The Six Five team discusses welcome Stacy Rasgon, Bernstein Research.
If you are interested in watching the full episode you can check it out here.
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: I am super, super, super excited to introduce Stacy Rasgon, Bernstein Research. Stacy, how you doing?
Stacy Rasgon: I’m good, thanks. Good to be here. Thanks for having me.
Patrick Moorhead: Yeah, so Stacy, I don’t know who gets on CNBC more, you or Daniel?
Daniel Newman: Or you?
Patrick Moorhead: Well, I think I’m available when Stacy’s not, like he’s going to the bathroom or something, they call me to talk chips on there with Fortt, but man, it’s great to see you. The first time we met I think was at a Board of Directors’ dinner for a chip company where we were their entertainment.
Stacy Rasgon: This is true.
Patrick Moorhead: Yeah, and I think they were entertained, but hey Stacy, let’s talk a little bit about what you do, what you don’t do. I know everybody in the chip space knows you, but we have a pretty broad audience here.
Stacy Rasgon: Yeah, you bet. So again, I’m Stacy Rasgon. I’m a managing director and senior analyst at an equity research firm called Bernstein Research where I look at the US semiconductor and semiconductor capital equipment space. And so what does that mean equity research ’cause you guys are tech analysts? I know you talk about companies, you do talk about stocks. I do that as well. There are some other considerations though. I also explicitly put out ratings and target prices on stocks. I articulate investment conclusions for those stocks and I write about those to my clients. Now that being said, I also do a lot of what you guys do. I spend my time immersed in the tech supply chain, in the semi supply chain.
My background for those in the audience that don’t know, I used to be on the technical side. So I have a PhD in chemical engineering. I used to build semiconductor manufacturing equipment, plasma etchers. My PhD thesis is on the plasma etch transfers a roughness down feature sidewalls during semiconductor manufacturing. And I did work at IBM T.J. Watson back in the days in their advanced lithography and processes group. Then, I went actually instead of going directly into the industry, I went into McKinsey instead and I spent my time at McKinsey doing business consulting for semiconductor companies, and now I’m on Wall Street and I look at semiconductor stocks.
Patrick Moorhead: I love it. I love it.
Stacy Rasgon: It’s fun. I’ve been doing this 16 years now. This is the longest I’ve ever been anywhere and it’s still fun.
Patrick Moorhead: Stacy, thanks for coming on. A question I get, and then we’re going to jump into Intel, AMD and Qualcomm, is how long is your timeframe of interest? Is it 5 years, 10 years, 3 years? How far does it go out ’cause I think that’s going to be important for this conversation?
Stacy Rasgon: So theoretically, target prices that I said are supposed to be a 12-month target price, so setting a target price a year. Now that being said, in this space you can’t constrain yourself. I’ve got, and it really depends on my clients, so I have clients who may just care about the next data point. It may just be the next week or the next month or next quarter. I’ve got clients that have a 5-year or 10-year horizon. And so you need to be able to sort of dial in and out depending on who you’re talking to and have a point of view across a variety of timescales. But in theory, the target prices, ’cause you got to put a stake in the ground somehow in theory, that is a 12-month target when I do put out target prices and ratings.
Patrick Moorhead: I love it, man. Yeah, I don’t tell many people I actually have a finance degree and I was supposed to be in Wall Street, but I picked a tech instead. So hey, let’s jump in here.
Daniel Newman: Hey, Pat, do you have a real job?
Patrick Moorhead: Yeah, no, I used to have a real job too. Dan loves when I point that out ’cause-
Daniel Newman: ‘Cause I never did ’cause that’s the joke.
Patrick Moorhead: No, Dan, you kind of had a real job. You were CEO, a distributor. Okay, hey, let’s dive in. Let’s do easiest first.
Stacy Rasgon: Yeah.
Patrick Moorhead: So Qualcomm, Stacy, big upside in automotive, smartphone, better than maybe some expected. So what were your thoughts on, what does your note talk about?
Stacy Rasgon: Yeah, so it was a good print. So Qualcomm has been an interesting just stock. It was death most of last year as were most of the smartphone names and it wasn’t anybody’s fault. It wasn’t anything that them or frankly any of the other folks in that industry were doing badly. It was just smartphones were horrendous. They didn’t really benefit. Like PCs during COVID had a lot of pull forward and for a while we had a lot of demand. Smartphones, we didn’t really see that and in fact, demand was worse. I think smartphone shipments were down eight quarters in a row-
Patrick Moorhead: Right, exactly.
Stacy Rasgon: … In most of 2022 and 2023 and the semiconductors, especially in 2023, there was just too much inventory out there, so they wound up under-shipping that low demand. It was just awful. Smartphones, however, bottomed near the end of last year, and that actually was very good for most of these smartphone names and they had a really big bounce, earnings went up and the multiples went up and the stocks were pretty good over the last six months. Going into this print, people were getting a little more nervous. There were some signs, people were worried about Apple shipments and everything else.
Qualcomm took their pain last year. They’re more exposed to the premium at Android tiers. They had some inventory correction last year, which has bottomed out for them and so they were able to put up some better numbers, I think, than some of the other companies were. The other area where they’re really, really benefiting is content. And this is where I’m actually starting to warm up more to some of those longer-term horizon kind of things. I’m getting more bullish on the prospect for AI smartphones and by the way-
Patrick Moorhead: There we go. Now, you’re talking my love language, Stacy.
Stacy Rasgon: Right. Hang on, maybe or maybe not. I am not bullish on AI smartphones per se in the sense that I do not believe that anybody is going to run out and buy an AI smartphone where they were not going to buy a smartphone already. I am not convinced that it will drive an upgrade cycle yet. However, it should drive content, and this is something where Qualcomm’s actually benefited a lot and in fact if you look, the whole 5G cycle didn’t drive any kind of upgrade cycle.
In fact, smartphones are down, oh, I don’t know, 20% or 30% off the peak during the whole 5G cycle. Qualcomm’s chipset revenues, however, grew at a 20% CAGR while smartphones were collapsing on the back of content increase, as well as some other things you mentioned, auto and everything. But content’s gone up a ton and I actually am getting more and more convinced that at least AI can continue those kind of content trends. And I do think when people do buy a smartphone, they will buy an AI smartphone because that is what will be for sale.
Now, I also think Qualcomm has been doing a lot more on the ecosystem work for AI smartphones, whether it’s software support, OS support, SDKs, pre-trained models. They’ve got a whole website with a ton of pre-trained models that are optimized to run on Qualcomm Silicon and nobody else in the supply chain is really doing that kind of optimization. So they recognize that there needs to be use cases in killer apps for this and we’ll see if that happens or not, but they’re doing what they can to try to create that ecosystem within this space that is really important. So I’m starting to really warm up to that thesis and look, if there is a killer app, maybe Apple comes up with something that everybody has to do and if it drives an upgrade cycle, even better. That’d be great.
Patrick Moorhead: Yeah, we have Google’s big conference coming up next week where they would likely drop something about what they’re doing in Android. And I think, by the way, all this discussion is very understandable ’cause we haven’t seen what Google, what Microsoft, we haven’t seen exactly what they’re going to be bringing out. With Build coming up in a couple of weeks and we’re going to see this. We have WWDC coming up that should talk about that.
Daniel Newman: You mentioned Computex.
Stacy Rasgon: Yeah.
Patrick Moorhead: Oh, yeah, yeah. Dan is going to be in Computex, I am not.
Daniel Newman: I’m going to be sweating profusely in Taiwan.
Stacy Rasgon: I will not be there unfortunately.
Daniel Newman: Guys, I will carry the water for you.
Patrick Moorhead: Congratulations, Dan.
Daniel Newman: Again, carry your bags, carry your, anything just to be in this kind of circle. Can I just hang with you guys?
Patrick Moorhead: Yeah. So hey, just a final quickie on Qualcomm. They have this $45 billion auto backlog. Definitely, big growth in the quarter, but definitely out of your one-year target, it goes out eight years I think is, is there… How do you factor that in Stacy?
Stacy Rasgon: Yeah, you bet. So they’ve given some medium-term as well as longer-term targets for revenues. What they’ve said is they think they can do more than $4 billion by fiscal year 2026, which is actually the out year of my forecast period.
Patrick Moorhead: Yeah.
Stacy Rasgon: And more than 9 billion by fiscal year 2030. So those are the targets they’ve given. They’re running right now annualized a little under 2.5 billion and it’s growing a lot. So they’re on that glide path to hit those numbers and what they said was, the last time they’d given a pipeline number was 30 billion a year and a half ago whenever it was and now they’ve taken it up to 45 billion. They’re not taking those targets up yet, but what they said is we feel even better about hitting those targets. Yeah, it’s looking really good. So I’ve got a glide path in my numbers for something that looks like that.
Daniel Newman: And the real revenue, Stacy, has been pretty impressive too ’cause that’s been the thing is like some of these pipeline numbers and then you’ll hear why-
Stacy Rasgon: Yeah, you never know, right?
Daniel Newman: But you’re seeing that trajectory of actual revenue that’s showing up in that business that sort of seems like the ramp is calibrated.
Stacy Rasgon: Yeah, yeah. You got to remember when people give a pipeline, those are engagements. You never quite know how much of that’s going to actually convert into actual orders, you don’t know over what timeframe. I mean it could be 45 billion over 30 years, who knows, right? It’s not going to be that long, but in general, when companies talk about pipelines, they tend to leave those details out. At least Qualcomm is putting up very reasonable revenue numbers along order numbers or pipeline numbers that look like that.
Patrick Moorhead: Hey, let’s bounce to AMD. EPSV revenue, small little beat outlooks, lower than expected. Your headline in your report was “Not awful, but not awe-full” as in full of awe.
Stacy Rasgon: Yes, yes.
Daniel Newman: Was that actually his headline? I didn’t-
Stacy Rasgon: Yes.
Daniel Newman: Nice. All right, well done.
Patrick Moorhead: Just wait for the Intel one.
Stacy Rasgon: Yeah, by the way, I should say, like we were talking earlier, titles are really important. My clients get a thousand e-mails from folks like me every single day. You got to give them some reason to click and so tight titles are best. I spent a lot of time on the titles. So AMD, the stock has been tougher recently, although I think year-over-year it’s still pretty good. It had a big ramp up last year, frankly in the wake of NVIDIA’s, when NVIDIA kind of blew the doors off the car in May of last year and really got the whole generative AI cycle going and AMD, at least from a narrative standpoint, has been benefiting from that. The stock has sold off a bit off the peak and it partially is because it’s a few reasons.
One is their core business has not been great and this is by the way, we haven’t talked about the broader cycle semis. There’s a lot of overhang that they’re working off from the COVID years, AMD and their core businesses getting hit by that, and the AI piece for them, as sizable as it is for them, has not been enough to offset some of the weakness in the core business. Plus I would also argue, in the context of the overall AI opportunity from some of their peers like NVIDIA are saying, even the multi-billion dollar numbers that AMD is putting out are fairly small in that context. NVIDIA might do a $100 billion in data center revenues this year. AMD is guiding for four plus. I mean it’s not that… Although AMD do their credit, they don’t have to do a $100 billion in revenue this year.
The issue with AMD stock and it did sell off off the printing, it was kind of like inline-ish, which the stock is fairly expensive and inline in that context isn’t always enough. The core business was weak in particular, what they call, Embedded, which is mostly the Xilinx business that they bought a little while back and their gaming business were already weak and getting weaker and some of those cyclical issues. And the AI piece, they took it up modestly, their prior guide had been more than 3.5, now it’s more than 4. And so that was kind of like expectations are already there more, so it just wasn’t enough. Now I will be honest.
We’ve been a little lukewarm on the thesis, but again to the title, I didn’t think the quarter was all that bad in the grand scheme of things. They are still, their core client and data center businesses, they’re not super because the markets are not super, but they were not horrible. They’re clearly taking share from Intel. We’ll probably talk about Intel. Intel had suggested when they reported a week or so before that they thought they held share. They clearly did not-
Patrick Moorhead: Yes.
Stacy Rasgon: … AMD still took share and the embedded and other weaknesses, presumably some of its cyclical, eventually that will hopefully come back and the AI numbers, they were in line, but I mean they’re still growing. It’s not awful, but “not awe-full” isn’t quite enough when expectations are high.
Daniel Newman: It’s not aw-full?
Stacy Rasgon: Yeah, this is something that people sometimes get confused when they look at how stocks trade in the wake of earnings and I’ll see sometimes, Pat even I’ll see tweets from you and you’d be like, “Oh yeah, the earnings were great, the stocks down a ton.” Maybe it’s the guidance that was bad.
Patrick Moorhead: Yeah.
Stacy Rasgon: Maybe it was the expected, it can be great, but expectations are much higher. The real question is how are they performing relative not to my expectations or my peers’ expectations, but my clients’ expectations, the folks that are actually doing the investments themselves. Those expectations are not written down. Part of my job, by the way, is to figure out what those expectations actually are, but you can get a good idea of where they are relative to the numbers by observing what happens with the stock when these guys report. So that’s what happened with them this quarter.
Daniel Newman: I know you might not be able to comment on this, Stacy, but my take on it all is that NVIDIA is setting the expectations of pretty much every data center chip company on the planet right now and the comparative is, well, we’re expecting NVIDIA’s 875% growth, I joke, but… So Lisa comes out and raises the number 3.5 to 4, but people, they wanted her to raise to 5 or 6.
Stacy Rasgon: You got to remember, so last quarter they raised it from 2 to 3.5 and the stock went down. Why? Expectations for AMD going the last quarter was that they would do 8 to 10.
Daniel Newman: Right.
Stacy Rasgon: Right?
Daniel Newman: There was no number big enough.
Stacy Rasgon: So they lowered it to 3.5, this is the quarter before. That was good because you want to get expectations down and so they lower it to 3.5, but people still think they might do 5 or 6. So now they’ve raised it to 4. Great. But people are still thinking they might do 5 or 6.
Daniel Newman: That’s fair.
Stacy Rasgon: And they’ve only got two quarters left now.
Daniel Newman: Yeah, I think that’s where I was trying to go was basically there was no big enough number that she could have said. If she said 6, they would’ve been like, why not 8? I mean it was like-
Stacy Rasgon: You have to remember, they’re relatively small, there were other constraints. Part of the issue for the whole space has been supply constraints. These guys are all ramping it up a ton. Remember, you don’t just snap your fingers and the chips appear, you have to make them and you have to put the infrastructure and capacity in place to manufacture them and that takes time, and it’s very difficult to gauge. And AMD because they are smaller, NVIDIA is pre-buying a lot of this capacity, there are limits to how much, even if AMD, and I don’t know what they, even if AMD had demand for $50 billion worth of parts this year, there’s no way they could supply anything like that. They’re probably doing the best they can. And again, it’s not bad. This business was zero just a few quarters ago, like a year ago. So even to grow it to 4 billion in a year from nothing is not a bad thing necessarily.
Patrick Moorhead: Yeah. I’m going to end on AMD. In December, I said 6 billion and we’re going to see where that ends up and that was-
Stacy Rasgon: It’s not an insane number. It probably won’t be 10 this year, but what AMD said is they will have supply in the back half to generate upside to that number if the demand is there, is kind of what they said.
Patrick Moorhead: That’s good. So let’s pop to Intel. So Intel, a big EPSV for the quarter and proof gross margins, lower than expected top line, and a lower forecast.
Daniel Newman: He’s cringing.
Patrick Moorhead: Stacy, your note was “Hello, darkness, my friend, I’ve come to talk with you again.”
Stacy Rasgon: Yeah, they’re in a tough spot. They know. Look, what I’ve said prior is this was always going to happen. It took 10 years to break it. It’s going to take 10 years to fix it. Intel had a Foundry. There’s a few things that are going on with Intel. One is just their numbers have been bad. They just have, it is what it is. Again, we can talk about some of the dynamics. The PC market has not been great. There was a lot of channel impact. During COVID, they were over-shipping. You had a massive inventory correction. I think they’ve actually been over-shipping again the last couple of quarters. So that has impact for the forward trajectory of their client business, plus they’re losing share in servers, plus traditional data center, in the wake of all the AI strength, traditional data center has been quite weak, server chips and networking, in that sense. So that’s been impacting them.
They were impacted on their Altera business, the same issues with AMD’s Embedded business, that’s very weak. So they’ve got issues just in general, near-term cyclical issues in their business, longer-term potential structural issues around share losses to AMD and then now maybe AI and accelerators taking share from traditional data centers. They don’t have anywhere near the AI or accelerator story that some of their peers do. So you don’t have that, and now they have this foundry business that they’re trying to build. They’re investing a ton of money that they don’t really have. They’re burning cash.
They did a Foundry Day a few weeks ago where they showed us the economics of the two businesses, the product piece of the company and the Foundry business. The Foundry business economics are, I think, unsurprisingly horrendous. They lost 7 billion last year, which was not surprising. What I found surprising was the Foundry losses are going to get even worse this year. That’s in the wake of them- they supposedly cut $3 billion in costs last year and presumably, the revenues are getting better. So what it tells you is there’s some real structural issues.
My biggest takeaway from the Foundry Day was come back in 2030. Again, they gave some medium-term and longer-term models from margins and numbers, which I still think look aggressive personally, and even if you believe them, they basically said this is going to look like a horror show for the next three, four, five years at least. So even if you believe in that long-term, and I can construct a bull case if you wanted to for what that long-term could look like, but it’s a long ways out. It’s hard. This is why the stock has been selling off and then, we had the Huawei news the other day, right?
Patrick Moorhead: Yeah, we’re going to dive into that. So Stacy, what would you have to see from Foundry and when for it to make a change?
Stacy Rasgon: You want to see volumes, right? They’ve announced customers and engagements, which is fine, but again, by their own, they showed their own forecasts. They’re not expecting any sort of material revenues till the latter half of the decade in Foundry, and it’s still an open question. I don’t doubt that they’ve got a ton of engagements. I think anybody that’s looking for leading edge process technology is going to be looking at them. Why wouldn’t you? Everybody wants second sources. The whole geopolitical question and how sustainable is having the world’s leading edge manufacturing in Taiwan, that is a real worry. I understand all of this, but these are not problems that like, you can’t just throw money at these problems and fix them.
Patrick Moorhead: Well, their long-term… The tech has to be there. Hey Stacy, it’s going to be getting advanced EV machines. That’s what it’s all about. Didn’t you read the memo?
Stacy Rasgon: Yeah, yeah. The High NA?
Patrick Moorhead: Exactly.
Stacy Rasgon: That has a whole other set of issues that we could talk about. Now, I will say back in the days when Intel was actually at the forefront of new innovations, they were actually doing better. So you can think, they were the first ones to do silicon-germanium and the first ones to do High-K/Metal Gate, the first ones to do FinFETs, right? And they were phenomenal for them, and you could make the argument that when they stopped, part of their big issue when they has the problems going to 10 nanometers is they didn’t want to use EUV. They decided they could do it using multiple patterning. As it turns out, they were wrong. In the meantime, TSMC aggressively went after EUV, and you know how this works, it’s a learning curve.
Patrick Moorhead: BK to the end was saying that EUV didn’t work and TSMC got it to work. So Stacy, let’s bounce and Dan-
Stacy Rasgon: We’ll see what High NA does for that. There are other issues with High NA.
Daniel Newman: As we move into the… Oh, sorry. Go ahead, Stacy.
Stacy Rasgon: We’ll see, but again, call me in a few years. It’s going to be a slog and I think it was always going to be a slog.
Daniel Newman: As we move into the back end of this and get into our topics of the week, by the way, thanks Stacy for giving us the rundown. Pat and I did this last week, so everybody out there want to check out our comments compared it to Stacy’s and then send Pat and I a note and let us know how great we did or how great we didn’t do. But the one thing I say is I talked to Bloomberg this week about Intel and they basically asked me the question, “In the near term, is there any reason to invest?” Boy, was that a hard question to answer because I think what you said was exactly what I said is there is a really compelling long tail, but the short tail is really hard.
Stacy Rasgon: It’s tough. Yeah, so it goes back to your question on what’s your horizon, right? It depends and look, the book values is, I don’t know, 27 bucks a share. It’s not that far off of there, right? They’re burning, they’re burning cash, although they’re getting subsidies, and again, I can argue the geopolitical benefits, but it’s probably going to be a slog.
Patrick Moorhead: Yeah, I’m interested to see when we have a TSMC-created compute tile, so it’s kind of mano a mano with AMD. Now, on the cost side, that isn’t good, but on the competitive side and getting that market share, I see that as positive. And by the way, the company has moved from a completely monolithic design to distributed and theoretically that means your net could die per wafer.
Stacy Rasgon: They’re doing the wrong thing. What is Pat’s alternative? It’s to lay down and die.
Patrick Moorhead: Exactly, and some people are like, “Oh, they need new leadership” or “they need to just split.” It’s like, no, no-
Stacy Rasgon: It’s exactly what he said. I think his biggest error was how he set expectations. So he came in and he was really a cheerleader. He said, “Everything’s fixed and it’s going to be great.” They had an Analyst Day in 2022 where they put out targets that were outlandish, but you also have to remember, investors, we’re not the only audience for what he says, right? It’s also the US government and its employees and its customers and its suppliers, and you have got to keep the boat from totally sinking even though there’s a hole in it, right? He has to maintain all of that, so the investors themselves are not the only audience for the messaging, but at some point you’ve got to put up or shut up and we’re at that point now where everything is now starting to hit him. I think it was always going to be like this personally.
Patrick Moorhead: Right.
Stacy Rasgon: We’re just at that point now. They’re going to have to get through. It’s going to be a while.
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.
