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:
- Intel Q2CY24 Earnings
- Qualcomm Q3FY24 Earnings
- AMD Q2CY24 Earnings
- Arm Q1FY25 Earnings
- Lattice Q2CY24 Earnings
- Microsoft Q4FY24 Earnings
- Amazon Q2 Earnings
For a deeper dive into each topic, please click on the links above. Be sure to subscribe to The Six Five Webcast so you never miss an episode.
Watch the episode here:
Listen to the episode on your favorite streaming platform:
Disclaimer: The Six Five Webcast is for information and entertainment purposes only. Over the course of this webcast, we may talk about companies that are publicly traded and we may even reference that fact and their equity share price, but please do not take anything that we say as a recommendation about what you should do with your investment dollars. We are not investment advisors and we ask that you do not treat us as such.
Transcript:
Patrick Moorhead: Welcome back to The Six Five podcast. It is my favorite thing to do for the week. We are slinging it all over the place. And back by popular demand, we have Stacy Rasgon. Stacy, how are you doing, my friend?
Stacy Rasgon: I’m good, thank you. How are you?
Patrick Moorhead: Good. Dan, how you doing? You look like you’ve had a couple hours of corporate meetings. I know how much you like those.
Daniel Newman: Yeah, look, I start late and I make up for it by leaving early. That’s how I do things at the office. But yeah, I mean look, we’ve got an exciting guest on the show today and I’m glad that you even introduced him before me. I think my goal of replacing myself on the show is even closer. So next week it’s going to be Dan’s out, Stacy’s in. Then in two weeks it’s going to be, the invite’s going to be off my Friday calendar and I’m going to come on and I’m going to be like, it’s Pat and Stacy.
Patrick Moorhead: Yeah, we’re going to go, yeah, I just read left to right Dan, I’m not technically *** but some people tell me that I am. Anyways, with that said, it’s great. Let’s fire this thing up. A little disclaimer here. Dan and Pat are not equity analysts. We are technology industry analysts. We are not professionals. We are not certified or trained. Stacy is though. Stacy, real quick blurb, you’re famous, infamous in some circles. If you wouldn’t mind introducing yourself, who you work for and what you do, that’d be great.
Stacy Rasgon: You bet. So my name is, I’m Stacy Rasgon. I work for a company called Bernstein Research where I cover the US semiconductor and semiconductor capital equipment space, the equities. Some background, so again, I’ve been here for 16 years. I was actually at McKinsey before this. I worked a hundred percent of my time then in their semiconductor practice and I served clients all over the value chain and all over the world. Before that, I did a PhD in chemical engineering at MIT where I actually built semiconductor manufacturing equipment. I did some work at IBM T.J. Watson Research in their advanced lithography and processes group, and so, it’s kind of been an interesting sort of random walk from there to here, but it seems to be working out so far.
Patrick Moorhead: Yeah, appreciate you being on the show. We’re going to talk about surprise earnings. We’re going to talk Intel, Qualcomm, AMD, Arm, Lattice Semi, Microsoft and Amazon. We are The Six Five, but we’re actually The Seven Five this week ’cause how can you not talk Amazon? I mean sure, we could have squeaked Apple in there, but we do have our limits. So let’s jump in here. Intel, Intel, wow, my phone was lighting up. I’m getting DM’d. I think I got 27 LinkedIn invitations from people at Intel. I mean, I got 42 new followers. I mean it was crazy nets, but here’s the net and net and we’re going to throw the red meat into the room and we can discuss it, but good quarter, lousy forecast, giant cost production.
Stacy Rasgon: It wasn’t a great quarter.
Patrick Moorhead: Well, they met revenue. How about that?
Stacy Rasgon: Sort of. Okay, yes.
Patrick Moorhead: Yeah, net-net boils down to confidence at this point. There’s a lot we don’t know and we’ll try to get info on details about Oregon to Ireland transition, Meteor Lake. What changed to drive the 15% layoff numbers, literally changed between last quarter and this quarter. What could be the downstream impact from the layoffs.
Daniel Newman: I’ll let Stacy go first.
Patrick Moorhead: Yeah, yeah, okay. I was just going to get my blurb out, and then…
Daniel Newman: No, no, no, no, no, what I was going to say though, Pat, you can go first, then Stacy, I’ll go last. I just wanted to laugh because I said something on Twitter and I said something along the lines of, had a less than ideal quarter was how I tried to be nice about it. And I got someone that said, is that the understatement of a lifetime that you said that? Because it’s like I’m already sad. I think Stacy, you said in the green room, I’m sad. It’s like, I’m sad. How do I say this without saying this?
Patrick Moorhead: Yeah, but I’m sorry, Dan. This was supposed to be your lead in too. I just got so excited ahead of myself. I’m sorry.
Daniel Newman: No, no, it’s okay. Go ahead, Pat. Finish your run ’cause you were on a nice run, but it was less than ideal for me.
Patrick Moorhead: I mean, in the end, what can we take to the bank now that the company says. And listen, I’m talking to their customers, they’re customer customers now. So Stacy, the floor is yours.
Stacy Rasgon: You bet. It was a less than ideal quarter for sure. So the quarter itself, you’re right. It missed the consensus, but they had actually lowered their number early in the quarter because they lost their export license to Huawei and the street numbers on revenue didn’t quite fully incorporate that. So you could argue the revenue was, they did about 12.8 billion, was kind of in line with that target. The gross margins were very, very bad. They printed 38%, so were back with a three handle on them. And you got to remember that, that happened a little while ago but really the last time that the gross runs out of three of them, I was in elementary school. It was 1986 was the last time. So it’s been a long time. They were 500 basis points below on that. They printed 2 cents. I think the street was at 8 or 10 cents expectations.
So the quarter wasn’t good. And you’re not the best quarter for, Q3, the guidance was horrendous. So they guided about, oh, I don’t know, a billion and a half dollars below the street on revenue. Gross margins, again, they guided like 38, it was like 750 basis points below the street. They guided to a 3 cent loss. It’s nasty. And they even sort of gave some color into Q4. Q4 is, I don’t know, another massive miss versus where people, so those things are really, really bad. There’s a few reasons for it. I mean one is I just think the original guidance, they had said last quarter that they’d get above seasonal growth through the back half. They have this little hockey stick for whatever reason embedded in the numbers, so that’s clearly not happening. There’s some PC/CPU inventory channel flush. This is something I’ve been concerned of.
We tracked this pretty religiously and we saw through the Covid swing, we had a big overshipment and then an undershipment. And I know in Q4 and in Q1, as far as I can tell, the PC/CPUs were back to overshipping by a pretty wide margin. So I’ve been a little worried about channel and that’s flushing out in the back half and so that, there’s some degradation on revenue there. Look, I think they’re still losing share. You can look at what’s implied for their data center growth versus say what their competitors are putting up and they’re clearly losing shares. So on the revenue side it’s not good. And then the margins are really horrible and that’s really where I think a lot of the deep concerns are going on. They’re attributing this to a mix. They’re basically, they’re ramping a new product for AIPCs. It’s called Meteor Lake.
And the early parts are made in Oregon. And then they would typically transition those over to their other fabs, in this case Ireland. And they made a decision, their claim is they made a decision to make that shift early. And they said what that does is it saves them a billion dollars in CapEx. So they think it’s good, and maybe it’s the right decision, but the issue is, early in the run, the costs are really bad. And so it’s massively impacting the margins into Q2 and Q3. And then I also think even as those costs come down, just the cost of that part is lower. I’m sorry, the margin of that part is lower, the cost is higher. A lot of it is outsourced to TSMC. A lot of the tiles, that becomes a bigger piece of the mix, I think, that impact. And then they gave kind of color for next year.
Margins are still going to be really impaired. I bet street numbers came down by five or 600 basis points for gross margins next year. And that’s the Lunar Lake ramp, which again, they’re very bullish. But Lunar Lake is entirely made at TSMC and so again are really bad. So it was not good. And so in response, they’re taking action. So there were some new stories in the days before that they were prepping a pretty big riff. And they are, so they’re laying off 15% of their employees. If I just take the Intel-specific employees, that’d be about over 17,000 employees. So they’re cutting OpEx. They’re cutting CapEx this year and, although it’s hard to say because I think the street CapEx numbers for Intel were already lower, but versus their plan, they’re cutting fairly significant amounts of CapEx. They also suspended their dividend. They had cut their dividend by two thirds a year, year and a half ago, and they’ve fully suspended it now.
So they’re in scramble mode right now, in survival mode to try to right the ship. And then there’s two other things that are going on, and these were already happening, but I think they’re accelerating. They are getting subsidies from the government. There’s the investment tax credit and then eventually the CHIPS Act stuff. And then they’re also getting what they’re calling partner contributions. What they’ve been doing is they’ve been selling parts of some of their factories effectively to private equity. They sold half of the Arizona fab that they’re currently building, the Brookfield, that one’s not operational yet. And then this quarter they sold half of their Ireland fab, this fab 34 to Apollo for $11 billion. They got that cash in Q2. Now they have to give up returns. They said that’s going to cost them, it’s like $700 million next year that they’re going to have to give up to Apollo.
Apollo doesn’t work for free. But they have cash coming in. So that’s what’s going, but I mean in general it was a horrendous, horrendous quarter. And then the issues that you kind of look at, it’s down, I don’t know, 27%, I’m probably off by 20 minutes on this feed, but it’s down a lot. I think the issue people wrestling is like, how do you even value it? There’s no real earnings, there’s no free cash flow, and there may or may not be any terminal value. That’s the question that people are having now. So it’s going to struggle to fly on the floor right now. I think all the funds that held it, like in the dividend funds, it had a small dividend. I mean they’re probably going to have to sell. It’s a mess. It’s a real mess.
Patrick Moorhead: Dan, what do you got buddy?
Daniel Newman: Yeah, so look, I know this is, Stacy, this is a mix like you said of being sad and sort of a victory lap because you’ve probably been, I’m sort of being sardonic about it because you’ve been sort of one of the more notable bears that just has not come around. And some of us, I mean look, I admit when I get things right, in fact I’m the first to always say it. And I also have to sometimes be the one to acknowledge when I get things wrong. And there’s some things I’ve gotten wrong here. I mean, I think on some of the parts, I think we kind of said about executing the say/do ratio thing that Pat likes to talk about. He’s done a lot of things, but it just hasn’t worked yet. It just hasn’t worked in the numbers. I mean the process, the getting things to market, the tape-ins, all the things that he sort of said would happen, have happened. It isn’t panning out, I mean, they’re raising money, they’re leveraging assets, they’re shedding business units. This is like a GE story. This looks like a little bit like what GE started doing and it’s like when do you get to the minimum viable product? Where the market’s going to be like, this is where you’ve…
Stacy Rasgon: I mean, that’s it. That’s exactly, I mean they’re spinning hope, right? It’s like, oh, we’ve got Lunar Lake. It’s a great part, supposedly, they say, but the cost structure is not good. But then when we get to Panther Lake, which kind of launches second half of ’25, so it’s kind of 2026 for volume, that’s where it’s actually on 18A, which is that’s finally the process that they’re saying is really good and we’re going to bring all those tiles back from TSMC. And so we’ll be making them internally on our better, at that point, hopefully we’ll have a better cost structure and so the margins will get better. So that’s the hope. But I mean even if you believe it, I mean it’s 2024 right now. That’s a long time to wait. We may have talked about this last time, I can’t remember, but they had that Foundry Day a little back and they were talking, we’re sort of splitting the economics of the different parts of the business and we’re trying to incentivize better behavior and better choices and transparency, those are all very good things they need to do that, but I think I said last time, my biggest takeaway from that day was come back in 2030.
Daniel Newman: Yeah, I was going to say, I got three quick bullets and then we got to move ’cause we probably could spend the whole show on… Go ahead, Stacy, go ahead.
Stacy Rasgon: I just want to say, there was one silver lining from last night. And maybe it’s in your bullets, but I do want to say, I was talking to Pat in the green room like, I’m having, initially I was having sort of reminisces of AMD back in 2014 where I mean that the investment controversy at that time for AMD was are they going to go bankrupt or not? And I think if this was a different scenario, if we didn’t have the CHIPS Act and all these partner contributions, we’d be having, going concerned conversations here. I think Intel will live. I don’t know what kind of a life it will be, but I think they’ll live for, and the reason is if you sort of add up everything they’re doing. You add up the cost cuts. You add up the CapEx cuts. The other cost savings. You add in the cash that they’re saving from not paying the dividend. And you add in all the subsidies and these partner contributions. It’s like $40 billion of cash, incremental cash on their balance sheet over the next two years. It’s probably 45 or 50 billion over the next three years that they would not have had. So that will I think save them from, they will live. So that’s the one silver line, but the problem is, I don’t know what kind of a life it’s going to be.
Patrick Moorhead: Yeah. Dan what are your three bullets, buddy?
Daniel Newman: Yeah, so I’ll make them real quick ’cause we’ve got some more positive earnings to talk about now that we can all wipe our tears and keep going. But bullet one is the Gaudi thing has to be done. They have to get it right. They have to land. And that one, what I’m saying though, look, you cannot be relevant if you cannot get an AI processor for the data center.
Stacy Rasgon: It’s $500 million this year. Who cares?
Daniel Newman: No, well that’s my point. So I’m not talking about now. I’m talking about at some point that thing has to get competitive.
Stacy Rasgon: Well no, it’s not Gaudi. So they’re talking Falcon Shores.
Patrick Moorhead: Falcon.
Stacy Rasgon: Which is the next one.
Daniel Newman: That’s too far out.
Stacy Rasgon: Well…
Daniel Newman: Literally. I know, I know, I know, I’m just finishing my thoughts though. They got to execute with what they have, because you cannot survive this. You’re right, I’ve actually said, I think I said to somebody, there’s just no thesis right now to go in now. The only thesis is basically extraordinary value with a 10-year horizon. That’s the only thesis I can really think of. The second thing though is they do have the wins and designs on AIPC, whether or not it’s super profitable, they need to saturate the market and show they can protect their market share. That’s the second item. And the third is, the world does need a strong Intel. The TSMC dependency is scary. When you actually break down how much we depend on Taiwan for everything.
Stacy Rasgon: I mean, I’ll say something and maybe it’s controversial, I don’t know, but this whole CHIPS Act thing like picking Intel as the national champion, I understand politically why that happened. But I mean to be fair, if we really wanted to get a ton of leading edge capacity built here in the US, we probably should have given all the money to TSMC and encourage them…
Daniel Newman: Totally should have. Totally should have.
Stacy Rasgon: Build as many fabs as they could have made it as easy as possible.
Daniel Newman: But the thing is they had to pick a US company. That was the only way.
Stacy Rasgon: I understand.
Daniel Newman: And second of all, and I’ll actually say something controversial back to you, Stacy, is they did not, they picked them in words, but with the money they gave out, they weren’t as nearly as clear that they were picking Intel. They gave almost as much money to Samsung, Micron and to TSMC as they did to Intel. So they kind of said blah, blah, blah, Intel, but the money they kind of said, we’re not as confident as we may be, we’re articulating we are.
Stacy Rasgon: Well, and maybe…
Patrick Moorhead: All right guys, let’s move. We got to move.
Daniel Newman: Pat’s the keeper. Pat says go, we go.
Patrick Moorhead: My final word on this is this brought up more questions than answers, but I’m going to be doing my channel checks and I’ve already got people phoning it in to give me information. So let’s go into the next topic and that is Qualcomm, beep, beep, raise. Diversification story here with automotive. Automotive is a juggernaut. Nothing really on PCs, said that it exceeded internal expectations which were appropriate low and some skews were sold out, but it’s a very small percentage. But I do believe the roadmap looks good and I think even looking at OEMs after the Intel news, maybe they’re doing a side eye at Qualcomm. Any impact from any potential AI lift is going to be through this new Orion Core that’ll be announced later this year. Won’t financially impact this year, but if AI in smartphones is going to hit in any year, it’s going to be 2025. And if we see what Intel, sorry, Apple is doing with Apple AI, it is a good portending thing. I think China will actually be leading with the AI smartphone, which I think again, there’s a lot of puts and takes for Qualcomm in China. You’ve got Huawei rising. You have Apple potentially declining. But you’ve got premium smartphones in China going up that does benefit Qualcomm because you get the APU, you get the modem and you get a lot more RF content. Go.
Stacy Rasgon: So Qualcomm, was a good print. The stock reaction was very interesting though. It was up like 8% in the aftermarket and then it was down and then it’s been weak since. And there were a couple of reasons. So the quarter was good. It was a nice solid beat and raise. The guide on the surface was a beep and people were worried because again, they also lost their Huawei export license. So there were two reasons that the stock sold. The first was that that guide actually had an extra week in the quarter. So you get extra revenue and everything. And I think it was misunderstood how much revenue it was, ’cause if you just do it at first glance, it’s 8% of a quarter’s revenue. The 9, they guided 9.9 billion. If you just take a uniform revenue distribution, the real guide without the week would’ve been 9.2, which was very low. They said some stuff on the call, you could actually back that out. It wasn’t that much. It was like $300 million, about a hundred million in licensing, 200 million in chips.
9.6 in my opinion was the sort of normalized guide, which would’ve been fine. The other issue though is they’ve made a practice of the last couple of quarters to give N plus one guidance to guide not just an extra but give some soft guidance to the quarter beyond. And they kind of guided the December quarter softly, but a little below where the street was. And so that’s kind of what, they tend to do this, again, maybe they should stop, but if you sort of look at their prior history, when they eventually get to that quarter, not only do they tend to beat their guide, they also tend to beat where consensus had been before they gave the guide. So I think they’re being conservative. I think it was fine. But I mean that was the issue. The other issues that Qualcomm is having is everybody’s worried about the Apple modem.
It’s potentially going away. They’ve actually told everybody for modeling purposes assume it goes away. People have assumed that’s the end of ’26, but there’s been some news flow that they might lose some small volume skews next year potentially as Apple kind of trials their modem and, I don’t think, it’s not a lot of earnings. I don’t think that the worry there is there. The worry would be that, does it increase the chance that the actual thing actually does go away? For real, because people are thinking that maybe it won’t. So it’s been disappointing. Now at these price levels, I think you’re already pricing Apple out. I could take all of Apple out next year fully and it’s trading in the teens today so. I think you’re kind of pricing it in, but that’s the worry that people have.
It’s like, well they’re guiding down December, I got these Apple headwinds. Maybe there’s no rush. That’s kind of where I think sentiment in the near to medium triple B, be it, just ignoring all the other stuff that’s going on with the macro and everything right now. I think that’s just where sentiment is. But in general, it was a decent print. It was very good. On the product roadmap and everything, they’re killing it. And you mentioned Auto, by the way, the adjacency story’s looking really good. The Auto killed it. Auto was really good. I think their long-term targets there are looking increasingly conservative. At least their industrial IoT business has bottomed. And then we’ll see what PCs do. I mean worst case it’s not a negative for them, right? So we’ll see if Arm PCs are a thing, I don’t know yet, but if they are, they’ll benefit.
Daniel Newman: Yeah. So you guys hit on a lot of the points. I think you and I, Pat, have been beating the diversification drum, and this is where it’s starting to really shine. I mean look, the handset business wasn’t that, it was fine, but I mean all their growth came in strength in automotive and strength in IoT, which are now, they’ve had these two additional, I mean automotive is almost a billion a quarter now of revenue.
Stacy Rasgon: Yeah, 811 million they get.
Daniel Newman: Yeah, 800 something, but I mean it’s 75% roughly up from a year ago. So it is changing the whole trajectory and the whole shape of the company. So now it’s got the potential of these arm PCs, which when we talked, we all probably talked to Christian, I know you and I did, Pat. I talked to him and I thought, I kind of said I’d like to hear more about how it’s going. The way they explained how it’s going was very, it was a little vague. It’s better than we had modeled. Okay, well good, I’m glad to hear that. But what does that really mean? Are people buying these? Are they flying off the shelf? I mean our testing in our lab shows that there’s some really great stuff. Pat, you and I both use them every day. I’ve been encouraged by my experience with them, but is this a supercycle? I mean we keep hearing that term. There’s going to be a supercycle.
Stacy Rasgon: I hate that term.
Daniel Newman: I know, I know, some people love it. But the point is what does that mean and when does it ramp? How excited are people? What value are they getting? You and I have both complained about the Microsoft Recall thing not necessarily being there and that was the exciting feature. So that’s a big one for me. I mean then we were hearing about, by the way, a supercycle on the handsets, right? There’s going to be a supercycle driven by Apple Intelligence and it’s going to make everybody run to the store. And I know you’re cringing, but just, I’m not saying I agree. I’m saying that’s the sort of the street, the bulls of the street are saying…
Stacy Rasgon: Where I get bullish by the way on the AI smartphone side is there’s actually not tons of units and upgrades. It’s content.
Daniel Newman: Yeah, okay.
Stacy Rasgon: I think that’s enough, frankly. If we get an upgrade cycle, great, that’d be wonderful.
Daniel Newman: So it’s more people willing to pay more for the things they’re using on the device. Of course, and look, I’m going to be really clear, AI is not XR. And not to say, actually Qualcomm, part of their strong IoT number was XR, which is rare and anybody’s making money on the VR/XR side, but they actually do well because the Meta new headsets are performing well. But what I’m getting at is AI is a real thing. The biggest thing across the entire industry though, I think people are fighting with is what AI are people willing to pay for and what AI are people just expecting as part of what would be kind of an iterative upgrade. And that could be in software, that could be in hardware, that could be in enterprise, that could be in consumer. It’s like if people pay more for SaaS, I’m supposed to get the next feature. That’s part of why I subscribe to your stuff. How much more am I going to pay for it? Same thing with a phone. Every time I go buy a phone, it’s supposed to have more stuff. That’s why I upgrade. I mean, by the way, I’m still using a 13 pro max because I’ve seen nothing in the last two generations that excites me enough to pay for another phone. And I’m a tech guy, so that’s what we’re up against. But Qualcomm’s numbers, I mean look, compared to where we started this conversation…
Stacy Rasgon: Oh yeah, oh yeah.
Daniel Newman: Steady state, good stuff, margin, operations, management, execution, not a lot to complain about over there. Diversification, looking good.
Patrick Moorhead: Excellent. And let’s dive into AMD. Dan, I’m going to give you the microphone on this one. I keep being the hog.
Daniel Newman: No, it’s okay. It’s okay. I’ll say a couple things on this one because I’m interested and intrigued, I’m hearing everyone’s, this was, I think my comment when AMD hit was I said finally a print that the street likes. There was finally a print that, it was the same day I think as Microsoft went, and everybody was just throwing up on Microsoft. But it was all about CapEx. Now, Satya got on the call, we’ll talk about Satya later, and he kind of recovered a lot of the ground over the course of his call, but AMD, two numbers. So Intel had a 9% growth in client and AMD grew like five times faster in client. So market share gains there, strength in their numbers there. The data center number was huge, and we all know why that is. It’s instinct. I mean Nvidia is the champion, undisputed, but there is a number two and number two in the GPU space is AMD.
And so AMD, while there’s still some questions on ROCm, questions on the software stack, are people going to use this? There is demand for this. And whether it’s been some of these AI clouds that are building up around it, whether it’s some of these, what I would call the hedges that are being made by the hyperscale cloud providers, not the AI cloud, not the Lambda Labs kind of things, but the actual AI cloud, so the AWS’s and the Microsoft, which is where a lot of the AMD is supposed to land, they need an alternative. They really do. The Tom Toms on the streets that I’m hearing is look, everybody knows they have to be, behold it to NVIDIA in many ways. And it’s the champion, undisputed on the product, on the software, on the stack, on the libraries. Having said that, they’re all feeling a little bit of the pinch, the margin control, the deal flow, that they just don’t have enough control.
And so AMD is a very logical second choice. There’s a solid roadmap. Can I say I hate the, I hate the 325 and 50 and just, my mind cannot wrap around those numbers. But there’s a roadmap, there’s clarity of a cadence. So in the data center, there’s an absolute strength there. The growth is there, the numbers are there, the margins are there, the customers seem to be there and there’s no one else. Your point about the 500 million, our market forecast shows about the same thing, 300 million last year, about 500 million on Gaudi this year. And so AMD’s got that momentum and no one else is there to compete with it. So the trajectory they’re showing and they’re on is good. So a lot to like. I won’t even touch gaming, that’s been pretty beat up. But that has more to do with where does that industry go? Where’s that going longer term. So I’m going to pass the mic here. You heard my initial comments.
Stacy Rasgon: Yeah, AMD’s quarter was decent. I will say the stock was about 140 when they printed and it was up after the print. I think if it had been at 180, which is where it was three or four weeks ago, I think it would’ve been down, but it wasn’t a bad print. So the expectation had come down. I think on the good side for them, they’re clearly still doing well in their core client and data center businesses and they’re still clearly taking share. And those were all good. And again, they did take their AI guidance up. They went from more than 4 billion for the year to four and a half. Again, I think three or four weeks ago, four and a half would not have been enough. I think where things are now, four and a half at the time at least was enough. So that was all fine. The negatives, and there were a few. Where it would be, look, on the PC side, again, they are significantly over living versus Intel.
Intel’s calling out channel flush out. Again, I’ve been worried about that. I am worried that AMD’s guidance, both their results as well as their guidance, they’re guiding for above seasonal growth for client in the back half on new products. There may be some incremental channel fill that may be incorporated into that guide. I don’t know yet. It sort of depends on the share interplay, but it’s something to watch, I think. I’d say the is like the embedded business in gaming are quite bad, still. Embedded its kind of bottomed, but the recovery isn’t as good as they thought. So numbers next year actually had to come down because as well as their OpEx going up, although I’m not going to knock them for OpEx again. Compare and contrast with Intel, I think it’s probably good to spend right now if you can. And then finally on the AI guide, the four and a half billion, people believe it’s higher than that. And you have to believe it’s higher because if it’s four and a half, we already know what they did in Q2, you can kind of ballpark what they’re sort of suggesting for Q3.
It doesn’t leave much room for an exit rate in Q4. So you need to believe it’s higher and it puts more pressure on the expectations for next year because I still think expectations, well, again after the blood bath, I don’t know, but expectations were for them to do solidly double digits in AI revenues. And so double-digit building in AI revenues. You need a good exit rate coming out of this year. So those are the things I think to watch. But objectively, by the way, in this set of guides, we’ve had a lot of semi-prints so far. There haven’t been a lot of good ones. This was more on the better side of the spectrum of what we’ve seen I would say so far during the season.
Patrick Moorhead: Yeah, good analysis guys. And I guess it’s interesting, in December I said that their AI number would be 6 billion. Just looking at the puts and the takes.
Stacy Rasgon: I could be wrong, but I feel like if it was going to be six, they could have said that they would do more than five.
Patrick Moorhead: See, that’s not Lisa though. That is not the way that Lisa Su rolls. I think the way that she wants…
Stacy Rasgon: Swinging before, so I don’t know, but you know her better than I do probably.
Patrick Moorhead: Well, it’s just I look at, look at the itty bitty numbers that came out and it’s this trickle in the way that they want to bring that out. I mean, we went from a number in December to January and then end of Q. So anyways, that’s kind of where, I mean, again, I don’t get paid to hit the exact number, but calling that it seemed reasonable to me that they could hit that and the only governor would be TSMC, which is the industry’s governor.
Stacy Rasgon: Yeah, I mean for what it’s worth, we’re modeling, I’m at five this year and nine next year, and I think 11 in 2026. So could be more.
Patrick Moorhead: You’re safe, you’re safe with five I think for sure.
Stacy Rasgon: We’re at five right now for this year.
Patrick Moorhead: Yeah, absolutely. And with all the downstream FOMO from all the hyperscalers, it’s going to happen. So it’s interesting on embedded in gaming, I like to look at self-inflicted wounds and market inflicted wounds and on gaming. I mean it is absolutely a challenge, with the exception, I do believe there will be new consoles that will come out. We’re in the fifth year of that. What’s a mystery to me is, are gaming, the gaming cards where Nvidia is an absolute juggernaut.
Stacy Rasgon: Right? Yeah.
Patrick Moorhead: Juggernaut there and embedded, listen, look at Altera, look at Lattice. They’re both down.
Stacy Rasgon: They were all over earning during Covid and it’s inventories. And look, auto and industrial markets in general right now are not great. Com infrastructure is not great, and they were all over shipping and over earning less so, there’s a bunch you have to work off.
Patrick Moorhead: Exactly.
Stacy Rasgon: It does look like it’s bottom, at least they’re looking for growth in the back half. Again, they weren’t less, they’re kind of suggesting single digit sequential for the rest of the year. I think a quarter ago they were suggesting could be double-digit. Intel’s still suggesting double-digit for Altera right now. Again, I think that’s a mistake.
Patrick Moorhead: Yeah. Final comment that I have on this is that I think AMD needs to be looking very seriously of how they can capitalize on the Intel news. And it used to be with the OEMs that AMD was the risky choice, right? Yeah. Embedded in that is you throttle your volumes and you throttle your volumes and the breadth of this. What AMD needs to really think through is how they can capitalize on this in areas like enterprise PCs and enterprise servers. This is areas where they’ve been traditionally weak and that takes massive marketing and sales investments to get that going.
Stacy Rasgon: By the way, actually talking on the call, on the public call about enterprise server gains, which I know they’ve been talking about forever.
Patrick Moorhead: Yes.
Stacy Rasgon: It sounded like they felt that they were a little more real now and again, I think they said something like a third of our new wind, which is not revenue yet, takes time. They said something like a third of our new winds are new enterprise customers that we’ve never sold to before. So I don’t know how big that is, but they were talking about it more than maybe they usually do.
Patrick Moorhead: Yeah. And what you’ll do is you’ll see that inside of numbers for Dell and HPE and Lenovo. So hey, guys, let’s move on to the next topic. And just like we had talked about, this is not Arm and Lattice or not Stacy’s, what do you call it? Official…
Stacy Rasgon: They’re not under my coverage. We don’t cover, A colleague of mine covers Arm and we don’t cover Lattice.
Patrick Moorhead: Perfect.
Daniel Newman: Hammer him Pat. We’re going to make him say something he’s not, I’m just joking.
Patrick Moorhead: Exactly. So hey, let’s dive in here. I mean, they had a beat and a beat, which was great.
Stacy Rasgon: Arm you mean.
Patrick Moorhead: Sorry. Yeah, moving into Arm, I had a beat and a beat. It was a good conversation with Renee Haas, but the guide obviously fell short even though Arm raised its guide $250 million.
Stacy Rasgon: Yeah, well they helped the full year. That’s the issue. So they’ve got to beat and beat the quarter, but they help the full year so you’re kind of implicitly taking the back half down if you do that. The other is you got to remember Arm trades at like a thousand times priced a quarter or whatever the number is, right?
Patrick Moorhead: Expectations, expectations based on the valuation are super high.
Stacy Rasgon: Yeah, that one’s always-
Patrick Moorhead: So it’s interesting, I think the licensing at 72% is eye-popping. But I got to tell you, I’m actually more impressed with the 17% royalty increases ’cause it’s so much more sticky. They have a product called CSS, which essentially, where ARM is doing more than giving you some soft macros that you can license, right? They’re doing software validation, they’re taking their IP all the way to the foundry. And they have about 50 different services they put in there. It’s higher revenue, higher margin dollars. Companies like AWS, Google, Microsoft with Data Center. So that to me is super impressive.
V9, right? You’re seeing, I think they went from 20% of licensing to 25%. Big boost, and by the way, higher margin. I think on the smartphone side, the licenses are double. I think I heard that correctly, which is good. They have about 75% less, sorry, 75% left to go, meaning there’s still room there. Long-term, it’ll probably be less 50/50 royalty and license, probably be 60% royalty, 40% licensing and an ongoing concern. That doesn’t mean licensing will go down, it just means the company thinks that royalty will go up. Unclear the financial impact of these new Arm-based Windows PCs. We know the challenge or the lawsuits going between Qualcomm and Arm right now.
Stacy Rasgon: They didn’t talk about that at all, I’m assuming. Did they?
Patrick Moorhead: I don’t know. They didn’t, I don’t even know how we would model something like that ’cause they are, sorry, they are getting licensing out of that based on the Nuvia-
Stacy Rasgon: Yeah, the only thing I find hilarious there is, I mean, for Arm in general, it’s important for them to be able to tell a story that is not smartphone-based over time. And Qualcomm is one of the biggest proponents of using Arm in things that are not smartphones and yet they’re suing them. So I don’t know.
Patrick Moorhead: Yeah. And customers just want this whole thing to go away. My final comment on this, what’s so interesting, Arm, in terms of durability, it’s not like you’re going to pull out, like if you’re a licensee, it’s not like you’re going to completely pull out and go to risk five.
Stacy Rasgon: Yeah, the ecosystem isn’t there. I mean, risk five is gaining traction in some-
Patrick Moorhead: Especially at the lower end, lower end embedded where there’s not a high level operating system. Yes, absolutely. But yeah, it’s durable.
Stacy Rasgon: Not there, right?
Patrick Moorhead: Yeah. Net-net, huge valuation people are looking for and they’ve kind of been bundled it with this AI stocks. Like they’re bundled in with the Nvidias, the AMDs, the Broadcoms, folks like that. So the expectations are pretty high.
Stacy Rasgon: Yeah.
Daniel Newman: I mean, look, by the way, you all watching the market at all?
Patrick Moorhead: Yes.
Daniel Newman: Jesus, wow. I didn’t actually look for a while and now I ran out of the screen to go vomit real quickly and then I just quickly reminded myself that it’s all paper anyway until you sell it. But yeah, it’s brutal out there. Arm did fine. I mean you guys hit it on a lot of it. I thought it was fine. I mean, you beat and nowadays, if you don’t do the trifecta, you’re not going to get the, you’re just not going to get the pop. And they didn’t do the trifecta, like you said, they did raise last quarter. But people right now, this AI play, it’s not about lean, mean operating beats, it’s about crushing. And if you’re not crushing, everybody’s just kind of like, eh, it’s it. But by the way, that’s fundamentally baked into the fact that everybody out there that has half a brain knows the economy actually sucks.
I’m just saying the economy actually sucks. I’ve been saying this for a while and it’s not like my moment of victory, but I’ve been looking at these adjustments that have been made on jobs reports. I’ve been looking at the actual inflation in the, I’m like, there’s no way this is such, and I know we’ve got to kind of dance the dance, but it’s just not that good out there. So AI carries the day and that’s what it is. And so if your number isn’t huge, if you don’t have a great AI story, if you’re not showing massive growth, and if you’re not guiding up, you’re going to just kind of get that blase result. And Stacy, I think you said it well, I don’t think it’s a thousand, but I think across the, I know I’m joking, but across the semi-space, I mean Arm trades at just this remarkably high premium.
And so that also leaves less latitude for anything less than amazing because people, so I think in the end though, they’re executing, they’re expanding. They’ve got these data center homegrown chips that they’re building on, they’re going to expand their footprint into AI. And even if they don’t expand it meaningfully, as AI expands, you can’t do it without CPUs. So they’ve got a benefit there, and no matter which handset kind of ends up being the winner in the AI handsets, Apple or anyone else, there’s Arm content. So there’s just a lot of nice tailwinds and they’re raising prices and they’ve been able to do that somewhat effectively. So all those things together, it’s been a good shot ****. Sorry.
Stacy Rasgon: Turn it off.
Patrick Moorhead: All right, we just got the explicit rating on YouTube.
Daniel Newman: Oh man, I’m sorry. I hope that the people that listen to our show are grown up enough that they can hear an F bomb when the market’s down 800 points in the morning. So off we roll.
Patrick Moorhead: Yeah, I mean it’s crazy. You see T-Mobile, Apple and AMD are actually up. Everybody else is getting slaughtered. Intel’s down 27%. I called 25% yesterday.
Daniel Newman: That’s good. I’ve got a message.
Patrick Moorhead: That’s where they were going to end up. And again, I’m not the equities guy, but it felt like there would be just a total meltdown and we have a meltdown. Hey, let’s pop into Lattice. Dan, you kicked us off but…
Daniel Newman: How much wealth do you think has been lost this morning? How much wealth has been wiped off the board? It’s got to be like a trillion dollars or something stupid like that this morning. Well actually, like you said though, Pat G might be happy because this is going to take some, the whole market’s like, oh, we’re down 25, but it would’ve only been like 15 on a good day.
Stacy Rasgon: It would’ve still been 25.
Daniel Newman: Hey, hey, silver lining’s playbook here, buddy. So Lattice has been in transition, new leadership, Jim Anderson went over to Coherent. Now you got Esam leading the charge. We had the chance to speak to him on earnings day. They’re still where, they’re still kind of in this position of digging out right now. The FPGA space was sort of interesting because it held a little bit longer. They were pretty well diversified in industrials and comm and PC and devices and data center. And so if you actually remember when other chip makers and memory, it all started tanking. Lattice had three or four quarters that they kept going up and it was like, wow, it was like 11, 12 record quarters in a row. Well, those chickens have come home to roost. And I mean the last three have been pretty rough and I’m looking for the good spots in the business. I think the best thing that I could say about their print was better than the other FPGA businesses right now.
Stacy Rasgon: Xilinx, and Altera, AMD, and Intel are, their FPGA business are pretty bad.
Daniel Newman: Even worse.
Stacy Rasgon: The general trends are the same.
Daniel Newman: Yeah.
Stacy Rasgon: By the way, you can look at the other analogs, the guys that do industrial and auto, it’s not great out there right now in those broader diversified markets.
Daniel Newman: By the way, when Qualcomm showed its big IoT number, I said to Cristiano, I said, I don’t get it. I said, I don’t see any data that’s showing strength. And then he reminded me about the XR business and I’m like, got it.
Stacy Rasgon: Yeah, consumers actually not bad. But with PCs and smartphones and XR and tablets, those already collapsed. So you’ve already hit bottom. We’ve had in general kind of an asynchronous kind of cycle where the different end markets have been rolling over and bottoming at different times. So it has the effect of actually elongating the whole thing. But depending on where you’re exposed, you could have already seen it or it could be still to come.
Daniel Newman: So with much of Lattice’s stuff sitting in industrial automotive communications, they just have a lot of parts of their business that are down. And industrials has been tough. Comms has been, it’s funny you said T-Mobile’s actually up, but comms, the actual hardware comm space has been a mess for just a long time. And so the overall space, Lattice I think is managing it well. They’ve been able to protect their margins, keeping it close to 70%. They’re still delivering, but they had had such a run-up, that on a year-over-year basis, it just looks terrible. And so they’re performing, they’re operating, they’re managing, they’re expanding their TAM with their mid-range products. And overall, like I said, I think it’s as good as they could do given circumstances. They’re not a huge company, but in this case, maybe that’s playing in their favor. They’re in that space. They’re focused on it.
Stacy Rasgon: Yeah, I mean there’s not anything that’s specific to them. It’s just, again, the markets are weak and they overshipped, just like a lot of folks did. And so you have to work that off. It takes time.
Patrick Moorhead: Yeah, I do the self-inflicted, market inflicted and absolutely market inflicted. You compare, like you guys said, to Altera and AMD, and they’re down less and they have been down less. The thing that’s really interesting here is that all of their business, 99% of it right now are these lowest power small FPGAs with small amount of LUTs and they’re doing-
Stacy Rasgon: Vivanta.
Patrick Moorhead: Yeah, now with Vant, mid-range, which very small percent, super small percent right now. So it’s all upside. And they’re going after AMD and Altera with brand new designs in leading edge fabs. And by the way, FPGAs, we’re not talking about three nanometer here, okay? Anything smaller than 22 is small, and I think that’s big. The other that I find interesting, it’s not just about the hardware, is the attach rate on the software. I think the company said they have a 50% attach to their software, and that does two things. First of all, it drives revenue because it’s an incremental cost. And second of all, it’s stickiness. As we know, hello Nvidia, write something directly to a piece of software and you might be like, well Pat people don’t dump FPGAs for ASIC at the low end. For the most part, you’re right. But you give them less of a chance and motivation to do that by connecting the software. Last comment, all that software works together between the low-end and the mid-range, so I think it’s going to, the company has the potential for a very bright future. So with that said, Stacy, those are all of the chip chippery. You’re more than welcome to stay. We can sling Microsoft and Amazon. What do you think, man?
Stacy Rasgon: I’d love to, but I actually have another call I got to get to.
Daniel Newman: Come on, Stacy. You don’t have anything else to do, but hang out with us. Don’t lie.
Stacy Rasgon: I’d love to.
Daniel Newman: You’ve got a string of tweets about his dog he needs to put out and he’s got to get back after it.
Patrick Moorhead: I love that dog. Chippy is my little friend. I love Chippy.
Stacy Rasgon: He’s sleeping over there on the floor.
Daniel Newman: I heard him barking earlier though. It’s good to see you, Stacy, thanks for playing. We’ll have you back on the Wheel of Six Five Fortune.
Stacy Rasgon: You bet. Anytime, just let me know. I’ll be back.
Patrick Moorhead: Thanks, man. All right. That was fun, man. Stacy’s great.
Daniel Newman: He is fun. He is fun. He knows a lot. It’s rare there’s someone that knows more than you. I’m not saying he does. I’m just saying it’s rare that it happens.
Patrick Moorhead: No, no, listen, we look at some things from different, some different angles.
Daniel Newman: I do learn every time we go on, you want to hit Microsoft?
Patrick Moorhead: Yeah, let’s jump in here. So listen, they cranked out a beat, beat, record bookings, cloud revenue, 21%. Azure up 29%, below investor expectations. They got hammered on the AI CapEx that they were investing in, but it’s like, they kind of did a Meta and maybe they didn’t do as well of a job doing the explanation. But I got to tell you, there’s so many great things came out on that call related to AI, right? Azure AI customers using data and analytics up 50%. 14,000 paid Microsoft Fabric customers. And by the way, don’t be confused. One of the biggest challenges out there with enterprise AI is getting your data estate and Microsoft Fabric goes across multiple types of data sets, multiple types of data location. The other thing that came out, and I think Satya refers to these as green shoots. Like you have a tree and then you see something new coming off of it that could turn into something big.
GitHub Copilot, 76,000 companies. Copilot inside of GitHub up 180% year-over-year. And I like the way that the company is kind of taking a certain, I want to call it internal case study, external case study for a product line, and talking about how the meaningful nature of what it does for AI. Copilot for productivity up 60%, quarter-over-quarter. And huge, huge companies like Capital Group, Disney, DOW, EY says they’re going to bring Copilot to 150,000 of its employees. So listen, Microsoft really is in the catbird seat as it relates to enterprise AI, and I still think they have opportunity on consumer AI. They made the early bet with OpenAI, and it is interesting that OpenAI is now, I forget, it was the 8K or the 10K, they listed OpenAI as a competitor. But that just makes sense. I mean, that’s the mature view of technology in this decade, which is a co-optician.
Microsoft needs to work like heck to give more meaning inside of Windows. If I look at what Apple teed up with Apple AI and how embedded it was into the operating system, it’s very impressive. By the way, Apple AI is also in the Mac and Microsoft just needs to keep plugging it away with case studies, customer case studies showing real payback from their tools. Final comment. I feel like Microsoft is getting serious for the first time on IAS. Traditionally, I mean since Azure’s foundation, it seemed to be PaaS and SaaS. And SaaS is obviously stuff like Microsoft 65, D65, stuff like that. But I feel like the company is for the first time, and they projected this for the last six months, is they’re going toe to toe with AWS on generative AI IAS.
Daniel Newman: So they’ve beaten every category except one. And it was the only one that anybody really seemed to care about this quarter, which was the cloud number. And by the way, they didn’t miss by a lot. It was 29%. I think there was a target of 30, 30 and a half percent. And so that was the really sticking point. The stock fell precipitously after they announced, but then it came back as Satya talked more. I think you hit a really good point on the 76 or so thousand in GitHub. What that says to me is there’s a fundamental shift in how software is going to be developed going forward. And that’s meaningful. That’s meaningful in terms of how companies are going to build software and how they’re going to hire. So as we go into this job deflation period of time and the increase in unemployment, that’s an area where we were sort of told along the way it was going to be guaranteed. And it may not in fact be, not to say we won’t need keep developers, we just may not need as many developers.
I think it was 60,000 Azure, people building in Azure AI was another number I thought that was really important. I think the market this quarter has been, there’s been two big themes this quarter, CapEx spend and value out of AI. So companies are spending a ton. So this Microsoft number and the Amazon number and the Meta number, everybody was really, really looking hard and deep at CapEx spend. Funny Zuck, Meta, another earnings we didn’t even talk about here. We didn’t really talk about Apple or Meta, both doing better. Apple’s actually up in a crazy day today. But Zuck actually came pretty much straight out and just said, we’re going to be spending more. We just can’t get behind. And the market loved it. I mean, they did have a great beat and a raise and their advertising and core business are super stable.
Patrick Moorhead: Yeah, what was the difference there? Why was the reaction different, you think?
Daniel Newman: Well, I think one is they build their product for their own product. So if you look at the difference of all this CapEx investment in these cloud providers that are building product to enable others versus Meta who’s building AI to enable itself, it also seems to have the ability to do a lot of it on its own. It seems to be doing it somewhat efficiently. And of course their margins are incredible. Their growth and their margins and their EPS, I mean, look at their EPS, it’s like over five bucks a share, if I recall. Playing this by memory, I mean they’re just delivering. So this is not Meta Labs. This is not a bet on something they don’t believe in. AI is going to be foundational to that company long-term. So he said it, he came out, he was explained it. People appreciated it. I think Satya was able to ease the market the same way through his comments on the earnings call.
But people just, it’s like they’re in this tug of war, spend, don’t spend. I came out and said, look, this is not cyclical. This is exponential. The ingestion slash digestion slash gestation period of AI is not finite. It’s not fully understood. We don’t know when value is going to be realized by every company. And sometimes you have to build factories. It certainly isn’t building a new fab, like what Intel is trying to do. I think there’s a lot of reasons to be more bullish, but record results, record outcomes, growth, 29%, which by the way was substantially bigger than Amazon’s growth. So both Google and Microsoft continue to gain share on AWS despite law of large numbers. But that’s meaningful too. So Pat, overall, I mean, it was a really good print. I just don’t see, but the markets, like I said earlier, the market sucks. The market sucks, and so the volatility is skyrocketing.
Patrick Moorhead: Good analysis, bestie. Hey, let’s jump into our final symbol here, and that is Amazon, AMZN. Dan, what do you got?
Daniel Newman: Well, first of all, I mean Amazon, it was a miss on revenue, narrow. I mean, God, terrible company, only did 148 billion of revenue in a quarter. They were supposed to do 148.8 by the way. They beat on earnings, had a good earnings result, they beat on AWS. That was a good number because that was the number that on Microsoft, everyone cared about. So you would’ve thought that might’ve kind of buoyed the company a little bit, but they had a little bit softer guidance, which again, sort of fits the theme that we are heading into somewhat of a recessionary pressure. It’s kind of interesting as the election comes, now all of a sudden, I thought we were supposed to hold this crap together until the election was over. So I don’t know how they’re planning to do this, but maybe there’s a rescue decision that’s going to come and get everybody really excited. Pat, I’m going to prognosticate. Hold on.
We’re going to have a rate cut. We’re going to have a rate cut now. It’s going to happen. Rates are going to come down now. Jobs are imploding. The market’s imploding like, okay, now can we cut rates? So that’s going to happen. If that’s good, that means inflation, I don’t know, we’ll see what happens long-term, but maybe we finally beat it to some extent. Just don’t go to the grocery store or try to buy anything. Pat, two things, two notes, because I know we’re running out of time. You got to get rolling. The cloud performance is good. AWS has a big mission ahead of it to prove it is not a third player in the public cloud for AI. That’s the big mission the company has. And even notes, there’s rumors or some confirmed rumors of Andy asking what the heck happened. So Andy’s basically, as I had speculated about the decision to move on from Adam, there was some contention about how the company handled its transition into AI and the AI cloud has flipped the rankings, not the size of the businesses, just the rankings of the businesses and AWS is pissed.
Andy is pissed. That was his baby. He wants to see that company on top of AI. They have work to do. That’s the area they have to prove. And so by the way, it’s something I think you and I will be digging into quite a bit over the next few quarters, where will this kind of play out in the end? Because AWS has the infrastructure. They have SageMaker, they have Inferentia, they have Trainium, they have Titan. Just work to be done to prove they can keep customers on it for AI. Advertising missed, but advertising is now a 50 billion business for them. People don’t think about them for that, but that’s a massive revenue stream for the company. But again, did not grow fast enough. And Pat, I’ll kind of wrap up by saying this. Very sturdy, diversified business, their retail business has actually seen its margins almost double. It’s down a little bit quarter-to-quarter, but double from a year ago, which is a huge thing. They’re making more money, they’re delivering more profit, but yet from a value standpoint, they’re, from what I’m reading and I need to assess this a little bit more, we’re not the equities guy, but it looks like they’re valued about where they were in 2020. So all the growth they’ve done for four years has not actually given any more value to the market, which is pretty interesting because to me, I think the company’s improved its margins, diversified, and performed pretty darn well.
Patrick Moorhead: Yeah, great analysis there. I’m just going to add a little bit of color on your very astute comment about AWS and generative AI. And there’s a holistic way to look at that, meaning how, and I think a lot of people are comparing the company to Microsoft that has both enterprise and consumer capabilities. Where is generative AI making more money for Amazon.com? The media businesses, their massive home electronics business with things like Alexa and Fire TV, and then of course AWS. And AWS has been a pioneer in so many things. I mean, they invented IAS at scale and they’re used to creating markets and then staying ahead. I do believe that they got a little bit, I wouldn’t say surprised, of course they knew about generative AI, but right now they are really, I think, need to be very loud and proud about its customers, the benefits that they’re deriving from generative AI and also being able to paint that company-wide picture.
Because if you can’t get tons of customers to go on the record and talk about all the benefits, then maybe get Amazon.com or get Panos on the devices side talking about how Amazon generative AI is the best thing since sliced bread. I also think there’s a lot of competition out there for homegrown silicon. AWS was the pioneer of that. We now have Google and Microsoft who by the way for that matter, Meta, but they don’t compete with Meta, building their own silicon. How does the performance look? We now have a competitive market and I think measurements between let’s say Graviton and Maya would be interesting. Or Trainium in Maya, sorry, sorry, I meant Cobalt and Graviton. But anyways, you guys get the picture. It’s a different market and I think that Amazon would serve itself well. And I do think they’re motivated and I do think they’re energized to get that message out. Dan, any final comments on Amazon?
Daniel Newman: No, I was just saying if we only knew of a lab that could do that kind of testing and then provide that kind of analysis to the market. Anyone knows anyone, you should ring us up. Maybe tweet at Pat on social media and let him know. I personally believe this week, last point, we got some attention around the TPU being used by Apple for training some of its LLMs and there’s more these flexible sort of accelerators that can be used for training inference purposes. And now we’re basically seeing companies come out with specialty products. And of course, I guess, is this conversation going to be a big thing over the next year, Pat? Because Nvidia kind of had this undisputed period. It was just like no one else is there. And then the only competition was going to be the other big flexible GPU from AMD. That was it. And then of course we’re waiting for Intel’s. And now we’re starting to kind of hear, well, are more people going to consider the route? Is TPU going to be an interesting alternative? It’s just an interesting inflection, Pat. I think we’ll have to get some of that stuff going in the labs. See how some of that stuff performs and maybe we’ll write some analysis on it. Of course, if we knew a lab.
Patrick Moorhead: We also knew a company that would do economic analysis too to go cut across its customers. I think if you guys know of anybody who does that, let us know too. Hey, how about that? We started off with chips. We ended with chips. Dan, what a whirlwind. And we are six minutes over, which I think is pretty good for us, ’cause we did seven topics and we had a special guest, Stacy, so just want to thank everybody for tuning in. Hopefully you are entertained and educated. Were you not entertained? I love that meme. One of my favorite Top 10.
Daniel Newman: Are you not entertained?
Patrick Moorhead: Those guns are scaring me, man. This is why I keep the camera below my guns.
Daniel Newman: Sorry. It’s all guns, no head.
Patrick Moorhead: I love that. Yeah, thanks for tuning in everybody. We appreciate you, Dan, thanks. Good to see you. Hope to see you maybe this weekend we can maybe break bread. We can come over, take some selfies.
Daniel Newman: Lets do it.
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.
