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:
- Qualcomm Q1FY25 Earnings
- Arm Q3FY25 Earnings
- AMD Q4FY24 Earnings
- Amazon Q4FY24 Earnings
- Alphabet Q4FY24 Earnings
- IBM Investor Day
- Coherent Q2FY25 Earnings
For a deeper dive into each topic, please click on the links above. Be sure to subscribe to The Six Five Webcast so you never miss an episode.
Watch the episode here:
Listen to the episode on your favorite streaming platform:
Disclaimer: The Six Five Webcast is for information and entertainment purposes only. Over the course of this webcast, we may talk about companies that are publicly traded and we may even reference that fact and their equity share price, but please do not take anything that we say as a recommendation about what you should do with your investment dollars. We are not investment advisors and we ask that you do not treat us as such.
Transcript:
Daniel Newman: Hey, everybody. It’s Friday. We are back with another episode of The Six Five podcast. Patrick Moorhead, how are you doing? Good morning to you. Looking sharp. Are you in bed still?
Patrick Moorhead: Thanks.
Daniel Newman: You in bed still?
Patrick Moorhead: Thanks, buddy. Yeah, up at 5:00 AM. I decided to do the pod from the home office. Got a flight out. Going to see my dad. I’m doing a keynote speech in Miami on Monday. It’s go, go, go, buddy.
Daniel Newman: Yeah. It was a go, go week, huge week, in fact. I made it back before the weather hit the East Coast, but I spent three days in New York this week. Very excited to not travel next week. Do want to put a little shout-out to my mama in the hospital, having some complications. Everybody out there, say a little prayer for her. She’s not doing super well this week. I know, Pat, you’re off to see your dad. I think it’s the same thing on your end. It’s getting every moment we can with these folks.
Patrick Moorhead: Exactly.
Daniel Newman: The days go slow and the years go fast. And all of a sudden, we look up, and I’m the old person that I used to see when I thought my mom was an old person, and now she’s quite a bit older.
Patrick Moorhead: Dan, you want to hear old? I am probably going to be a grandparent next week.
Daniel Newman: Oh, man. You hear that, everybody? You hear that, everybody? Grandpa Pat, I mean, so now he can be Grandpa Six-Pack when he gets that topless pick. And, by the way, everybody, I think it’s important. We’re going to run a poll here on The Six Five of how many of you want to see Pat doing shirtless photos on Twitter, on X? Do we reinstate the poor kid that got fired from DOGE for being a moron when he was younger, or do we take shirtless pics of Pat or both? That’s an A, B, or C option for you today.
Patrick Moorhead: Oh, and think we threw around the idea of doing a shirtless show, so-
Daniel Newman: Shirtless show? Shirtless show, the shirtless Six Five. We could call the show the Six-Pack, the Six-Pack with Pat and Fatty. Anyway, great to have everybody with us. As we know, it’s been a big week in Wall Street with earnings. This year, Pat, has just absolutely been raging between coming straight out the gate with CES, then Stargate and Davos, then DeepSeek, then the big tech earnings wave. And, I mean, I don’t know if it’s going to slow down, but I think I told you. I said, “I think I’ve done 18 TV appearances.” I mean, that’s probably usually two or three months of normal sort of appearances.
Patrick Moorhead: Crazy.
Daniel Newman: You’ve been all over the media and the tube, and, of course, this stuff isn’t trivial. These are massive inflections right now about what the future of AI looks like, what the future of enterprise spending, government, and, of course, we didn’t even mention tariffs and stuff. Well, you and I are not economic advisors or economic policy advisors. The tech economic policy and all this stuff does kind of … It all converges upon itself. And so, whether it’s where semis ship, the materials required to build them, the energy required to run infrastructure, the impact that tariffs might have on corporate spending, we do see all this stuff come together, which it makes a lot of fun for us, but-
Patrick Moorhead: Totally.
Daniel Newman: … I won’t bore everybody with all the details, but let’s just say it’s been a ripping, roaring year, and we got a lot to cover this week, Pat. So, it’s going to be another one of those sort of extended edition shows. We’re going to try to keep you all on time, get it done in 50 minutes. We do The Six Five, but it’s never actually five minutes a topic. We’re going to cover Qualcomm, Arm, AMD, Amazon, Alphabet, and IBM Investor Day as well as Coherent earnings, so almost all earnings and definitely all investor related stuff. I love this crap. I know sometimes, Pat, I got to pull you by the hair. I know you love it too, deep down. I know you do, but we’re going to cover all that today. Now, a quick reminder for everybody out there. This show is for information and entertainment purposes only, but we will be talking about publicly traded companies. We’re not here, licensed to give you investment advice. Don’t buy any stocks because of what we say. Do your own diligence is what we would recommend. That’s the only advice I give. Do your own damn diligence, fools. Okay. All right. So, let’s see. Pat, the sprint is on. Let’s start. Let’s start with … Draw it out of a hat. No, I’m just kidding. We always know this in advance. Let’s go, Qualcomm.
Patrick Moorhead: Yeah. So, Qualcomm, Net-Net, I mean it’s just a bunch of records, right?
Daniel Newman: Ooh.
Patrick Moorhead: Record revenue, record non-GAAP EPS, record handset, record automotive, and with a really good guide, and then Wall Street rewarded them with a sell-off, right? And, my gosh, it took me so long to figure out what was going on, but, really, it was two things. No, there were actually related, right? It was sequential decline in QCT handsets due to seasonality in Apple, and then Skyworks laid an egg related to Apple, and then, essentially, they just got completely caught up in that. And, unfortunately, all the great accomplishments they had, investors forgot about. I would’ve thought for sure, given their guide, because what we’ve seen so far the previous week is that all these companies had really good quarterly earnings, and then they had this shaky guide, right, but Qualcomm came in with a really, really good guide.
So, handset growth, really driven by the new Snapdragon 8 Elite platform, and, essentially, this is the chip that outperforms Apple’s Bionic on most tests out there. Samsung just brought out their latest S25. They did a full sweep there, and then a lot … Obviously, automotive is still cranking, and automotive was up 61%. IoT was up 36%, and that’s pretty staggering numbers when you pull back on it. Some of the questions, right, that came up during the analyst call that I listened to, QTL guidance. I know there were some chatter about, “Hey, we thought that would be bigger. Why haven’t you guys closed?” Huawei, pretty used to that stuff. One really, I thought, exciting announcements that Qualcomm made on Snapdragon PCs were the market share numbers. And, essentially, it said that they had 10% Copilot+ PCs in US retail, and I think that’s really good. I think the biggest they’d ever hit was one or two percent, so this was a big number. The other thing was that they did this at the same time that the Microsoft Windows goodies, most of the goodies that people were looking forward to hadn’t even shipped. So, all in all, really solid earnings, and if it weren’t for catching the Apple cold related to Skyworks, I think the company’s stock price would have been a lot higher.
Daniel Newman: Yeah, Pat. I think you hit a couple good points. You dove into the details. I continue to zoom out at the Qualcomm business, and you look. Record was a QCT 10 billion. That’s the non-licensing part of the business. For a long time, those numbers between licensing and the technology were a lot closer. The diversification did not span beyond mobile and handsets. And during the time since CEO Cristiano Amon took charge, after we kind of got through this period of time with all the lawsuits, every FTC on the planet trying to take down Qualcomm, you had Apple. Then you had the Broadcom takeover, and it was a little history lesson, but the history was the company had a lot of sort of headwinds that were slowing its innovation and its ability. Well, when Cristiano came in, he’s much more product centric than CEOs of the past that were much more kind of the legal, law related. Mollenkopf was definitely focused on getting through that period of time. Well, Cristiano came in and said, “Look, we don’t want to just be a handset company. We want this connected, intelligent edge,” was what he called it, but what he really meant by this is, “We’re going to go from handsets to handsets to cars to PCs to IoT and VR, AR, XR,” so the product diversification. And so, when you look at the business, right, it’s like are they on the right trajectory? If I’m an investor, is this where I would want to be looking at a company in terms of its direction?
Patrick Moorhead: Yeah.
Daniel Newman: It’s pivoted off some of its dependencies on Apple. Of course, when you look at the Apple number, you got to remember. Qualcomm still makes money on that too, and this is the cool part about the business with the patents and licensing is even if no Qualcomm technology is actually in the phone, there’s still licensing that would drive certain amounts of revenue attached to the continued sale of iPhones. That’s what makes businesses like this pretty cool. But look at automotive, Pat. We didn’t really talk about it, but it’s running really close to a billion a quarter in revenue now. We’ve talked a lot about that 50 billion dollar pipeline, but so, you would say, have they successfully pivoted to automotive? I would actually argue that they sort of took the moment where NVIDIA went all in on data center, and they sort of took a lot of that sort of momentum they had in automotive. Now, I know NVIDIA’s going to get back focused on that again now, but that was a really big move. They took a lot of design pipeline. They’re driving a lot of revenue there. They got into this PC space. You mentioned the 10% of retail, but they’re making good product. I mean, it’s a good, competitive product, and I think that third competitive product in the market’s really important. And to your point, the actual, real, best use cases for an AI PC are still really pretty nascent. So, now, what we have is lower power, higher performing, very lightweight, good-looking devices that have the opportunity now to add a lot of new revenue streams. So, when I look at the overall business for Qualcomm, I would say they’ve succeeded in the mission, and the mission was not to be dependent 100% on handsets. Also, one last thing I’ll note is they have kind of an interesting China outlook.
Patrick Moorhead: Yeah.
Daniel Newman: They’re very dependent on China, but they also benefit from the fact that all the different handset makers use Qualcomm technology.
Patrick Moorhead: That’s right.
Daniel Newman: And so, if there’s tariffs and a ban, and it becomes more nationalistic, and the Oppos and Xiaomis become the bigger volume, and Apple actually moves out, that could actually be an advantage to Qualcomm-
Patrick Moorhead: Totally.
Daniel Newman: … if Apple actually continues to seed market share in China. And, by the way, then you add the Samsungs and others that … They don’t ship from here. They’re not US products, and they’re not shipping, so they’re, what, out of Vietnam? So, anyway, the point is there’s a kind of this interesting tailwind they could come from a … what most people are seeing as a more difficult macro situation. So, I look at all that, and I say it’s a pretty good time to be invested in Qualcomm, and the company’s made a lot of progress. Let’s move on to another company in the same, sometimes, vein, definitely mentioned in the same sentence sometimes. I’ll take this. I’ll put the grenade in my hand, and I’ll talk about this for a minute, but Arm had a really good quarter. The company’s revenue rose almost 20%, 983 million, so just under a billion. Now, as a bit of a context there, this is a company that was doing about a billion a year a few years ago, and now they’re doing a billion a quarter. I would like growth like that. I mean, it’s a little bit slower than Futurum, but, I mean, I’m happy at that level. They’re making more money. This has been a big focus of the company. One of the big challenges for Arm along the way was it had this really high volume royalty business, but it actually wasn’t super profitable.
Patrick Moorhead: Yeah, yeah.
Daniel Newman: And the profitability gap was because they were basically living their life in this very low power, very low
Patrick Moorhead: IoT, for the most part.
Daniel Newman: What’s that?
Patrick Moorhead: IoT space, a lot of those
Daniel Newman: Basically connected devices, very low cost, low per unit, so they were doing billions of units at very pennies kind of thing, and it was driving a lot of revenue, but it wasn’t super lucrative. And so, now, you’ve got the company growing at a great clip. It’s beating. It’s raising/meeting its guidance. It’s part of this Stargate project. Clearly, we’ve seen over the last week, between Meta, Alphabet, Amazon, Microsoft, that the whole CapEx is over because of DeepSeek was absolute, positive horse crap.
Patrick Moorhead: Nonsense.
Daniel Newman: God, I don’t even know. I need to put a sizzle reel together how right I was. I’m going to put that out there. You should do the same. I mean, like I said, I actually think we’re going to go faster with AI. DeepSeek taught some things. I mean, it’s not an all negative situation. It was just people’s ability to see that lowering costs wouldn’t grow, thinking that it meant less growth.
Patrick Moorhead: Well, they thought the entire market was training, which there was a lot of training going on, but every time you even had to run those training … So, the workloads that were trained, that inference was very costly.
Daniel Newman: Yeah. And, by the way, now we can do more inference across more applications, consumer to enterprise at a level of scale. You bring down the cost, the resale value. You got margins. It pushes out from sort of core to more industry and enterprise use cases. Money can start to be made by handset makers and device makers as well as by app makers. There’s just lots of different places to make money. Trillions of concurrent inferences, agents and assistants all running at the same time. Someone’s going to monetize this, and in this case, finally, it won’t be just NVIDIA. So, that’s a good thing, but NVIDIA … By the way, a lot of that CapEx money, still going to go to NVIDIA.
Patrick Moorhead: Absolutely.
Daniel Newman: I mean, I like Broadcom, Marvell. Others are going to do fine, but there’s a lot of dollars to go to NVIDIA, and I think the market sort of recovered. It was … been really interesting few weeks in NVIDIA stock. We talk enough about them. We’ll come back to them at some other point. The other thing I’ll point out is the company’s diversification from royalty to licensing is looking really good. It’s a little under 50% licensing, but that’s all about the company’s done really well in areas like data center and automotive, the higher end PC servers. So, areas like Amazon, so you saw Amazon’s big CapEx number. You see what they’re doing there, but you also see that Amazon … I think the number’s something like more than half now of their compute, a server compute. And a large chunk of it, and going forward, if that number keeps growing, is Graviton. That’s Arm.
So, Arm’s doing really well there. So, that’s a for instance. And, by the way, Apple, if you think they’re doing well, that’s Arm. Qualcomm, there’s Arm technology there. All of this, this is where Arm is doing really, really well. I’m optimistic about it. I think the company’s ability to diversify, and what I keep pushing on when I talk to … You and I both got to talk to Rene, I think around earnings here, is the ability to continue to show the market they can get more margin by this more custom work, the CSS work they’re doing, the customization work that they’re doing. That’s where you diversify away from that kind of high volume, low margin royalty revenue.
Patrick Moorhead: Yeah. So, if I look at just the raw numbers, I mean, they’ve had a consistent streak of a double beat and raises, and if I look at the revenue beat, was in line with the other quarters, right? It wasn’t the highest quarter. It wasn’t the lowest quarter of the last four. EPS basis, right, they beat between 17 and 15%, but it was really about the guide, and it’s almost a carbon copy. Arm has had similar degrees of performance, and there’s a giant, giant sell-off. In fact, I’d predicted that going in in some of my show notes that I had sent to Yahoo Finance and CNBC, and sure enough, it happened. I think a lot of that has to do with the valuation. Stuff that’s just not being factored in here right now are things like Stargate, right? They do appear as if they are the CPU of choice. They were the only CPU that was listed along with the Stargate announcements. Probably the only thing that could indicate that they don’t get 100% of that is if there’s outside financing from somebody related to an Intel or an AMD or something like that. So, that’s not even priced in here for the opportunity.
A couple things happened with the guide. Well, in terms of kind of an EPS, people wanted a bigger number, and if you go and look at their forecasts and what they’ve been spending, they’re spending a lot on R&D, right? And I know that’s not a great thing for the stock market, but it’s a great thing for stuff like me, things that I think are important. I think increased R&D keeps that pipeline going, and I think in a year, when we see what’s cooking with Stargate and new opportunities, we’ll see if the payoff was there. There were a couple analysts, too, who were kind of picking on the percentage of royalties for v9. They thought it didn’t grow faster, meaning it was 25% in the December quarter, the same level as the prior quarter. There were a couple questions, even on the call, that came up from … brought it up, and then there were a couple notes from New Street as well that related that, but to me, not a … I don’t know, something to be expected, just kind of weird.
Final thing I’ll reiterate is that you’re seeing increased penetration of Arm-based hyperscaler servers, right? Whether it’s AWS Graviton, whether it’s Google, or even Microsoft, those are all going to be going up and to the right. Microsoft and Google will be at a much steeper percentage pace, but the mass right now is Graviton, where you have 50% of their capacity, increased capacity, for the past, I think, year, year and a half or so, has been going to Graviton. And knowing AWS, they’re just going to keep putting the pedal to the metal, and we’re going to talk about them and their hundred billion dollar CapEx. Arm is going to benefit from that as well. Kind of a dark-horse year, so what is Meta going to do, right? Meta hasn’t shown up. They’re a top five hyperscaler. They haven’t even showed up with their Arm-based server part yet. Will it happen? I don’t know. Big relationship between Meta and-
Daniel Newman: AMD?
Patrick Moorhead: Sorry, Meta and AMD, and then Meta … It’s funny. I was going to call them Avago and Broadcom, so who knows where that custom processor business is going to go?
Daniel Newman: Yeah. That’s really interesting. And I didn’t say this, but this feels like one of those quarters where the market really feels that a lot of these companies are priced super well. And the reason that is is because good results are actually leading to selling, which means people were sort of buying into the earnings. They were seeing the momentum, but the pricing was … only had more upside if these companies absolutely destroy the results or if your name’s Palantir because now, Palantir’s worth more than AMD. It’s catching up to some really big companies. Palantir’s bigger than three Intels at this point, but that’s just the world that we live in sometimes. I’m not saying they’re not a great company. It’s just super interesting to see how that … Sometimes the market can be completely impossible to read. Pat, let’s jump into AMD then. Yeah. So, this one, there’s a lot to cover here, and it’s another one that had a very interesting response.
Patrick Moorhead: Yeah. Beat on top line, beat on EPS, the beat on revenue was in line with the previous three quarters on average. The beat on EPS, lower right from March to June to September and December, but they had an absolute smashing quarter, basically records everywhere, record client, record data center, and then a colossal sell-off that continues to this morning. What could possibly have done it? It’s very simple. In the previous 2024, they had made an Instinct forecast. It went from a hundred million dollars to exceeding five billion, and the company has decided not to forecast that number for the year, and that just sent off red flags after red flags, even though they are actually pulling in the MI350. There were also comments that said that the Instinct number was going to be back-end loaded, which, again, people look as a bigger risk here. So, AMD, I understand why they don’t want to constantly and chronically giving that number, commenting on that number, particularly when it’s back-end loaded, but it’s super lumpy with a ton of dependencies, right? Dependencies meaning things like how MI350 actually shows up, right? I mean, the silicon is up, a couple key things from the cloud, MI350 running large-scale LLMs within the first 24 hours of receiving silicon, validation ahead of schedule. They are planning to sample lead customers this quarter.
Another positive thing out of MI350, net new hyperscale customers in preparation for at-scale MI350 deployment. So, that has got to be AWS, or it has to be Google, right, unless … I mean, because they’re already in at Oracle and Meta and Microsoft. So, unfortunately it just shadowed, overshadowed everything, like 50% hyperscaler share, and my assumption was in Meta, Google, Microsoft, and maybe AWS, really a trifecta for EPYC and Instinct, right? Increase adoption across cloud, enterprise, and supercomputing, and that hasn’t been the case that all three were up into the right. Client has fourth straight market share gain quarter, and you don’t hear this often. We are going to gain more, right? Quote, unquote, “We believe we can grow client segment revenue well ahead of the market.” That is gaining market share. And what else here? Yeah. I think those were really the highlights. So, the company, just amazing figures for the year and for the quarter. I mean, data center was up 69%. Profit was double in data center. Client up 58%. Profit was 10X of where it was a year ago. Shout out to Jack Huynh and what he has been doing to the business there.
Daniel Newman: Yeah. So, the issue here and the kind of falling knife that has been AMD, I think, has a lot less to do with really anything that makes sense and a little bit more to do with the longer term sort of viewpoint on this AI future. Why do I say that? Because, to your point, you got to go down the list and look at the growth, and then you kind of look at how its forward PE score has come down substantially. So, it’s trading less even though it keeps growing, and why is it doing that? I don’t know how else to say this, but not being able to forecast Instinct was just one of those things that there was just absolutely no way the market would forgive.
Patrick Moorhead: Yep.
Daniel Newman: Lisa knew this. They had to know this. They had to know that this was going to be one of those step back to go forward. I think the problem was is it’s kind of a pay now or pay later thing. If you get too aggressive and you miss, you’re going to get hammered all year long.
Patrick Moorhead: Yep.
Daniel Newman: If you come out now, you’re going to sort of get the penalty right now, and then, if you can sort of exceed expectations, people will see that as the opportunism. It’ll start to push it forward. You kind of mentioned this, I think, somewhere along the lines that, clearly, there are just things people do and don’t care about. Now, not to switch to Intel for a minute, but when they basically canceled the first GPU, the Falcon Shores, you would’ve thought, “Oh my God. That’s a huge thing,” and the … It would’ve sold. The stock went up, and why? It’s because nobody believed that product was worth a crap. I mean, nobody believed that was going to be a competitive part, and so, they were actually favorable to the idea of, “Just cancel it. We’ll wait.” I mean, the stock’s already trading at book value. Let’s see if you can come out with something better if you actually don’t rush it to market.
Patrick Moorhead: Yeah.
Daniel Newman: In this case, it’s just a situation where, look, all the great work being done with EPYC, it’s not to say that that’s not great work. It’s just I think people have sort of largely put into their spreadsheet that the future of data centers isn’t CPUs.
Patrick Moorhead: Yeah.
Daniel Newman: And it’s going to be more accelerators and AI. So, seeing EPYC do well and seeing Instinct not get a forecast probably came with a certain amount of penalty points. What’s that?
Patrick Moorhead: Sorry. I was saying goodbye to my wife.
Daniel Newman: Oh, bye.
Patrick Moorhead: I thought I was on mute.
Daniel Newman: Bye, Paula. Anyways, no, I didn’t hear you. I just saw your lips moving. Anyways, and then PCs, I think as a whole, we know that that market just doesn’t trade with a ton of premium. PC is a exciting space, but it’s not a … It’s not the data center market. It doesn’t come with those margins. It’s super competitive. The fight for market shares there, everybody knows that Intel’s getting super aggressive. So, even if AMD’s doing well, they’re going to have to fight with a more aggressive Intel that’s basically fighting for its life right now in terms of maintaining its market share. And then you got new entrants in Qualcomm. And then you got areas like embedded that tend to be higher margin but have been slower over the last few quarters. And then, so, what does it all come back to, Pat? It all comes back to Instinct. And so, my take on it was no Instinct number was a leading indicator of a lack of clarity or optimism. It’s one or the other. And so, my guess is it’s more clarity than optimism. I think that there’s a lot of reason. I think you and I listened to Mark Papermaster with Oracle talking about software. They did get zero to five billion really quickly. I think part of it’s like can you realize how amazing that is to go from zero to five billion
Patrick Moorhead: I know.
Daniel Newman: The problem is NVIDIA went from zero to a hundred billion in forecast in a very similar period of time.
Patrick Moorhead: Kind of, yeah.
Daniel Newman: It was very small if you go back two years. So, I’m just saying, I think that’s the comparison. Whether it’s a right comparison or a wrong comparison, that’s the comparison that most people are using in their head. And then
Patrick Moorhead: No, I’m going to do … I’m just-
Daniel Newman: Yeah, yeah.
Patrick Moorhead: NVIDIA wasn’t zero. They’ve been doing-
Daniel Newman: No, no, no, no. You’re right. You’re right. You’re right.
Patrick Moorhead: … crypto, right?
Daniel Newman: It had a few billion dollars in that space, but it was smaller than its gaming business not that long ago. It wasn’t very long ago.
Patrick Moorhead: Oh, 100%, 100%.
Daniel Newman: And the use case didn’t really exist in the current makeup of the use case, but the bigger point is just this is where there’s a lot of competitive pressures. I think you and I would both answer that they are the clear number two in terms of flexible GPU architectures for big AI applications. The question isn’t just about NVIDIA though. I mean, NVIDIA’s fighting for its own relevancy in this space right now. It’s about Marvell and Broadcom and hyperscalers building custom chips and Grok and SambaNova and these very specialized inference chips. And all these things that are going on is uncertainty, but the one thing that we know for sure is that the market’s actually getting bigger. So, even after DeepSeek and six million instead of six billion, there’s a lot of growth here. So, I think, for AMD, my recommendation is if they can nail a meaningful competitive software stack, the hardware will follow. I think it’s the problem is trying to move the hardware faster without the software will continue to provide it a challenge. All right. By the way, five billion, not a big, big challenge. They’re clearly doing pretty well.
Patrick Moorhead: Yeah.
Daniel Newman: Let’s move to Amazon, Pat. So, lot to cover off on Amazon. You and I are less commerce folks, but this is interesting. The company now has three 10 billion plus a quarter businesses now. It’s not just a cloud and commerce. It’s also cloud, commerce, and advertising business. And, by the way, they’re going to start shooting rockets into the sky soon, and they’re going to have a fourth business that I expect that will compete with SpaceX that’s going to also make money.
Patrick Moorhead: Kuiper, right?
Daniel Newman: What’s that?
Patrick Moorhead: Kuiper-
Daniel Newman: Kuiper.
Patrick Moorhead: … led up by Panos Panay.
Daniel Newman: Kuiper. Anyways, feels like a Street Fighter character or something, Kuiper. Anyways, so-
Patrick Moorhead: Kuiper.
Daniel Newman: … the company, double-digit growth, 187 billion in the quarter. It beat its earnings by a long shot. It guided sort of conservatively, and, by the way, this is why I think everything’s been selling off. I didn’t specifically say this, but everything’s been selling off. Some of it’s been the guide. There’s been a little conservatism in the guides from these companies. I think everybody’s still a little unsure. I mean, we saw those tariff … Wasn’t it only the Monday of a week ago that tariffs were going to take our economy to zero? The market had gone down 10% overnight. Crypto sold off 35%. It seems like every quarter, we get one of those days, the Japan carry unwind thing, the DeepSeek thing. It’s becoming more and more frequent. But, anyways, the revenue is growing very, very sharply. AWS still grew 19%. Now, that means, I always have to say this, that at 30, 31% for Azure and Google, that I … There’s probably some share shift that’s going on. There’s also a law of large numbers thing with AWS. It is so big that growing 20%, you’re talking about over a hundred billion dollar run rate now, think 107 over the last four-
Patrick Moorhead: Yeah. Jassy claimed 115 billion run rate.
Daniel Newman: Okay, 107-
Patrick Moorhead: That’s crazy.
Daniel Newman: … I think, over the last four quarters, maybe, and 115 run rate.
Patrick Moorhead: Yeah.
Daniel Newman: And so, that’s really, really good. The company did come out, Pat, and say they’re going to spend over a hundred billion in CapEx, so-
Patrick Moorhead: And that’s the biggest number we’ve heard so far.
Daniel Newman: And then they made some other strong announcements. This is the company that’s really kind of taking the challenge of infrastructure on the nose. They’ve been really successful at Graviton. I think they’re trying to double down and really get utilization of Trainium2 to hit the market. They threw out a number of about 40% more efficiency. The only thing that then … the current Hopper, I believe, so I think that’s versus H200 maybe. I’ve been trying to make sure I’m 100% accurate on unpacking that one. This is another situation where the efficiency’s great, but software is going to be a big factor is Trainium not the same as a GPU in terms of being able to be flexible for all the use cases, but for specific use cases, this is exactly what we’ve been saying, higher performance out of a lower cost of … and power envelope.
So, that was something to be optimistic about. How much of that 115 billion ends up where? That’s a little bit of the assessment that people need to do because a lot of people, I think, make the assessment that 115, that’s all going to NVIDIA. It’s like, no, there’s a lot of parts and pieces in there. You got to buy the high bandwidth memory. You got to buy the CPUs. You got to buy all the networking gear. You got to buy the power and conditioning gear. I mean, in some cases, this is physical building sites running in cables, power, cooling technologies. All that stuff is part of the spending numbers, but there is big dollars being spent. Every one of these companies is indicating they have clear line of sight to enough revenue to justify this type of CapEx expense. The other thing, though, is you have to look at these companies’ balance sheets. What else are they going to spend the money on? I mean, these companies that have trillions and high hundreds of billions and trillions, this is the big bet. This is the future. They have to invest big. They have to invest out in front of it, and don’t be surprised that the average person, the average analyst investor cannot see the future. I would bet on the Musks and the Jassys and the Satyas to have a better understanding of the revenue future than the people that are working at banks that are set to evaluate them.
Patrick Moorhead: Yeah.
Daniel Newman: And so, there is a little bit of that. I keep going back to trillions of agentic and assistants operating concurrently, all driving pennies. As I said in the movie Office Space, Pat fractions of a cent rounded off into an account with lots and lots of transactions . I think that’s what the future looks like. I’m very optimistic about Amazon’s business. And on a positive note on commerce, Pat, they did overtake Walmart this quarter, so biggest in the world now.
Patrick Moorhead: Dan, great discussion. I’m going to hone in on a … You left me, did leave me a little bit of oxygen, but-
Daniel Newman: Big company.
Patrick Moorhead: … I mean, this company beat on EPS by 25%, okay, which it wasn’t a decline. In fact, it was the largest beat in four quarters. So, and then revenue, if I look at the average of their beats, they missed in June, and they beat in March and September. It was right along that average. So, yeah, there’s kind of this reckoning. I did a lot of digging on this AWS number. I went through the transcript, and there was a tremendous conversation about supply chains constraints for AWS, and I got to tell … Oh, and, hey, we’re going to get good in the second half. Isn’t that crazy? Microsoft talked about the second half, and we’re going to talk about Google and the challenges they had. AWS said, “Better in the second half.” The funny part is nobody has said GPUs yet, not a single person. What Andy Jassy did talk about a lot was even cooling apparatus, right, which, by the way, could be related to this rumor about NVIDIA GPUs and cooling systems out of these ODMs. So, here we are. We have three people talking about the supply chain issues and two saying they’re going to get healthy in the second half. The only thing I can think about is GPUs and cooling, but nobody has said the word “GPU.” So, kind of some investigation I’m going to do. I’m going to do my specific channel checks and see if some people will tell me directly. And, by the way
Daniel Newman: What do you think, What’s your gut?
Patrick Moorhead: What’s that?
Daniel Newman: What’s your gut?
Patrick Moorhead: My gut is that it’s GPUs and everything else to make a GPU server, cooling … Also noticed that Lisa Su talked about back-half loaded, right? So, there’s something about something that is going to make this back-half loaded that has to do with AI because the capacity constraints were pointed out by Andy that this was an AI challenge, but with all that said, this company has the best array of silicon at scale in-house and also merchant silicon. Their CapEx is staggering, folks, a hundred billion dollars. What is that? Was that twice what Google put out there?
Daniel Newman: No, I think Google said 75, but they were like 110.
Patrick Moorhead: Okay. Sorry, and then-
Daniel Newman: No, no, I’m just saying. I think it was 65 Meta, 75 for Microsoft and Google.
Patrick Moorhead: Microsoft was 80-
Daniel Newman: 80 or 75 there.
Patrick Moorhead: … 80 billion dollars, so-
Daniel Newman: It’s huge.
Patrick Moorhead: … and if you look at where they have Trainium and Trainium2, Inferentia, the homegrown silicon they have, they’ve got a pretty wide array to get there, but listen. Action item for Pat and his team, figure out what this second half bottleneck is, but all in all, I mean, we’re looking at 115 billion dollar AWS business that is incredibly, ridiculously awesome profits, and that has not subsided. And when you’ve got growth, big growth on big numbers with big margins that are going to continue with increased investment, I think they’re going to do just fine.
Daniel Newman: We didn’t even really mention, but these companies, too, have this amazing customer zero story. I mean, Amazon, all this AI is amazing for commerce advertising. It’s not just to sell instances in the cloud, Pat. I mean, all this infrastructure they build out, and I mean, Meta is probably the best story in this particular case, but I mean, there’s a lot of that going on with these companies. Google and its search business, they can build things. They can use them, make money off them, and then also resell them, so it’s pretty powerful. I tend to think the sell-offs are more about perfect pricing and a little bit of uncertainty. Let’s jump to Alphabet.
Patrick Moorhead: Yeah. So, Alphabet, miss on revenue, the first miss on revenue in a very long time, at least four quarters. They had an EPS beat that was very, very small, a beat of about one percent. Previous quarters were 14.7, 2.5 and 25.6. The interesting thing though, I think what rattled people was the Google Cloud number, and isn’t it wild that Google Cloud can’t make the stock rock and roll, yet it comes in, and they miss on that, and the entire stock craters? So, I went in, checked the transcript, and sure enough, capacity issues, right, issues in operations, very similar story, as we’ve discussed, with AWS and Microsoft. One thing that I want to point out though is … were some of the big wins that Google Cloud had. So, first and foremost, big increase in OpInc dollars and OpInc percent, huge acceleration of Vertex AI, which is an end to end AI platform. Check out my callouts on X that I pulled from the transcript. And also, last but not least, Google Cloud is outpacing market growth on a percentage basis and hence gaining market share.
Daniel Newman: Yeah. You and I always have to put that little caveat of inconsistent reporting that always makes it a little bit hard to know what’s kind of banged into these cloud numbers that they all use, but, I mean, I think over a period of time, when you see the delta of mid-teens to low-twenties versus high-twenties to mid-thirties that you have to imagine that there is some tightening of market share. I’ve said a lot that AI was a big reset moment. I mean, I still think Google and Alphabet, but Google specifically, as it relates to the cloud, has done really good in terms of this era and being in a good position in this AI era. I think some very, very good moves have been made.
I think Vertex has been a compelling product. I think there’s sort of provenance in the AI space with Brain and DeepMind. It gave the company a bit of a permission early on to come in and show they’ve been to customer zero with TPU. They spent almost a decade building AI accelerators for their own use cases. They really, barely started to make that any sort of meaningful revenue instance, but I would imagine at some point in the future, that will continue to be a diversification play for the company. I mean, they trained Gemini and Pro on their own silicon, not necessarily winning the market share yet. Gemini’s definitely behind, but I think based on its install base inside of Apple phones, the install base on Android platforms, the install base with my grandmother and everyone that uses Google and doesn’t use LLMs as a search mechanism does give Google … I think they probably have the second-biggest pass on winning early market share besides Apple when it comes to the generative users because I think their sort of core install base is so substantial.
And, by the way, did you see OpenAI is supposed to be raising money at 260 billion valuation? In less than a year, they’ve gained another hundred billion in valuation while continuing to lose money, and someone copied their model for four dollars and six shoestrings a paperclip, but I really cannot understand how this is not … I think Masayoshi Son, I think he’s going to come out that he’s wearing a Bernie Madoff mask. I don’t know if this is a secure kind of racket or what’s going on. I just can’t get to there, but maybe there’s something with AGI that you and I just don’t know about. I mean, they don’t call me-
Patrick Moorhead: I don’t know, man. Let’s talk about all this in nine minutes.
Daniel Newman: Did I get off-topic? I do that a lot. I had a lot of commentary across the web about this. I’m just not spooked at all about this stuff. I think the spend is legitimate. I think there’s a reason these companies are smarter than FinTwit. I apologize. It’s just the case, and let’s get on to IBM’s Investor Day.
Patrick Moorhead: You were there, buddy. You were there live.
Daniel Newman: I know. I know. I know. I know. Just kind of taking a breath.
Patrick Moorhead: You’re just everywhere, dude. Dan’s everywhere.
Daniel Newman: Dude, I talk a lot, so I was just like I needed to … I need to get a little … get my wind back, but yeah. This week on Tuesday, the company brought together the investor community to basically do a victory lap. Unlike all the companies we talked about today, they all sold hard on their earnings. You know who didn’t sell hard on their earnings and actually got a-
Patrick Moorhead: Uh-huh.
Daniel Newman: We talked about this two weeks ago with the super green candle.
Patrick Moorhead: Yep.
Daniel Newman: That was IBM. I mean, okay, when in the world did IBM get the mega green candle of growth after an earnings report? Well, what happens is when you change your forecast from something about a one percent model to a five percent growth model, that’s like a 500% or four or 500% growth, but you’re also … another company that’s doing a zero to five billion dollar revenue victory lap as it’s seen its generative AI business go from zero to five billion in just six quarters. But this was a broad day. Company kind of marched every one of the businesses on stage. Arvind, I think, could feel very happy to know that kind of long before AI was an important topic, like it is today, they changed their kind of whole core business thesis to hybrid cloud and AI. I don’t know if they’re going to have to change it to AI and hybrid cloud at some point, but the ability to kind of say, “Hey, we got it right with Red Hat. Hey, we’ve got it right with small language models,” they showed some really compelling data how, basically, they were able to do small language models with Granite and InstructLab, fine-tuning to get 96% performance approves.
You got to look. The tweets are in the show notes. I don’t have it all. It’s a memory, but much greater efficiency, much more performance, smaller parameters, and they’re showing the ability to do that, which is super impressive because this is what I keep saying. The small language models, the tuning, the forking of bigger models, and using them for industries, this is going to be the scale enterprise use case, and IBM is sitting in a really good position in this particular part of the market. There’s also some other little wins, like talking about the billion dollars of IP revenue related to the research it’s done on quantum. And nobody gives the company any valuation for what it’s done, but it actually has one of the most compelling quantum roadmaps in the marketplace, and they’ve got the market use cases. We don’t talk about this a lot, Pat, but early in any technology transformation cycle, there’s always money to be made in services, and IBM is one of the only technology and services companies that does the full GSI consulting business as well as has a huge technology and software portfolio, and so, that’s really promising. And then the last thing I’ll say is Rob Thomas got on stage and actually showed how their software business had actually exceeded the rule of 40, which I don’t think anybody thinks of IBM software as a rule of 40 company. I think they got to rule of 41. So, that’s a really impressive stat for a company that has a huge software portfolio and doesn’t necessarily have that sort of typical SaaS rocket ship growth, so hit a lot of things right on a lot of different cylinders, Pat. It was a really positive day, and I think that’s something people hadn’t come to expect out of IBM, but I think they’ve really turned a corner under new leadership over the last few years.
Patrick Moorhead: Yeah. I wish I could’ve been there. Had some on-site customer visits that I was pretty committed to. Wish I were there with you. IBM stock in the last six months, it’s up 36%. Last month, right, it’s up 13%. And this is a impact from not only its earnings but also this Investor Day. I don’t think that analysts incorporated the kind of 5X growth between prior year and forecasted year, but I like some of the nuggets, whether it was quantum, whether it was the cost of these SLMs to run. And, Dan, we always talk about the end beneficiaries of this. IBM is an enabler of enterprise, big enterprise, biggest enterprise, particularly regulated AI, and they’re not just going to dump everything into OpenAI and call it a day, right? So, huge opportunity for the company, and kudos to IBM and Arvind, quite frankly, having the vision to turn the company from a services business into a product software company.
Daniel Newman: Yeah, and really quickly, because I want to get my stat right, so we’re doing this Signal65, one of our favorite children, Pat. We’re doing this validation. Some of the numbers they showed about cost efficiency on performance on small models, so Granite’s seven billion plus model plus InstructLab versus ChatGPT-4 Turbo got four percent better performance at 98% lower cost, Pat.
Patrick Moorhead: That’s crazy.
Daniel Newman: And Llama 3 70 Billion got six percent better performance at 75% lower cost. So, these are the efficiencies you and I have been beating the drum on. So, now I want to move to sharks with … I mean, companies with laser beams. You and I had a great conversation in the show notes with Coherent CEO Jim Anderson. They had earnings this week, Pat. Take us home with the rundown.
Patrick Moorhead: Yeah, so listen. A top line beat that was higher than the previous three quarters, an eye-popping beat on EPS by 41%. Previous quarters were 20%, 2%, 25%. Since Jim Anderson came in, actually last six months, the stocks went up 66%.
Daniel Newman: Woo.
Patrick Moorhead: No, I know. Last five days, it’s up 13%, and Jim’s been in there since June of 2024. This is a company that benefits biggest right now on the back of AI growth in the data center. All of this switching that goes on optically, that’s essentially Coherent, not only the transceivers, even the physical chassis that it goes into. Coherent’s in a lot of businesses, right? OLED display manufacturing, they don’t actually just do the manufacturing, but what they help, it basically laser beams to cut OLED wafers to put inside of devices. And here was another cool thing that I thought here. The telco market that has just been in the toilet for, what, two years, two and a half years, Dan, when I look at equipment makers like Nokia and Ericsson and all related companies, but it looks like there is a recovery in Datacom, two sequential increases in revenue. Industrial is not good, still is not good, and I’ve never heard a person say that it was good, and that was very consistent with what we saw from Coherent. So, laser, silicon photonics, 3.2-terabit transceivers on the roadmap, shipping 1.6-terabit transceiver products, sales anticipated in 2025, I like the roadmap. I like where Jim is taking this company, and seeing what he did at AMD and Lattice, I didn’t expect the stock to be up into the right this quickly, but they’re taking advantage of the AI crazy.
Daniel Newman: Yeah, you hit it. This was a bit of a company looking for the moment. I mean, we all know that photonics were going to be a big opportunity as we try to move more data more quickly. Historic mechanisms may work in some cases, but we know in some other cases, we’re going to need to be able to transport data faster. Who are the companies that really do this? I mean, really, where are the guts? We talk about pick and shovels, and a lot of people talk about that. Well, this is like the pick and shovels of the pick and shovels. Whether it’s the modules or supplying the technology, even materials, they kind of do all the different parts there.
You hit it on the head. I mean, industrial’s moving slow. Comms is transforming a bit slower too, but this is a data AI moment, and if you’re talking about the transport of data AI, this is something that Coherent is really well positioned for. They supply a lot of companies you know that are doing silicon photonics, so they do the full module, but they also supply a lot of companies as well. So, they make money a few different ways, and this is sort of Jim’s specialty. He seems to be able to come into these kind of companies that are in the semi-space that have some exciting opportunities, and he seems to know how to really maximize quickly and get them moving in the right direction. It’s also a testament to Jim just as a leader, kind of how his style is, his culture. He’s very hands-on.
Patrick Moorhead: You ever notice he never takes credit for anything and never has?
Daniel Newman: Hmm.
Patrick Moorhead: Not even a hint.
Daniel Newman: No, it’s awesome. Super-
Patrick Moorhead: Not even a hint.
Daniel Newman: … super humble. I heard he’s also very good at driving race cars, but we haven’t had a chance to do that yet, but he also … I love that he went out and … I don’t know. It was like 50 countries and visited people of … There’s some leaders that they come in. They get in their big office. They get their teams and their, there’s other people that who … They’re flying around on planes and sleeping in hotels and living out of a bag, and they’re showing their company that they’re in the trenches and ready to go to war.
Patrick Moorhead: Yeah, executives that come in and tell them what they have permission to do and permission not to do.
Daniel Newman: Yeah.
Patrick Moorhead: It’s rough.
Daniel Newman: They approve POs and different things. Hey, I mean, we have some companies, Pat, all jokes aside, and we have customers and companies that you and I both know that literally still have CEOs. These are multi-billion dollar companies that approve every hire. I mean, there’s various levels of joy-sticking. I think what it comes down to this is Jim’s really found a great balance of being a leader. He’s also very technical. He’s very competent, and you’re seeing it early. All the credit can’t be given to the new guy. There’s had to have been some good work that went into this coming in, but he was definitely the man for the moment, and the company was also well poised for the moment, Pat.
Patrick Moorhead: Yeah. I’m looking forward to their Analyst Day this year, right? They can lay pretty much everything out. That’s when we’ll see the grand plan.
Daniel Newman: Let’s go see the grand plan. All right, everybody. That’s it. The bonus episode, the Seven Five, the Seven Ten, whatever it is we do here, 248 in the books, Pat. I know we did tease that there might be a new format. It is coming, we promise. It’s just been crazy this year.
Patrick Moorhead: We want to make it right. We want to perfect it before we roll this
Daniel Newman: We want it to be perfect rather than good. We’re going to let perfect get in the way of progress. That’s how we do things here at The Six Five. Just kidding, everybody. Thanks so much. Coming soon, some updates on The Six Five Summit as well. That’s coming to an audience, town, computer near you, but for this episode, for this show, for 248, for Patrick Moorhead and myself, subscribe. Be part of our community. We love you all, except those that we don’t. We’ll see you all later. Bye-bye.
Patrick Moorhead: Bye-bye.
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.
