old but I was walking uphill a couple degrees. That’s why I’m so fit, everybody. It’s the walking.
Patrick Moorhead: Yeah, I’m pretty sure those triceps are from walking, Dan.
Daniel Newman: Well, I walk on my hands. That’s how I got those.
Patrick Moorhead: Can see that.
Daniel Newman: Good one.
Patrick Moorhead: It’s a good one.
Daniel Newman: Of course, I’m listening to pundit, op Pining about what’s going on now with the iPhone ban for Chinese government, Chinese employees or workers. Doesn’t everybody work for the Chinese government? I don’t know how that actually works, but I’m pretty sure. Like I said, we’ll adjacently get to that topic later today, but just so much cranking along. I guess I’ll do the quick disclaimer for everybody that’s watching this show. If it’s your first time. Remember, we focus on the analysis. We try not to do too much of the news. We’ll do the recap where we have to do, but we really try to get beneath what’s going on with these different companies. But the show is for information and entertainment purposes only. While we are going to be talking about publicly traded companies, please don’t take anything we say as investment advice. That’s the disclaimer.
Pat, like I said, we’re going to cover four earnings from last week and their implications. Then we’re going to cover a couple of pieces of news from this week. First, it’s going to be Dell. We’re going to hit MongoDB, Broadcom, Elastic, and then we’re going to talk a little bit about Huawei and their new crazy, sweet phones. Are you using one yet? Don’t answer that, tell me later. And then I’m going to talk a little bit about the announcements that came out of IAA, at least some of them. It was a huge event, couldn’t cover them all, but we’re going to cover some of them. Pat, why don’t we dig in because we both have to get on with our day and we want everybody out there to get on with theirs? Dell Technologies.
Patrick Moorhead: Yeah. Dell Tech absolutely had smashing earnings. They were up 10% the day after and they’ve kept that going. The interesting part is their business was way off. Data center was off 11%, PC was off 16%, but they absolutely… 53% EPS beat and a 10% revenue beat. Again, folks, if you don’t cover the markets, don’t get confused by investor expectations versus guidance. They’re two very different things. What this basically said is they outperformed on what Wall Street analysts thought that they would do. The interesting part here is they not only crushed it but they gave a really good and positive forecast.
One thing I wanted to highlight was this was Dell’s first time really drilling down into what they’re doing in AI. Dan and I have talked about this many times on the show. I think there’s different tiers of companies who will be taking advantage of AI in different ways, but also different timeframes. You have your data center play, you could even split that into hyperscaler and on-prem. You have the data center edge, let’s say a drive-through of some sort, and then you have the far edge, which are things like devices, smartphones, IoT PCs and tablets. And each one of those has a different way that they’re going to be able to take fuller advantage of generative AI. What Dell Tech did, in fact specifically Jeff Clark, by the way, was joking, hey, you should have Jeff Clark talk about every earnings, the stock goes up 11%. I’m sure he would find that funny.
But what Dell did is they did bring out an AI TAM that said it’s going to increase 19% CAGR over the next few years to $90 billion. They also created a new classification called AI servers, which is nothing more eloquent than saying the servers that have giant GPUs in them, said it was a 20% of their server mix in the first half. They also said they have a $2 billion worth of backlog of a specific server. This is one you could put four H100s in. It’s called the XE9680. They said their pipeline is significantly higher. So, I’m thinking five billion, is it 10 billion? Then they went over in the different ways, a macro view, Dell validated designs for AI. Dan, you and I have had two guests on The Six Five that talked about this.
Then there’s the Dell Professional Services and then there’s the Dell Precision workstations that in many ways, shape or form is doing the initial programming, but also as we’ve seen with ML and DL, workstation programs are capable of taking advantage of that. Finally, to make a long story longer, they went through six customers that are all about AI. The one big surprise to me was Workday because thinking interesting, Workday thought that was a SaaS play. And they would plug into IaaS and then I remember, oh, okay, Workday timing actually was built before the cloud, but was part of internet very similar to Salesforce. It started with Web 1.0. But anyways, the market loved what they heard up over 10% and they’re sustaining it even when they got downgraded yesterday.
Daniel Newman: Yeah, I think the overall market reaction was very positive. I think what you said in the beginning, Pat was probably the most important nuance.
Patrick Moorhead: I should have just stopped there.
Daniel Newman: Well, it’s just the most important nuance and the nuance is there’s an argument because Dell had a great quarter relatively. Someone said to me once, “When you put the bar on the ground, it’s easy to hop over it.”
Patrick Moorhead: That’s good.
Daniel Newman: I’m not saying…
Patrick Moorhead: That’s your second one of the day. Do I need to start having a tally for the-
Daniel Newman: Yeah, I’m a funny guy.
Patrick Moorhead: … smart, funny things you said?
Daniel Newman: But this by the way, isn’t unique to Dell, so don’t take that as a knock on Dell. What I mean is the market largely saw the pullback for Dell. The market was very aware and so was Dell of the PC pullback. They knew that there were many quarters of lower demand, and so the guidance has been largely reset, and that’s part of the process of a market moving from an upturn to a downturn back to an upturn is going to be readjusted guidance. Of course, everybody evaluates a company based on future earnings. That’s how valuations are done for businesses. So, seeing that the company can now perform means higher multiples in the future, but then there’s all the extraneous factors that could be higher for longer interest rates.
We’re seeing Fed policy continue to indicate that we may see a longer lag, but then the market’s also largely very strongly associating value right now with AI. So, I think Dell’s announcements around AI providing classification clarification around how it’s going to monetize AI was encouraging. And of course, I think what we learned from looking at Lenovo, HPE, from Dell and others was all of them have pent-up demand because basically right now, any company that can get their hands on A100, H100 inventory from NVIDIA can sell it through. And it’s also a margin enhancement because everybody’s got a ton of price elasticity right now where they can get it. If they can get it, they can sell it and if they can sell it, they can sell it at a premium, which offset the fact that quietly the CPU market has just tanked. It just absolutely tanked.
This is what’s really happened there, for anybody that’s curious about why that happened, it’s because there hasn’t been more budget enabled for IT. What’s happened is all of the budget has been moved over to the GPU spend or for AI spend. And this is going to be what I believe is going to be a short-term to midterm trend. It’s going to go over four to six quarters, maybe a little bit more. And I’m not saying at that point it’s going to be a huge reversal, but what’s going to happen is we’re going to realize a lot of inference is going to be done still on a CPU. It’s still going to be done on an edge device on a PC. Pat, you’ve written about this. I’ve written about this. We are in this major flux of investment around training, but we know a lot of inference for applications can and will be done on lower cost CPUs and that should eventually swing a pendulum back somewhat that way.
The most interesting thing, Pat, that I think we are going to watch, which is a little bit less related to Dell, and I realize I’m talking more macro, but you really hit the Dell stuff well, is that will NVIDIA see a steep cliff? Meaning you’re seeing these numbers and these guides of just these extraordinary highs and at some point… Remember with Zoom Pat, when Zoom had that high from the pandemic and then remember they were still growing, but people saw the stock value, it dropped by like 80%, 80% because what ended up happening was it was growing at a much more moderate pace, but it was growing at a pace over the huge post-pandemic surge of demand.
What I’m saying is you might see a point now where NVIDIA sustains the size of revenue that it is or even is growing at mid-single double digits for GPUs and data center AI. And people are going to say, “Well, that growth has really slowed down.” But remember, it grew at X hundreds of percent for multiple quarters. That’s going to be really interesting, Pat is what’s going to happen when more and more AI moves, when we see the pendulum swing from training to inference once these companies have fully built out at least their immediate needs for capacity for AI training. That’ll be interesting, but you know what the good-
Patrick Moorhead: It is going to be interesting. NVIDIA either has to find another next thing, which I think they’ve done a pretty good job going from ML gaming to ML to DL to generative AI, or they go vertical integration and go after Intel and AMD on CPUs, which this new grace is really optimized for a CPU plus a GPU. But it would be interesting to see if they could field a competitive generic processor, which by the way, still you can get a hundred percent margins on. That’s a very good business.
Daniel Newman: By the way, I know this was a segment on Dell, so let me just close out the loop. Dell will benefit on all ends of this because they have the low end to the highest end and they’ll be putting the NVIDIA parts inside their highest end AI servers. But of course, they also have a huge line of servers for just traditional compute, which can handle a lot of acceleration for traditional AI inference or at the edge. That’s part of the beauty of Dell, why they do a hundred billion in revenue is because it has one of the broadest portfolios on the planet, but this will be an interesting thing to watch. So, good coverage on the Dell side, Pat, I’m sorry, I got a little macro there, but you know it’s a Friday.
Patrick Moorhead: No, no, I think part of this podcast is it’s organic and I think we get into what we want to get into and I think our audience appreciate that.
Daniel Newman: I’m into it, buddy. You know me, I’m an economist at heart.
Patrick Moorhead: Are you?
Daniel Newman: I’m an economist and a financial advisor at heart, but by the way, this is not financial advice, so I’m just being very clear about that.
Patrick Moorhead: Yeah, don’t go there, Dan. Don’t go there.
Daniel Newman: I won’t even recommend stocks to my kids. I’m like, “Don’t ask me.”
Patrick Moorhead: Yeah, exactly.
Daniel Newman: I always buy the wrong stuff. If I actually took my own advice, I’d be a lot better off.
Patrick Moorhead: I love it.
Daniel Newman: Anyways, so let’s go into MongoDB. We’re going to talk about a couple of software companies that are also inflecting very positively around the AI boom. Look, we all know that the data platform, the operational data layer is what’s going to make these really killer applications for AI. So, which are the companies that are going to get a big tailwind? Apparently, MongoDB is one of them. It had a very strong quarter and the market is starting to see, I think, its story move from being a really developer-centric product to a business-centric product where companies are leveraging the technology and its capabilities to be able to build, implement, and continuously innovate on AI’s capabilities for application.
You’re talking about a company that saw its revenue grow 40% year-on-year, which in software SaaS is very, very strong. It saw the SaaS part of its revenue grow by 38% and now is 63% of its Q2 overall revenue. The company is still not profitable, so on a gap basis, but it’s getting very, very close to being able to do that and it’s creating a lot of cash flow. You and I were at the event in New York City, so we had time to talk to many of the company’s executives. We had a chance to talk to many of the company’s leadership as well as its customers. Talked to some large banks. We talked to some other large companies and startups that are building on Mongo, a very positive sentiment out there, Pat, on the company. You’re seeing vector search, which is all about simplifying generative AI and semantic app builds. And we’ll actually talk about that more when we talk about Elastic’s earnings too, because we’re seeing this moving across the industry.
We’re also seeing that it’s deepening its partnerships. Part of this earnings, and also at the event we were at Pat is the company was busy announcing that it’s expanding in its cloud integration. It had a big announcement this quarter around Google Cloud, but really MongoDB has deep ties and roots across the hyperscale cloud community. Across the board, Pat, this is just one of those plays that if you believe that this generative AI boom is sustainable, then you’re going to have to believe that there are critical components, services and tools that are going to be required to enable this continued build out. Vector search, operational or developer data is going to be critical to basically building out the apps of the future. You can see it here with what MongoDB is doing. You can see it there by the momentum. And like I said, I think one of the really important key trends is this is moving from really a developer, nerdy tool to something that the market’s beginning to see as a very critical business component to companies that would want to build generative AI apps for the future.
Patrick Moorhead: That was quite an analysis of the things you pointed out. What I think is so interesting about MongoDB is how they moved from a best in breed database to what’s now a data platform. We’ve seen a couple companies who are doing that and it just makes sense. A company starts off, they have a killer app and they then try to extend that into something that is more and that’s bigger. And it’s a challenge because you try to grow, grow, grow, and you have to make expenditures. But absolutely, it’s working out for MongoDB. 78% gross margins is pretty incredible at that growth rate. It’s not like they’re giving away products, they’re growing. And if I look at what makes a successful data play, one of them is the hybrid multi-cloud.
MongoDB doesn’t support every feature on every cloud, including its on-prem capabilities, but it’s enough to get the job done. They used AWS as an anchor point and then moved to Google Cloud, and that’s exactly the types of companies you want to be working with. I totally agree with you on the vector adder and it’s called Atlas Vector Search, which by the way competes directly… It’s funny, it sounds like it competes directly with Elastic, but the two actually work together as partners. Congratulations to MongoDB and eye-watering 40% growth at 78% gross margins.
Daniel Newman: Let’s talk about a company that might be acquiring a company that might be a really huge mega deal in the near future that had a kind of… What do you want to call it, Pat, when you just constantly do what you do and it’s always solid? Broadcom.
Patrick Moorhead: Yeah, it’s interesting. Broadcom does so well so often that anything other than a complete blowout quarter gets a complete yawn. If you look at the prior quarters where… It’s just amazing how consistent the company is. The prior three quarters beat by 1.75, 1.65, 1.64 on EPS and then on revenue beat by 0.3, one by 0.2, one beat by 0.34. Investors freaking love this consistency. They did come out with growth, 5% growth in both semiconductors and software, but Wall Street wasn’t that impressed this quarter. One thing about this is I look at these AI darlings and I put Broadcom in there, there has been a softening across all of those, whether it be Marvell, Broadcom, NVIDIA, and even AMD. In the stock, Broadcom has been up 56% in the past six months. What’s happening structurally inside of the business is that AI is growing leaps and bounds, but the traditional non-AI networking is not able to make up for that mega growth in AI.
And it’s amazing how many people don’t understand that. Sometimes I just wonder if investors actually understand stock or not. But then again, potentially the same themes that brought the stock up 56% in six months are the same ones that they’re finally understanding, hey, it’s not a one for one in all of Broadcom’s business if NVIDIA goes up. There’s a portion of Broadcom’s business that goes up when NVIDIA goes up.
Final thought on VMware, I think everybody’s waiting for this. It’s going to add just a huge element of dynamism to the company in terms of what they can do. The cash coming in, the $2 billion investment in the cloud is going to be exciting. I am expecting some divestitures and I think as we’ve seen in every type of acquisition, there’s going to be some overlaps, particularly around the OpEx side, I’m most excited about is the way that I think VMware is going to be run going into the future, which is going to be more like a startup as opposed to a giant monolithic software company.
Daniel Newman: Yeah, I think you hit some really good points there. Nobody quite knows how it’s going to run. There’s been departures within the executive ranks, the comings and goings. There’s been some talk about whether Hawk is going to take the VMware part of the business and become his central focus with Charlie running the traditional chip and semi, but Hawk’s also not one to… He’s a delegator. We know the GMs of his businesses and people have a certain opinion about Hawk, but one of the things is you got to know him, he really does entrust the leaders of the various business units to run those business units.
The Broadcom way… Avgo we should call it, in terms of its stock, A-V-G-O Broadcom, that one, I’ve always wondered why they didn’t change the ticker, but people would follow. Don’t worry, they’d get it. But anyways, huge margins. You look at most products up over 70%, huge earnings on every share, very large dividend on a per share… A very significant dividend. I don’t want to say very large because the shares are very expensive, but I’m saying for people who own, it’s a highly rewarding stock. But the behavior is really, it’s about profit maximization. He does what most companies should aspire to do and it’s sometimes scrutinized, but he invests heavily in the parts, pieces and products in the business that make money. When he acquires companies, he’ll shed and divest anything that doesn’t create revenue and profits.
You can be pretty sure that that’s going to happen with VMware. It’s a very safe guess to bet that there’s going to be some things that are going to get chopped off. He’ll find buyers for it. And by the way, that’s just the way it goes because what he’s going to do is look for the cash cows in the business. He’s going to maximize the cash cows and he’s going to get rid of the parts that don’t make a lot of sense. Most big companies, they tend to fund their pet projects and their loss-leaders by having a couple of cash cows. And then there’s Hawk and his style is, I only want cash cows. And that’s why the company performs so well. The company’s guidance, it was a little bit calm from what the market wanted. You said it really well, Pat, there were these AI darlings that were the first to be able to say we are going to be impacted by a tailwind of all this AI investment. That hasn’t changed.
But I think what a lot of people don’t realize is that this is going to happen in phases and that the investment is not all happening at one time. For instance, a company like a Broadcom though, you’re tucking pennies on the dollar for networking against every GPU that’s sold. That’s what some of the people were missing is they’re thinking, well, how much content from Broadcom gets sold with each one of these NVIDIA pieces? And then of course, NVIDIA’s got some competition within its own ranks. Some of these companies, if they’re all in with all NVIDIA networking, there may be very minimal content, but also it’s still there. But over time, as this scales to maximum proportions like traditional compute data centers, Broadcom is going to be riding that wave all the way through. I’m really interested, Pat, by the way, this is the quarter for the VMware deal. If it doesn’t close this quarter, it might not close. Having said that, my stance has not changed. This is the quarter it’s going to happen. But again-
Patrick Moorhead: Why would you think it might not happen? I’m just curious.
Daniel Newman: I think it’s going to want to be done. What I guess I’m saying is I think it is going to happen. Nothing has really changed, but I think Hawk was steadfast in having it happen this year. So, I think there will be some disappointment. There will be a very hard push to get it done this year.
By the way, one of the things you and I have learned is just the longer things do linger, the more likely they don’t happen. That is just the thing that’s happened with big deals. It’s the NXP deal with Qualcomm. It’s all eyes on Activision right now as it’s yo-yoed, to and from, of whether or not it’s going to pass. The more these things linger on, the more likely of scrutiny. At some point, a lot of these deals have… What do you call that? Poison pills in them where if they don’t happen by a certain date, there’s a breakup fee. And we’ve seen that happen as well. Again, I’m not familiar if there is one in this deal, but things do have to eventually close because, Pat, you and I know, having never done M&A at that level, but one of the things you do know is that M&M creates a lot of cultural uncertainty inside of businesses when it’s going on big deals, especially deals of this size.
You got to acknowledge, for Broadcom, it probably hasn’t created all that much. But for people at VMware, I guarantee you, the last six months, there’s been a lot of uncertainty, especially knowing that you’re moving between cultures that are so different. Pat, I’ll leave it there having said that. Another strong quarter, I think I tweeted something along the lines of just another boring quarter of just doing exactly what you say you’re going to do by Broadcom. Nobody likes that. It’s no fun. But anyways, onward upward. Here we go. Hey, let’s talk about a company we’ve never talked about in earnings before, Elastic. We brought Elastic up. By the way, Pat, you know how you and I love a good deck. One of the best decks I’ve seen from an earnings deck for a company that people probably largely don’t really understand much of what they do was Elastic.
We’ll put it in the show notes, but if you actually check out the deck, it was very, very interesting. Elastic’s really trying to tell a story right now. We talked about vector search, we talked about the whole idea of search in the relationship. When you’re doing generative AI, you’re talking in natural language to a machine. When you talk in natural language in a machine, there’s a lot of complexity that goes into taking that natural language and returning you something generatively that not only looks and reads like real text, but that’s actually accurate. We talk a lot about hallucination. The company did a really nice job of telling its story of how it’s grown through these inflections and that AI is the next inflection. We’ve seen huge growth. We’re seeing data grow exponential. We’re seeing basically the ability for natural language and the ability for API that can connect all this data that companies have, enterprise data as well as internet data and create these applications that are going to be not only in compliance, they’re going to take care of PII in privacy and data.
They’re going to allow human-like interactions and of course they’re going to be able to take foundational and models that have been pre-created and use them in a way that’s utilities to individual businesses. This is that trend and shift of training to inference that I talked about is this is what’s going on, but who’s one of the beneficiaries? One of the beneficiaries is going to be any companies that enable intelligently plugging into your applications and then doing enterprise search. And enterprise search is really the ability to use your data. For everybody out there, LLMs are great, but as we all know, it’s table stakes. When everybody is asking questions of the same dataset, it’s giving you basically what I would call search 2.0 or new filtrations of recommender engines and filtering systems, which has been the baseline of AI. And if you don’t believe me, go back and look at where Jensen Wong spent a lot of his effort in his early days of building out foundation frameworks, Jarvis, Merlin, recommenders, filters.
This was where it started. Now it’s how do we plug in and do this very, very quickly? Well, Elastic focuses on observability within Elastic search platform, focuses on security within its platform, and then it focuses on actually having an extensible, flexible search that you can use. One of the things that I also really like about any company that’s a little bit smaller like Elastic, that’s doing things. We’ve had the chance to talk. I’ve spent some time with Ash, their new CEO, a very bright guy, but the company’s doing a very good job of telling its story through its customers. And I can’t tell you how much I admire that. It went down the list and it looked at who’s using its observability and why. It was able to talk about Comcast, Orange, a European company. It was able to talk about BMW. And it actually breaks down in their earnings deck, but really why companies are using it, how they’re using it, and how they’re going to basically be able to drive it to create fast responses, handle multi pediment scale.
And then of course, relevance. Relevance is what you talk about when you talk about hallucinations, inaccuracies. Not only can it search through all the data paired with the public internet, large language models, but can it find you the most accurate, most relevant without hallucination? And that’s what I think Elastic is trying to accomplish. It’s going to be a competitive market, but they were early to this market. They have a very, very important subset of customers. They plug in with the public cloud providers and they’re really making a nice pivot from the first generation of filtering of the internet through traditional search to generative text where I think Elastic is going to have a significant growth opportunity. Strong quarter strong growth numbers, Pat, I’ll turn it over to you.
Patrick Moorhead: Yeah, I agree with you on the deck. In fact, it was 34 pages of information before they actually got into the first quarter ’24 results. Revenue was impressive, not overly impressive at 17%. But one thing that I really like here, and again, I was talking earlier about best of breed, narrow startups that grew, then basically going from a search capability to observability, security and then search. And then you have generative Ai, which depends on a vector database and essentially what the company has been doing for years. For all you listeners out there who don’t understand what a vector database is you have a standard database, which think of it as a giant spreadsheet in the data center that’s text and numbers. Vector databases are for things like images and videos and any content that is not text or numbers. Essentially, you create vectors to be able to do pattern matching and run generative AI algorithms against that.
My prognosis for Elastic is they’re going to see some mega growth once the generative AI kicks in. Well, what do you mean, Pat, kicks in? We’ve got all this stock market height, probably 1% of data center usage is generative AI right now. Also, it does-
Daniel Newman: Good point.
Patrick Moorhead: … machine learning. Generative AI doesn’t mean that analytics, machine learning and deep learning is going to go away. I like to look at these different methods as a bag of golf clubs. You wouldn’t just take one golf club out to do 18 holes, use your driver on the putting green or use your putter to drive with. Maybe after a few drinks, but not on a day where you’re actually trying to…
Daniel Newman: Ever seen Tin Cup when he played the entire round with his seven iron?
Patrick Moorhead: I did. That was pretty impressive.
Daniel Newman: It’d be harder to do that with a driver for sure.
Patrick Moorhead: A lot harder. Those aren’t going to go away here. So, I was impressed too. You mentioned something about their customer set and it is impressive how they can get their customers to say yes. Go to slide 26 of this quarter’s investor deck, and they’re huge companies and they’re not all born in the cloud. You have born in the cloud stuff like Uber, and then you have the, I call it brownfield companies with people like Home Depot. So, a pretty impressive crew here. Dan, we may or may not be going to one of their events coming up.
Daniel Newman: Rumor has it.
Patrick Moorhead: Rumor has it, The Six Five may be going to their big event.
Daniel Newman: Maybe, we’ll see. All right. Pat let’s talk about something a bit more controversial. I know you used to be a big power user of Huawei devices. Thanks for lending all of your personal information to China. Appreciate that. Kidding, maybe, I don’t know, whatever. But the big news, Huawei is back in the phone business. They’d disappeared or at least they disappeared outside of China for a while.
Patrick Moorhead: Yeah, this is a great intersection between a lot of conversations that we’ve had on The Six Five and one of them is the ongoing saga of China versus U.S. regarding cutoffs in the technology. I always like to tell the big picture of this. This goes back decades to when China was blocking companies like IBM and Cisco. The thing that’s true today, which is most online services are blocked. U.S. online services are blocked in China. And China essentially copies the Facebooks, the Twitters, Google search and things like that. Here we are today where the U.S. thought it had cut off Huawei from advanced semiconductor tech, but also you have to get a waiver if and when, from the U.S. federal government, if you’re going to provide anything to Huawei. But here, this mystery, I call it the mystery phone, the P60 and the P60 PRO come out. On day of launch, the company didn’t even talk about what processor was inside, didn’t talk about whether it supported 5G or 4G, which quite frankly was bizarre.
And then you saw phones that were provided to reviewers. Now remember, this is a China only phone. Huawei sampled this well beyond Chinese phones. So, even the people who were writing about the phone didn’t even know what it supported. And oh, by the way, the timing just happened to coincide where the highest-level U.S. diplomat on trade Raimondo came over and was visiting. And then the entire Chinese press got around that, it was a feeding frenzy. A lot of questions about this phone. And then what popped up was Hynix memory. Hynix is not allowed to sell flash memory to Huawei without a waiver. And even Hynix is trying to figure out where this came from. Was it black market? Was it left over from 18 months ago when the initial sanctions were put in? Nobody quite knows. And by the way, why didn’t Huawei use indigenous flash? Why didn’t they use indigenous memory from China? That’s a weird one.
You and I have talked about how Micron has been cut cutoff or greatly reduced from certain types of applications here. There was some dialogue about Qualcomm and what this meant. Here’s the bizarre part. Qualcomm zeroed out Huawei a long, long time ago, so no revenue. But for some reason, again, I question the intelligence of the stock market on this one. SMIC, which by the way, is the foundry that supposedly created this seven nanometer N+2 chip, which by the way is close in characteristics to TSMC’s six nanometer. It’s not leading edge, not even close. People are wondering how they could possibly provide the volumes to Huawei. And the final bizarre thing about this is I think 10 days after launch or something like that, the company did say, no, this supports 5G, but there’s not a single video in any of the reviewers that shows it connecting to a 5G network.
Again, this is why I wanted to call this Huawei’s mystery phone. Apparently one of our state senators who leads a committee on commerce has opened up an investigation on this because if you think, how could SMIC do this without U.S. technology? Hey, did they use any of the software tools from companies that we cover? Did they use any of the core IP building blocks that have restrictions in China from the U.S.? Anyways, it’s the mystery phone. We might be talking about this a little bit more in the future, Dan. I don’t know.
Daniel Newman: We all know for sure that they don’t pay for a lot of software. That’s a well-known thing in China. That goes back to that joke I made about VMware. How many VMware licenses does China own? Anyways. But there’s so many weird implications here, Pat, and this all does tie together a little bit. I’ve talked a lot about China and microaggressions for a while. There’s so much conflict between are we doing this well or are we not in terms of how we’re managing this. I shared an article, Pat, a week ago that talked about how the U.S. is using all these cheap IoT sensors in military grade devices from China rather than using just slightly more expensive US-based technology. I know it’s not related to a phone, but I’m saying it’s weird that we pick something and then we get really focused on it and then we don’t pick other things. Are they really a threat or are they not a threat?
I was listening to one expert online this morning, a policy expert, talking about China’s aggressive consideration of going back for Taiwan, and it could happen in the next seven years. It’s funny how we just shrugged that off because that’s a freaking huge deal. It’s a huge, huge deal. But it’s like, well, it’s not a problem yet. But there’s certainly consideration that China’s ramping up to take military action with the U.S. Look, us cutting them at the knees on AI is not a small, inconsequential thing. We’ve also known for a long time that China’s a massive market for so many of these technology companies. So many of them are dependent on being able to sell stuff in China. And when you take away the ability for them to sell stuff to China, it does actually cause an impact to growth for US-based companies.
So, there is that push and pull effect going on. Can we sell and what should we sell? What’s a military grade risk? Is it just about military and defense or is it really just about technology leadership as a whole? The war for AI is a race, and the race is won by slowing them down. I think what we’re seeing with the SMIC and the seven-nanometer thing is they’re going to find a way with or without EUV to be able to get access and the way they’ll get access. We don’t actually know for sure, Pat, but I think you said something to me about gray market. You and I are talking about there’s always a way this stuff ends up where it’s not supposed to end up. What you brought up about SK, what you brought up about these different parts, but then of course, if China wants to put off this perception that there are this powerhouse national capable of building their own stuff, but then they’re still having to get all the capabilities, the intellectual property, the building blocks, the memory from outside of China.
How capable are they and how much of this can they actually build? How much, how fast can they build? At the same time, Pat, didn’t they just ban the use of iPhones from a whole bunch of their citizens? Mostly the government ones. But that was a pretty big step. By the way, Apple got pounded this week based on the proclamations of potential lower sales in China. They already knew Apple was making some strategic moves to go to India, building more in India, at least from a manufacturing standpoint. But the Chinese market is incredibly nuanced and complex right now. And them wanting to build their own phone, Pat. The Qualcomm thing, I think you nailed it. It’s a total nothing burger. It’s as bad as the Apple and Qualcomm relation conversations. Qualcomm has already for over a year or two years now in their investor decks communicated what’s happening with Apple. It’s no surprise and no secret there. Same thing with Huawei. That’s a weird one to me.
But the real question is can China build at scale competitive devices without using intellectual property that’s part of this ban or part of these chip controls? Pat, unless I’m mistaken, to build a super phone that’s going to compete with a Snapdragon or to compete with an iPhone, they need leading edge technology that’s probably going to be processed below seven nanometer. Unless I’m mistaken, I don’t understand entirely how they’re going to do it at scale when there’s no technology to manufacture those chip sets, at least that I’m aware of, that they have the access to get their hands on. By the way, I’m not saying I have all the information here, but if I’m mistaken, I don’t understand how this is going to get done at scale. By the way, China has a few billion people. So, if this thing started to sell at scale in China, are they going to… How much capacity could they make of these devices?
Patrick Moorhead: Not a lot.
Daniel Newman: That seems to be a problem for me to try to math this thing. How big of an opportunity does it really become for Huawei? By the way, does Huawei have any chance outside of maybe selling their phones in China and Russia? Are there any other markets that are going to really adopt these? I’m going to leave that there. Very interesting though, Pat. And there’s a confluence of things coming together that are just making this China thing more and more interesting to keep tabs on.
Patrick Moorhead: I call this the hallmark topic. It’s the gift that’s going to keep on giving here.
Daniel Newman: So, let’s have a little fun and let me wander off to IAA, Pat. We had The Six Five represented this week at Qualcomm IAA. I’m missing my bestie, but I brought our next newest bestie part of The Six Five, Diana Blast, brought the connected series there and I can’t wait to put those out. She did a great job. The IAA, it’s an event that converges mobility and technology and then automotive. And every other year they flip between transportation and mobility, so this is the mobility year still full of cars, vehicles, drones, two-wheelers. Spent a lot of time with Qualcomm, spent some time with CEO, Cristiano Amon and automotive chief Nakul Duggal as well as chief marketing officer Don McGuire.
So, the company was in full force there, had all its big wigs there. Automotive has been a massive focus for Qualcomm. We’ve talked about this a lot though on the show, so I’m not going to beat that to death. But that 30 billion pipeline is coming to fruition with these continued design wins that are coming from BMW, from GM, from Mercedes-Benz. By the way, there were some killer concept cars there, Pat, really, really cool concept cars, which is by the way, it’s bringing love together for me vehicles-
Patrick Moorhead: Well, it’s like torture. You see these cars that may or may not actually ever see the light of day.
Daniel Newman: We just got to get rich enough that we can buy the concept car.
Patrick Moorhead: Yeah, probably not.
Daniel Newman: Mercedes had a killer cool car and so did Porsche, which by the way I… By the way, is it Porsche or Porsche?
Patrick Moorhead: It depends on where you are. The-
Daniel Newman: I’ve never quite known what’s right.
Patrick Moorhead: German pronunciation would be the er Porsche.
Daniel Newman: Okay, well, they’re right and we’re all wrong.
Patrick Moorhead: Well, they created the name.
Daniel Newman: Nein.
Patrick Moorhead: Dr. Porsche.
Daniel Newman: Anyways, Qualcomm focused on three or four really interesting things at this year’s. One is we’re moving from ADAS and sensor-based AI in these vehicles to generative AI in the infotainment and management maintenance of the vehicles. Some digital assistant updates we’re focused on is how generative AI is going to be part of these building blocks of the digital chassis. You’re going to be able to have an interactive digital assistant. And this brings a really interesting conversation about we’re moving from device-based automotive control for infotainment to moving to fully onboarded generative capabilities because we all know that most cars today are, a lot of people are plugging into Apple play.
A lot of people are plugging into Android-based systems. By the way, this still could be Android based in terms of the interface, but in terms of are we using the smartphone as the compute or are we using the vehicle as a compute? And then for generative tools it’s where does that processing actually exist? Very interesting, but some very future forward-thinking about how this digital assistant is going to work together. Another one was the two-wheeler, Pat. We’ve talked a lot about cars, but this was pretty cool, so-
Patrick Moorhead: And that was teased in the Financial Analyst Day you and I attended in New York this year.
Daniel Newman: Yes. One of the things that Qualcomm really focused on new partnerships with companies like Harley Davidson. By the way, it’s not only cars that are going to move to electrification. It’s not only cars that are going to need a digital chassis. So, the ability to put building blocks of navigation and telematics and ADAS and infotainment and then be able to provide the capabilities to do this with a smart chip set and then all on electrification, it’s going to come to bikes too. Qualcomm is able to identify a fairly large market expansion opportunity and also an opportunity to drive more design pipeline. There’s a strong amount of demand. It’s not just two-wheeler, I think it’s just the ability for Qualcomm to be the provider of intelligent platforms for anything that rolls or flies or floats. I’m waiting to hear about the boating interfaces that come next, Pat.
Patrick Moorhead: Yeah, the corporate yacht.
Daniel Newman: The corporate yacht and the yet and the yar and all the other things that you move around in. That was another strong thing. There’re a couple of other things that were interesting too, is more partnerships, another one with BMW was announced. And Pat, this is just a trickle of constant for Qualcomm and you’ve seen they’ve flip-flopped with NVIDIA. NVIDIA was big… And NVIDIA, while it’s gotten all this data center growth in business, quietly Qualcomm’s come underneath them and swiped a lot of design pipeline. Remember, this doesn’t happen over one or two years. What happens is over four or five, six years, this gets designed in and the revenue comes. That’s why a $30 billion pipeline is only yielding three or 400 million in revenue on a quarterly basis because these designs take time to come out but then the revenue pivots.
What has happened is you’ve seen a steady flat to decrease across NVIDIA’s revenue. It’s because a lot of these new designs have gone over to Qualcomm, which has been a real strength in its diversification. It’s been necessary given the rest of the portfolio and what’s gone on in the market with mobile devices for instance. And then the last thing, Pat, that was a big one, is a design partnership with a mega hyperscale cloud provider for vehicles of the future. That was a commitment that the companies, AWS and Qualcomm made to co-innovate around the future in automotive and mobility. And that came out as well, so lots of announcement. I realize we’re getting close to that time and you and I are going to have to run. So, I’m going to hand you the mic to give the last thoughts on the Qualcomm stuff. Since you weren’t there, I figured I’d try to cover most of the ground here.
Patrick Moorhead: Yeah, you pretty much covered every single announcement related to Qualcomm there.
Daniel Newman: Yeah, I was there.
Patrick Moorhead: What I would like to dive into though is the very provocative announcement. It was really a three-way announcement between AWS, Qualcomm and BMW. By the way, very rarely do you see this from AWS. While I don’t have all of the skinny on it, I feel like I can say enough to say that Qualcomm likely brought AWS BMW and BMW is going all in on the A100 or a variance of it. As we know about Qualcomm, their IP, it’s building blocks. Similar building blocks, just more of it that you might find in a smartphone will be super-sized when it comes to Qualcomm’s automotive. If you look at the A100, it’s likely a super-sized version of that.
BMW needed a development environment. And by the way, AWS and BMW said this is about the Neue Klasse, that means the new class of cars set to launch in 2025. That is not a lot of time to be able to crank this out. But there was a quote from Dr. Nicolai Martin, an SVP at BMW, where he tied together AWS and Qualcomm technologies, which we don’t see that often. And I have to wonder what on earth is NVIDIA thinking about all of this. And you gave a little history. In fact, the history goes back all the way to where Intel was the preferred supplier for BMW with Mobileye and then that moved to NVIDIA. And it looks like BMW is all in on Qualcomm at this point. As we found to have a development environment, taking those capabilities to the public cloud where you essentially have endless capabilities will be an interesting thing to follow.
I just have to wonder, is this truly the, I would call it the credible beginning, meaning Qualcomm, there were a lot of… What’s the right word, Yeti sightings on Qualcomm’s AI 100, but it never really made a connection. This is the first public cloud connection we’ve seen for Qualcomm’s data center Silicon. I’m interested to see what’s going to happen in the future here.
Daniel Newman: There we have it. Pat, you did it. Can you take the rest of the day off?
Patrick Moorhead: I’ve got one or two more meetings.
Daniel Newman: Terrible day off. Just want to say you did a terrible job taking time off. You’re fired.
Patrick Moorhead: Thank you.
Daniel Newman: You’re fired. All right everybody. Great show. We covered a lot of ground but we got to go. Hit that subscribe button. Join us for all of our Six Five episodes. Pat, where are we going to be next week? We’re going to be in beautiful San Francisco. We’ll be at Dreamforce 2023. It’ll be a firstie for you, bestie.
Patrick Moorhead: It will, yeah. I’ve been to a lot of shows in my time, but I’ve never been to a Dreamforce. Pray for me in San Francisco, hopefully nothing will happen there.
Daniel Newman: Well, let me just say it’s unlike any other Moscone event you’ve ever been to. So, let’s let this one go. We’ve done it. We’ve done the show. Hit that subscribe button. We appreciate you. We’ll see all later. Bye-bye.