All right, everybody, we're going to get started with the next presentation and ask the group in the back to move outside. So we're really happy to have Akash Palkhiwala, the CFO of Qualcomm, up on stage with me today. So Akash, I know it's a very long commute for you to get up here, so we chose wisely for Qualcomm, if nothing else. So why don't we just start with a little bit on the near-term business conditions and specifically where we are in the smartphone cycle? Last quarter, basically, people have thought that you'd have a snapback like we've seen to a certain degree in PCs, but it just really hasn't happened on the handset side of things. So what's Qualcomm's view of when we will start to see a rebound in that side of the business?
Yeah. So as you know well, there are two factors kind of impacting the handset market. One is kind of just overall sell-through and then the weakness we've seen, especially in China. And then second are the inventory dynamics, where there's definitely more inventory in the channel than we had anticipated at the beginning of the fiscal year. And we've been very clear in our kind of framework that as we get through the rest of the calendar year, we expect the inventory to drain down. So, you know, in our minds, there's a short, shorter-term phenomena, which is getting through kind of these headwinds, which we're confident that we'll be able to get through quickly now.
But as we think about the company, we step back and think about the long term, and fundamentally, not much has changed. Of course, we're watching the market size, total handset market size very closely because that's a key driver for us in terms of scale. But when you look at our product leadership, our roadmap, how we compare to our main competitor, the new technologies that are coming in for us going forward, we're very excited about what's in front of us.
So when I look at the handset market, I was I didn't think there would be the impetus to double order, build the inventory, et cetera, like you would have had on the PC side during the pandemic side. So it's a little surprising to me that it's taken this long, and this isn't a Qualcomm-specific dynamic at all, so no, no blame on you guys. But is there anything else, like, kind of post-5G, you know, refurbished handsets, anything that is a little bit more structural in your mind than cyclical?
So again, I would just divide this into two parts. If you look at the total size of the market, what we've seen in history from the cell phone market is as new technologies come in, the replacement rates go up and then normalize over a couple of years and then go back again, up again when a new wave of technology comes in, s o we saw that through 5G. Now, of course, COVID came in the middle of it, and so it exaggerated the high and the low on both sides of it. But this is a consistent trend in the phone industry.
I think when you step back and say, "Okay, what's the next peak that will come in?" We're very clear in our minds that the AI phenomena that's happening on the cloud is going to migrate to the edge. And as that happens, that clearly creates a driver for devices to have an impetus for growth. And you know, this was a theory, I don't know, six months ago. Three months ago, we were beginning to see use cases come up and OEMs think about: How do we create kind of the next set of Gen AI phones?
Clearly, Google has been active on this, Microsoft has been active on this, and now as we stand here today, we are even more confident that that's a change that's going to happen because of what we are seeing our partners doing already in terms of getting ready for launches in 2024. So we're pretty optimistic, t he other thing I would say is, the way the replacement rates work in phone industries, you're somewhere in the 3 to 4 year range of replacement cycle, and the last big cycle happened in 2021. So just naturally kind of following the timeline, you'd expect, 2024-2025 timeframe to see benefit independent of Gen AI. And then Gen AI, if those use cases take off, that'd be an incremental tailwind to it.
So I want to go into the AI side in a minute, but first, let's just get the last potentially bad news thing out of the way, and then I promise we'll talk about good news things. The Huawei trying to do its own processors and modems, et cetera . Talk a little bit about what is going on kind of behind the scenes with that. They seem like they would be building on a lagging edge technology. I wouldn't think they'd be competitive with you. Maybe they don't take as much share in the market, and so some of your other competitors make it up. But just kind of talk about what's happening with that and size the risk that you see to Qualcomm.
So when you step back, I think everyone knows the history of actions that have been taken against Huawei, and as a result, we only had a license from the U.S. government to sell 4G chips to them. So along with 4G, of course, we're selling modem processor as well, and an integrated chipset. As we go forward, Huawei has made a decision to launch a 5G chip to the best of our understanding, and so we don't have a license to sell 5G to them, so they'll have to find another source if they want to launch 5G devices. And that's the process we're going through, t he rest I think you'll hear. You're hearing the same rumors as we are hearing on how they are trying to accomplish that. We'll see how it plays out.
Now, when you step back and think about, kind of I'll go back to our product roadmap because we feel that against what they're trying to do, we'll have a strong advantage in terms of power, in terms of performance. Our roadmap on technologies, especially CPU and AI, is very, very strong as we go forward. So we think we're in a great position. Our OEMs obviously have picked up a lot of the share that Huawei had before. And these OEMs are looking forward to great launches. So we've sized the impact of Huawei to us going forward. It's already included in our guidance, and we're just looking forward to working with our partners for new phone launches.
All right, let's get on to some more positive sides of things. So in the AI world, I think AI going from, you know, training at the core to inference at the edge is inevitable. Talk a little bit about the use cases of where you, as Qualcomm, can benefit the AI at the edge side of things.
Sure, s o our view on AI is relatively clear. If you think about processing, the processing has always been split between the cloud and the edge. There are certain things that happen on the cloud, the rest happens on the edge, and we see AI as not something that's very different from that. Obviously, we've seen AI first happen at the cloud. But when you think about the edge, the advantages that edge brings to cloud processing cloud AI is very significant in our minds. The first one, and the most obvious one, is cost. Once you've bought a phone, you just have that computing power sitting there versus using up expensive resources in the cloud to run use cases.
So it's very straightforward that there's a massive advantage of running AI use cases on the phone, on the edge, versus running in the cloud. The second is latency, t here are several use cases, especially if you think about gaming and how GenAI will play a part in gaming. As an example, one use case is creating avatars on the fly during a game based on the environment you're in, based on your individual preferences, what characters you like, as an example. That would have to happen on the device. You couldn't make it happen in the cloud. There's not enough time because of latency reasons. The next one I would point to is security and privacy.
One strong way to guarantee privacy, and which is one of the key topics for GenAI, is to do it on the device rather than do it in the cloud, and keep the information on the device. So those are all obvious ones, but in our minds, running inference on the device is about new use cases beyond these things, and content creation is a great one. You could take a photograph, as an example, and then you can alter the photograph based on text or speech instructions, and using GenAI models to go do that. You've heard about Stable Diffusion. It's a great way to create art on the phone, where you could take your finger and start drawing something, and a GenAI model in the back can start creating art based on what you're drawing or how you're moving your hand.
Or you can give verbal instructions to draw something out. So there, there's just a lot of use cases, and I won't profess to be the expert at use cases, but the point is, once you have this capability on the edge, the ability to do different things with it is tremendous. The second point I wanted to make is, this is not just about phones, right? So clearly, with phones, we're going to see GenAI get adopted with devices launching later this year. You will see it across all OEMs, and let's say in the next kind of 12 months, you're going to see a whole new set of new devices come through on the phone side. But as we go beyond that, we're going to see benefits show up in PCs.
Clearly, Microsoft is very focused on this, and as they transition their operating system around the use cases and the Copilot use cases that they've outlined, that's something, as we enter that area, we'll be able to give a significant benefit too. The last two I'll highlight is ADAS. Obviously, AI, not just GenAI, but overall AI, is at the center of ADAS, and so it's something that we've had a lot of success with, and we're seeing that continue to drive adoption of new technologies within cars. And then the last one is IoT. I think we're going to see GenAI use cases in various IoT devices. And what that does for us, in our minds, is you go from microcontroller-centric, wireline-connected devices to processor or AI-centric wireless-connected devices, and that's closer to home for us.
So those are the technologies that we are very good at, and we'll be able to bring those to those devices. Last thing is, as we think about monetization, the amount of content growth that happens on the chip side, in our minds, is very significant because you're going to see increased, increased processing capability, increased memory bandwidth. Obviously, if depending on the use cases, you're going to have better graphics that are needed, better processors that are needed as well, CPUs. So there's a lot of opportunity for us to kind of increase content as these use cases get adopted. And then, of course, to the extent that we do a better job than our competitors in phones and outside phones, we have a benefit in terms of picking up share as well.
So that, that's a very good answer across all those different edge nodes. It's interesting to me how much of that is processor-based versus modem-based. So talk a little bit, I know that's nothing new to you about apps, processor, Snapdragon for many, many years. You're probably the world's leading low-power processing company. But talk about the processor expertise at the company and how you evolved to penetrating processor-centric markets rather than kind of communication-centric markets.
Yeah, s o I think when you look back at the history of Qualcomm, we started with the modem, and I'd say we're clearly the best in the world at modem in our mind. And that expertise or that lead hasn't gone away. But increasingly, we are a processor company rather than a communication company now. And if you just take our premium-tier chip and we've we showed a web slide about a year ago in our earnings release, is about 85% to 90% of the content is driven by processor, and the remaining 10% to 15% is driven by modem. So if you just think about Qualcomm, really, we're first a processor company and an AI company on the device versus a modem company.
Now, when you step back and think about how our framework of AI on the device is three parts. First is we have a CPU, s o anything that runs on a CPU can run on our CPU as well. And the advantage of a CPU, obviously, is it's in all devices, and it's very easy to port use cases on the device. And we have, in our minds the most power-efficient CPU implementation there is on the edge. So that's the first one. The second is GPU, w e support all open standards like OpenGL and others. A nd so GPU use cases can be ported onto our chip as well from a Gen AI perspective. The third engine that we have is our NPU, neural processing unit, which is specifically designed for AI workloads.
So what it does is, think of it as a new computing engine that is needed for the most power-efficient operation of AI workloads. When you get to the edge, the single most important metric is power. So it's performance divided by power is how you measure how good you are in implementing technology. And we feel like with our NPU, we have a significant advantage over anyone else in implementing power-efficient AI operations.
How does NUVIA fit into all of this? When you bought it, some people thought that might have meant you were moving back into kind of the server CPU side of things, but it seems like it's much more focused on the edge and the power-centric side. So talk a little bit about NUVIA and how it fits into this whole?
Yeah, we were pretty clear that what we were looking for is really to acquire a strong team and a strong heritage when we did the acquisition. We've used that team and our architecture license with Arm to develop highly power-efficient cores for edge devices. We're starting with the PC chip that's coming out later this year, in devices mid-next year. We're going to then use it in phones, and then we're going to use it in automotive. So you should think of this as something that will be pervasive across all of our devices, and we're very confident that we'll have CPU cores that are aligned with our architecture but really have a performance advantage versus our competition.
So let's pivot a little bit over to diversification from a high level. I know everything you just talked about in AI will do that. But between your handset market and your, in the QTL, the licensing side of things, you're still a very heavily handset-oriented company. How important is diversification, and is the speed of doing that something that you consider when you determine whether to do it organically or via M&A?
Yeah, that's a good, great question. So when Cristiano became CEO about 2 to 3 years ago, he kind of changed the emphasis of the company, not just how we were communicating externally, but how we were operating internally to prioritize diversification, and especially focusing on auto and IoT. And the way we started developing technology, that became the first thought, versus "Let's make technology for handsets, and then let's use it in other areas." So we've seen a tremendous change in the way we've been operating internally, and I think, that's showing in our results. Now, we've done extremely well in automotive, I mean, by any measure in my mind. And if you think about we started off with telematics, we extended to infotainment and digital cockpit, and then finally we extended into ADAS.
And the design win pipeline kind of shows how well we've done. The same applies to IoT, w e've obviously in some ways kind of been impacted, especially in the short term, with the environment, as post-COVID and from an inventory perspective as well. But when you back up from it and think about the longer-term trend of the demand for technology within IoT, we think we're in a great place. So that's absolutely the emphasis of the company. It's a part of our M&A strategy, If you look at the last couple acquisitions we've done, it's been focused on things outside handsets. Nuvia was clearly, t here was a handset component to it, but as I said earlier, our primary focus was entering the PC market with a very strong CPU story.
We also acquired Veoneer, and that gave us software assets within ADAS, and that has been a key driver of our strong design win pipeline. So we've been looking actively at acquisitions that accelerate our organic diversification plan. Now, one last point.
Sure.
When you step back and look at the numbers, we've grown IoT and auto significantly over the last three years, even with kind of the challenge we have in 2023. But handsets has grown tremendously, so the ratios have not moved as much as as much as we'd hoped, but it's for a good reason, because we've been able to massively increase content on a per-handset basis, even while the handset market has been smaller.
So one difference between the prior management team and the current was the size of the deals you, w ell, you've concluded since you've been there, but attempted prior, and then unfortunately, the regulatory environment got in the way. Are smaller deals, more tech-specific kind of rifle shots, the more likely strategy for the company, like the NUVIA and the Arriver/Veoneer, et cetera?
I'd say our strategy is largely unchanged. And we outlined this about three years ago at our Investor Day, and I'd say we've executed on the framework we outlined. We've been focused on two things. One is acquiring very small acquisitions where we're mostly acquiring teams or technologies, which mostly go under the radar, but several smaller acquisitions. The second area has been the ones I mentioned, which is very larger teams or technologies, and NUVIA and Veoneer being examples of that, which I think has worked very well for us. And the goal of all of these is to accelerate our organic plan. How do we get to diversification sooner through those transactions?
We've also looked at larger transactions, and for one reason or the other, we've chosen not to pursue any of them, again, consistent with our framework. I don't fundamentally see us taking a different approach going forward. To the extent we see a good fit, we will, we'll look to pursue it.
So why don't we go in a little bit into your revenue segments that you break out, and let's start with the handset one because it's the biggest. There's plenty of speculation in the market about the timing with which your, one of your biggest customers might internalize their modem procurement. I believe you've said, just take that out starting at the second half of next year. Any sort of update on what's happening there, or is that still the base assumption you'd wish us to operate under?
Yeah, that's still the base assumption. I mean, I think we've been very clear that we have a very strong modem roadmap. We'll have the best products out there, and then it's the customer's decision from there.
Is there something that, that the support necessary for any customer, if they're making those sorts of decisions, that there's leverage on the cost side if the revenue goes away at Qualcomm to offset that sort of thing? Or if they want this on a more positive side, if they wanted to stay, that you're not just going to keep your cost structure high for something that they might, you know, kick the can down the road for one year, and you would just have to deal with the problem a year later. Like, what, what are the levers that you have to adjust?
We've been, I'd say our framework has been straightforward. We've set targets for our operating margins, and we're planning to execute to that. As you've seen through this year, we've taken cost actions as appropriate. At the last call, we also said we're planning to take additional cost actions. And so, I think no change to really our framework of how we are operating the business, and we're committed to the targets.
So getting away from that hot-button topic into something you probably can discuss a little bit more, the pricing within the handset market, you know, 5G starts at the premium hero phone level, and then, you know, it migrates downward. How does the pricing year to year change for Qualcomm as a node or a G matures?
Yes, I'll start with the same kind of framework I discussed earlier, where it's less about the G, it's more about the application processor. So if you look at the last three years, and we showed a trend of this in the same web slide I mentioned earlier, the content has grown very significantly in the premium tier, while they've all been 5G devices. So even you leave 5G aside and just look at content growth, it's been a tremendous trend over the last several years. We've also seen mix shift up as the device just continues to become a more important device for people, and it's already been that in developed markets for the last several years.
But in emerging markets, the amount of content consumption that has increased over the last two or three years has been massive, and so that drives a stronger device mix change, which has already benefited us. So you look back and you say, "Okay, every year, content continues to grow, especially with an application processor." There is still a portion of 4G to 5G transition that needs to happen. And then the mix shift has moved up. So the proof is in the pudding. If you look at our per device revenue per MSM metric, and of course, we don't disclose total MSMs at this point, but I think you can back into it. You will see that there has been a very significant increase in revenue per device, and it's all these factors coming together.
So yeah, and that's, that's kind of the underpinning of my question. Just to solidify this, it's not a cyclical reaction or a G reaction. It's something that's more structural because of the processor content and those sorts of things, is, is your view?
Absolutely, I think. And it's historical and fact at this point, that the content has increased independent of G. And I think the best way to look at it is premium tier, where we have increased content by a lot in the last three generations. And in each case, they have been 5G phones.
Well, some of the performance benefits and leadership that you have has been very evident in the market share increases you've had at some of your key customers over the years. And just, you know, obviously, there's the iOS side that's difficult to talk about, but in Android, even, it's increased substantially. Is that something that you believe you can hold, that nearly 100% share in some of those high-tier flagship products, or was that kind of a more one-time or two-year pop that you saw over the last year and a half?
I think when you step back and you look at how we've done in terms of increasing share and increasing content, we think we're in as strong position going forward as we've been in the past. You're always going to see some ups and downs across the OEMs, but that's not how we plan our business. We step back and look at our overall share of available SAM, and it's been very strong. We think it's going to continue to be strong. It's all about the roadmap, product roadmap, and we've learned this over a very long period of time in this industry.
I think if we were sitting here 10 years ago and having this conversation, we had about 10 competitors in, in handsets, and it's now down to, one or two players. So, I think our strategy has worked out well. We've obviously grown our revenue a lot when you look at a longer-term time frame versus just 2022 to 2023. And it just shows the strength of our technology portfolio.
And if anybody in the audience has a question, just raise your hand and I'll hopefully see it. But, let's pivot outside of handsets to the IoT side, the next biggest part. And you talked a little bit about this before in the diversification efforts, but talk about the trends in the consumer. I think you have three buckets in there: the consumer, the edge networking, and then the industrial.
That's right
-side of things. Talk about where we are in Qualcomm's penetration in those markets and where you think the future growth could be the most exciting.
So let me talk through each one of them. Consumer, there's two large opportunities, and I consider them step function opportunities for us going forward. First is PC. We're seeing, w hen you look at the PC market, we see the PC market changing completely to an AI-based PC market, and, as that happens, we will be the leader in that part of the market. And so it'll start as a small portion of the market. Eventually, we see the entire PC market transitioning over, and we're going to enter that portion of the market as the leader, going forward. So we are in a very good place, I think, to tap into a very large, established silicon market that is going through an inflection point. Second is XR or metaverse devices.
It's difficult to predict when it will become a very large market, but I think we're set up very well. We have most of the designs, and eventually we see that market becoming larger, significantly larger than what it is today. And if and when that happens, we'll have a chance to grow significantly with it, right? So those are the two large opportunities within consumer. The key thing to note from a financial perspective is it's all incremental investment for us because we are leveraging technologies that have been created for handsets. And so financially, you don't have to make a very large leap of faith in terms of our share in those markets to make it material for Qualcomm. The second is edge networking, and that's where I'll highlight one market. So there's two parts to it.
There's a Wi-Fi access point market, which we're already the leader. The second is 5G as a broadband technology into homes. 100% reuse of technology we create within handsets going into this market, and we feel very confident that as 5G gets deployed broadly in India and Indonesia and all the emerging markets, it will become one of the home broadband technologies. So, very well positioned there. And then the third is industrial, which, as I mentioned earlier, that market is going through this transition from a microcontroller market, a portion of it. There are obviously microcontrollers will still be very relevant, but a portion of it is going to transition over to processing AI and wireless connectivity.
If and when that happens, we obviously have an entire portfolio of products that is 100% reused to go into those markets. So again, financial leverage is very high when we enter those areas. So, pretty optimistic when you look at a longer-term view of how we can leverage the technology we've created for mobile in those areas.
So we only have about five or six minutes left, so I'm going to go fast. We can kind of do rapid fire in these, and forgive me that they warrant more time each, but we just don't have it. On the automotive side of things, you guys have grown incredibly impressively over the last few years. Most of the things that you talked about at your auto tech analyst meeting have come out to be ahead of plan and understated. Where are we in that evolution, and when do you think there would be, if there is, a stairstep up or a step function increase in that market, or is it more linear?
I think financially, we've kind of outlined our forecast in the longer term, and so, that's what we're executing on. I think fundamentally, the way the trends are working out for us is very favorable. If anything, we're seeing an acceleration in terms of how cars transition over to modern infotainment and digital cockpit. And then we're by far the leader in that area with very, very broad presence. And what has happened there is we have the design wins. Not much has translated to revenue because the cars are launching in 2023 and 2024, and so you're going to see the step function as those cars launch.
And then ADAS, our view was always that Level 2 , Level 3 is where the action is going to be for the next several years, and we focused on that part of the market. What we bring to auto OEMs in that area is a roadmap that goes from the highest premium tier to the lowest tier, and software assets that come with it. We've also talked about a combined SoC that allows you to do cockpit and ADAS both with one chip. So we just bring, in our minds, tremendous advantages to a very large global OEMs who are looking for scale, who are looking for the best technology, and who are looking for an ability to combine across all these paradigms. Like phones, we went in with a system solution.
Our Digital Chassis comes with all the connectivity chips you need, all the digital cockpit chips you need, and ADAS chips and software. And the benefit of our solution is not any point solution, but the fact that you can put it all together.
So the last segment doesn't get as much attention, but it's incredibly important, at least for profitability, is the QTL side, the licensing business. Any sort of changes other than, you know, if the handset market is strong or weak in a given year, that you worry about? I get questions on, you know, does, does there need to be another renegotiation? Does that happen when 6G kicks in? What's the view, kind of, nearer term, and more importantly, mid to longer term on QTL?
So, QTL, we've been very happy with where we stand on QTL. I mean, we've obviously licenses that are signed for the longer term. There are smaller licenses that come and go as negotiations come and go. But when you step back and say, with all the major OEMs in the world, we have longer-term license. We just extended Samsung all the way to 2030. We have several years left in a couple other licensees as well, large, very large OEMs. So we feel pretty comfortable about where we stand. Now, when we get to 6G, the way to really think about 6G is, first, we hope it happens as soon as possible, and so that's something that Qualcomm is definitely going to lead.
Second is, when 6G devices come out, we expect them to also have 4G and 5G in the same device. So our technology licenses, and from a QTL perspective, still remain relevant as long as you have those technologies. Maybe the last point I'll make is 6G is going to build on the things that are good in 5G, right? And so when you think about spectrum bands, we're looking at bandwidth, we're looking at leveraging millimeter wave spectrum. All things that are strong in 5G, where Qualcomm has an advantage, are the basis for 6G. So, we're pretty excited about what the technology brings forward, and, QTL is going to be in a great position when that happens.
So I just want to wrap up in the last two minutes we have, talking a little bit about the margin front. You talked about reexamining the OpEx side of things before, and you know, I assume you're not going to really update us on that today beyond what you said. So let me just focus on the gross margin side of things. Qualcomm's done a great job of stepping the gross margins up in QCT, at least the implied ones that we can get to, over the last three, four, five years. Where is the sustainable gross margin in QCT? Because my calculations, you're kind of upper 40s to near 50%.
I would think with your dominant position, the strong IP you have, admitting that you have a big customer or a concentrated customer base in handsets, but a lot of fabless companies will do, you know, I don't know, 55% on a 50% to 60% range. Where do you think the long-term center of gravity is to your gross margin?
Yeah. So we've been, we've been consistent in kind of our view on gross margins. We think that where we are at is a good way to model the business going forward. As we look forward, especially as we look at IoT and areas that we are going into IoT, we think there's an opportunity on the upside. Auto, as we look at connectivity and cockpit and then ADAS and software assets that come into play, there's an opportunity for upside as well. So structurally, financially, I think we should model where we are at. Clearly, we are focused on kind of the key drivers I just outlined, where we have upside opportunities.
The mobile landscape is, handset landscape has been clearly defined in the past, and so we've been in a certain range from a gross margin perspective. We're optimistic that as kind of things evolve and as our competitive advantage kind of extends in certain areas, we'll be able to add to it. But, I don't think of that as a base case financial forecast. We think where we are at is a good way to model the business.
The diversification over time is probably the longer-term tailwind to focus on.
Clearly, that's something that benefits us. Yeah.
Great. Well, I think we are running out of time, and so we're going to end perfectly on time today. So Akash, thank you so much.
Thank you.
for joining me on stage.
Thank you for having me here.
I appreciate it.
Thank you.
Thank you.