QUALCOMM Incorporated (QCOM) Earnings Call Transcript & Summary

Please welcome Vice President, Investor Relations, Mauricio Lopez-Hodoyan.

Hello. Good afternoon. Good afternoon, everyone, and welcome to QUALCOMM's 2024 Investor Day. It's great to be in New York. It's great to see all of you here today. Thank you for attending. Before I start, let me just thank all the teams that work in today's program and the executives for spending time on these presentations. Now for some housekeeping. We will make forward-looking statements in today's program regarding our business and financial expectations and other future events. I would like to refer you to our most recent 10-K for a description of our business and associated risks and other important factors that could cause actual results to differ materially from those in the forward-looking statements. We will also use non-GAAP financial measures as you find in Regulation G, and you can find the related reconciliations to GAAP on our website. Lastly, today's agenda includes presentation by Cristiano Amon, Durga Malladi, Alex Katouzian, Nakul Duggal and Akash Palkhiwala, followed by a short Q&A session. And with that, please join me in welcoming QUALCOMM's President and Chief Executive Officer, Cristiano Amon. But first, let's kick off the day with a short video from one of our partners.

Good afternoon, everyone. I'm glad to be part of Qualcomm's Investor Day and to talk about our shared commitment to using AI and generative AI to transform industries with the vision to build on the cloud, deploy on the edge. AI is going to improve almost every customer experience we know or can imagine. Amazon and Qualcomm have been long-standing partners at the forefront of innovation. Together, we're making it easier than ever to build and deploy AI integrated solutions from cloud to edge. AWS and Qualcomm are combining our strengths with the Qualcomm AI Hub and Amazon SageMaker, making it easier for customers to build and deploy powerful AI solutions on edge devices. For example, in the automotive sector, we're working together to democratize vehicle software development, allowing automakers to experiment and integrate cloud technologies into their development processes. Our collaboration is working to transform the auto industry with the goal of making vehicles more intelligent, adaptable and safer. In Amazon's own fulfillment and transportation network, we're partnering to use AI to improve driver and worker safety and deliver faster and more efficiently, including with our new vision-assisted package retrieval technology coming in 2025, which will help our drivers quickly find the right packages in their vans. In testing, it saved more than 30 minutes per delivery route. We're also working together on advanced robotics and drones. And for Amazon Devices, we're working together to bring on-device AI to our customers in their homes on a variety of different form factors that we think they will find inspiring and useful. There is so much more to invent. Together, Amazon and Qualcomm have amazing opportunities to innovate for customers, and I look forward to continuing our close collaboration.

Good morning, everyone. Thank you so much. Good morning, everyone. Thank you so much for being here with us today. I'll tell you a little bit what to expect in this 2024 Investor Day. We're going to give you -- if you remember, we outlined a new strategy for the company back in 2021. And we came here to kind of walk to what we have done since then and what we're going to be doing next. Thank you so much for being here. I really appreciate all of you attending. So I want to start by highlighting what is our mission. Our mission is really to enable intelligent computing everywhere. We have been on this trajectory realizing that the technologies we have developed over the many years and continue to develop can be very relevant to a number of different industries beyond mobile. And that is the mission that we have been pursuing since we outlined our strategy back in 2021. And I think that creates an incredible opportunity for the company and a significant expansion of addressable market, and that's what we wanted to highlight to you today as well as show you how we've been executing. But the interesting thing is we have a new Qualcomm today. The Qualcomm today is very different than the Qualcomm from the past. And hopefully, you'll be able to see it as we expanded and became a partner to industries beyond mobile and automotive, in personal computing, in spatial computing. We're heading towards industrial. We have been doing that in networking and other personal computing devices. And the company feels different. The company is on this mission to generate growth and diversification. And we can do this by relying on our 3 technology pillars, which is about being the leader on everything wireless connectivity from cellular to WiFi, to Bluetooth, to GPS, with best-in-class connectivity. At the same time, being a computing company, I do believe and I said that back at Computex, with the launch of Snapdragon X Series for the PC, it was graduation date for Qualcomm as becoming a computing company. And that's based on our asset of high-performance, low-power computing across multiple computing engines from the CPU to the GPU, to the digital signal processors, to the NPU, to image signal processors and more, and also the unique ability to do intelligence at the edge. So I think that's part of you're going to hear as we're going to have many videos today. We're going to do this Qualcomm style. In addition to what we're going to tell you that we have been doing and will do, we want our great partners to also tell you how they feel about that. And 1 common topic of this presentation, it's going to be how AI is going to come at the edge, and that leverage our #3 pillar, how can we run AI inference at the edge with an incredible equation of performance and power consumption. So I want to go from here to our strategy back in 2021. So if you remember our Investor Day 2021, we outlined the new strategy for the company. We said that in mobile, we are going to be focused on the value share of mobile. We're going to be focused on Android, and we want Snapdragon to be the platform of choice for premium and high-tier Android smartphones. We also said we'll build a platform to be the preferred partner for the digital chassis of the entire industry. Number three, we will drive the evolution of the PC into a new platform. We will build the leading platform for spatial computing and be ready to power the digital transformation of the industry. So that's what we said back in 2021. So what I'd like to do is to walk you through how we have been doing and I go to the first three that we have started, there was definitely a science behind doing many things at a time, which I'll walk into the framework but I want to walk you to each and every one of you. So our strategy is really working, starting with phones. If you look what we did in phones, we generate on a market that is flat, sometimes grow single digits, sometimes have not yet recovered the market was still a bigger market when you think of the total mobile market before the pandemic. And we'll be able to generate 20% year-over-year growth in Android revenues. If you just look at fiscal '24. So it has been a market that we're incredibly proud how we have been able to grow ASP, drive silicon content and increase share on a market that is now growing. And the other thing that we have done, and I think the points in time that, that happened was just with the launch of the current Snapdragon 8 Elite that we did at Snapdragon Summit. We'll restore performance leadership to Android. Now the fast performance CPU, the best CPU in the industry, in addition to the GPU and the NPU is on Snapdragon Elite. How are we doing in cars? We have now a $45 billion automotive design win pipeline. We disclosed that back in July. 1/3 of that is driven by ADAS. We have now made the Snapdragon digital chassis, an industry asset. We're working with virtually every single OEM in the industry, and we have built the industry's most complete and scalable platform for software-defined vehicles. And then the latest addition of our execution strategy is what we did on PCs, and we're just the beginning of that opportunity for Qualcomm. Snapdragon X Series was the first platforms to enable Copilot+ PCs. We restore performance leadership to the Windows ecosystem. And just from when we announced it, which this is -- and I mentioned this in the last earnings call, we launched this in May. Between May and today, we saw a 2.5x increase in the number of design wins, which show that we're getting traction, and we're the beginning of the ramp of this incredible opportunity for Qualcomm. Spatial computing, we have invested in this early, and we believe we are about to see an inflection point of this enabled by Gen AI. We have built the platform, which has been the platform of choice of every single device. But more than that, we have built very long-term strategic partnerships with the key ecosystems that are going to drive this opportunity going forward. And then we have industrial. We see an opportunity to expand the SAM enabled by AI at the edge. And I think some of the messages you probably saw from Amazon, you'll continue to see it throughout the day is there is something changing that is actually creating this huge opportunity for AI at the edge. We look at this as the next wave of technology innovation in the industrial segment. We have launched a complete new portfolio within the quarter, which is the Qualcomm IQ Series for next-generation industrial IoT. And we have built a framework to scale. We took our time to understand how to build a channel for this opportunity, and we're going to provide details of that framework for you today. So in summary, this is how we feel about the incredible opportunity ahead for Qualcomm. We have put a strategy in '21. We're not changing our strategy. We've just been busy executing on that strategy, and it's working. It's working not only the new approach that we take to mobile, and that has been reflected on the big change from the time to now when you think of the operating margin of QCT. The creation of new business in creation of a leading position in each one of those business for automotive, for PC, for the upcoming spatial computing in the next wave of innovation and industrial. So what I'm going to talk about next is how are we doing this? And how do we feel that we have a unique position, not only to have executed on those opportunities since 2021 but continue to execute on them. So the first thing is our technology road map. We're incredibly proud in this company about the technology road map of Qualcomm. With those 3 pillars, I walked you at the beginning of the presentation. It is the industry-leading technology road map for both at the system level of semiconductors at the edge. And I want to highlight a few things of what makes this technology road map very unique. Custom design leading IP is not a collection of license IP. Each and every one of our IP is custom design is designed for the market with the cycle of innovation we want to deliver, and we build complete hardware and software platforms. As a matter of fact, if you look what we have done for every new opportunity beyond mobile, we have been building a software platform. Snapdragon Digital Chassis is a software platform. You're going to see that in both spatial computing, you're going to see that in industrial. We have a unique ability to scale our IP across opportunities. And I want to double-click on that one because the one thing that we have been able to do in the company as we prepare for this strategy. How could we scale our IP from a smartwatch all the way into what you do in autonomy, which is a data center type performance of a silicon. And how do we do this within the way that you actually create a leading platform and you do this within the spending envelope. So we have this ability to scale. We created the most differentiated road map. In every industry, we enter. We didn't enter each one of those new industries to provide a second source. And we have delivered on our operating margin target while investing ahead of revenues. If you think about it, we have been investing on a number of opportunities ahead of revenues, and we've done that still delivering on our commitments of operating margin. And we also took a very focused approach to add new capabilities. So that's the first part of our execution strategy. And I'll walk you through the second part next. But before I do that, since this is a growth in diversification Investor Day, we're not going to talk much about handsets. we're going to talk about really the growth opportunity of the company. So I just wanted to show you 1 thing on handsets as we move on to the next conversations. Snapdragon 8 Elite, we're incredibly proud of it. It's one of the most powerful processes we've ever done in mobile. It's now the industry leader in handsets across every performance category. The world's fastest mobile CPU, world's fastest 5G and WiFi technology, the fastest NPU. But I wanted to show you this metric. We have 5x the premium tier revenues on Android relative to the primary competitor. And I think we'll continue to build on this opportunity as we really focus on building Snapdragon. And the reason I made the comment about how we're doing everything with our spending envelope and deliver on our operating margins. If you think about what Qualcomm was doing before, what Qualcomm is doing now, we have not missed the beat on mobile. We have not missed a beat on mobile, but be able to build all those leading platforms across different industries, and that's what we'll continue to be doing. So with that, I get to the second part of the presentation. How have we been executing on these opportunities? And we created a framework for growth, and we've been following that diligently. First, we select an industry that we believe there is an attractive SAM and undergoing a technology transition where our technologies are going to be very relevant. Then we identify the inflection point happening in that industry. We'll develop the platform and acquire when we have to, and then we drive scale. That's the framework. And I'll walk you to what we have done today. So first of automotive. The inflection point was car was becoming a computing space. The car is now a computing space, like a phone is a computing space, the laptops of the computer space, the car is a computing space. Vehicles are becoming software-defined. And then you have the addition of assistant driving and autonomy. The technology assets will build a Digital Chassis. We did the acquisition of Arriver. So we have an ADAS stack. And we sell -- we did a very unique agreement in the industry, which is the co-development of ADAS and autonomy with BMW. We're now moving, as you know, to scale is being reflected in 5 consecutive quarters of record revenue in automotive, 45 billion design win pipeline translating into revenue as we have new cars launched. We're working with all major OEMs and Tier 1s and we're going to expand now to 2-wheelers and micro mobility. We also apply the framework to the PC. We saw the inflection point. It was conversions of mobile and PC, which has actually started to be done by Apple. The competitive landscape change between the Windows and Macs. We saw that as an opportunity, especially as the ecosystem did not have confidence in the existing players to actually deliver a solution that will restore leadership to the Windows ecosystem, and we had Windows transition to AI as part of the Microsoft transition to AI. We built Snapdragon X Series. We acquired talent from NUVIA. We deliver a whole new Qualcomm Oryon CPU, and we have the industry-leading NPU. A lot of people ask me about an exclusivity agreement with Microsoft. The reason when May Copilot+ was only launched as Snapdragon because actually, it could be the only platform that would run it. It has nothing to do with any exclusivity. Then we're building now scale and reach. Global retail and commercial channel presence, we will talk about that in detail. We have a strategic Qualcomm and Microsoft go-to-market agreement, and we have now 58 design wins launched or in development across leading OEMs. So we've been preparing for this moment. And I think we have following the framework, and we think that this business is now well positioned to start being material from a financial standpoint. The next one, it's what is happening with XR. The inflection point has always been the merging of physical and digital spaces. But now we have something new. And you started to see that materializing with the AR, the Ray-Ban smart glasses of Meta. Smart glasses are becoming a personalized wearable AI. The technology assets, we purposely built SoCs for XR. This is where power is incredibly important. But we don't only did silicon. We did a lot of software. We develop our suite or perception technologies. And we came up with the ability to do the battery life that is necessary for this very small device. We're now scaling to strategic long-term collaborations with the ecosystem that's going to drive scale. And we have now every single design win across everyone that is building those devices. And the next one is what we're now in the process of doing, which is the new industrial. We believe that end-to-end industrial platforms are going to fundamentally change because of edge AI. So that's the inflection point. We have a new wave of transformation at the edge and Edge AI changes the equation from a cost perspective, from a personalized perspective and some -- and from a user experience perspective. The technology assets. We have broad connectivity and processing road map. We actually have augmented some of that connectivity recently with acquisitions. We have built a special design industrial platform, which is the Qualcomm IQ Series. And we have been building a software and cloud platform, some organically developed, some through acquisitions. And to build scale and reach, we're identifying halo customers that were codeveloping this solution like we did with Alto. We've built this entire framework for IoT, and we're building strong relationship with IT system integrators as well as distributors. And with that, we're starting to see that Qualcomm is becoming increasingly relevant to a number of different industries. I remember in the early days when we talk about ecosystem and we list all the operators in the world, we list all the infrastructure provider in the world will list the test equipment providers and the mobile OEMs. It's very different now. We're becoming relevant to a number of different industries. And again, the story of Qualcomm, which is always a story of collaboration and joint innovation. That picture is fundamentally changed, with the number of partners in all of those new industries that is going to be part of the Qualcomm growth and diversification story. And with that, we feel that we have a unique position. Qualcomm has a unique position on the edge for an expanded TAM. By 2030, we now have visibility of our TAM to be to the order of $900 billion, $50 billion connected edge devices is our projection of cumulative shipments for that entire industry from 2024 to 2030. And that is the opportunity that we have been executing on. So as we get to the end of my presentation, I want to talk about AI at the edge. And I think many of you had asked through the many quarters as we go to this journey about how we think about AI, how we think about AI monetization. But one thing is very clear to us. In that strategy that we have been executing AI is the best tailwind to that strategy. It's really accelerating. It's a generation opportunity for Qualcomm in the role of the edge is one that we're pursuing with the ambition to become a leader. And it is scaling very fast at the edge. The reason that happens is performance, efficiency, privates and security, low latency, reliability, the ability to fine-tune models for personalization, the change in the user interface, there's often a human on the other side of an edge device and costs. And that opportunity is going to make a difference for each and every one of our business where we're focused to grow and diversify the company. We built a platform in mobile. But now with Gen AI, we believe we can further differentiate. We're uniquely positioned to have Gen AI in automotive. When we announce our new platform, and we're going to provide you more details, there was a 12x improvement in AI capability to feature-proof cars in a lot of the Gen AI used cases that come into cars. And for PCs, for spatial computing for industrial, it is the inflection point for those businesses. Starting with the transition of Microsoft, the Copilot+ PCs, the personal assistant in the change in enterprise digital transformation with AI at the edge. And we believe we have all the assets to be a leader in the segment because we are the incumbent for the edge. We have custom design IP that can be optimized in each and every one of those opportunities. We have the best-in-class performance in our DNA. Our company DNA is always going to be performance per watt. We are doing AI integration with software in the platform in a way that is incredibly novel and easy for developers at the edge. We're going to present that later today. We're building an ecosystem. We have a road map of own device AI, and we have the scale. So that is the reason I'm saying this is an incredible tailwind for the opportunities we have at the edge, and we're focusing and executing on those. And as I get to the end of my presentation, I wanted to leave you with some messages. We're successfully executing against our diversification strategy. Gen AI accelerates demand for technology. There is a significant opportunity for growth across all the target industries. We're building a very strong ecosystem of new customers and partners. And this is what actually differentiate Qualcomm from many other semiconductor companies, especially at the component level. We build very unique and strategic relationship with many of our customers, and we're delivering on our operating margin target while investing ahead of revenues in all those opportunities. And the goal is to be continue to transform Qualcomm into a diversified growth leader in the industry. Thank you so much for the opportunity. I would like to introduce you to Durga, who will speak next. But before please watch this video from Satya Nadella. Thank you.

Hello, and thank you so much for having me at your Investor Day to talk about our deep and long-standing strategic partnership with Qualcomm. There's no question we have entered a new age of computing. Up until now, scaling laws have helped us build and serve powerful language models in the cloud. But today, we are going beyond the cloud to the edge, removing fundamental constraints of power and space, reducing latency and always ensuring privacy and security. This will require collaboration and innovation from cloud to the edge to deliver these breakthrough solutions and improve people's productivity and creativity. And that's why our long-standing partnership with Qualcomm is so important. And we are working together to power intelligence at the edge across devices and industries. And we are so excited about Qualcomm's innovation across mobile, auto, industrial and of course, the PCs. In May, we announced Copilot+ PCs, the fastest, most intelligent Windows PCs ever built, including new devices from across our OEM ecosystem. which were powered first by Qualcomm's industry-leading Snapdragon X Series processors and NPU. And we have been so energized by customer feedback. And this is all just the beginning, right? Thank you again for the partnership. We look forward to working closely with Qualcomm to shape the future of computing in the years ahead.

All right. Well. Okay, over the next 15 minutes, I'm going to talk about a few things. I'm Durga Malladi, I run our technology road map planning across all of our business units in Qualcomm. And in the next 15 minutes, I'm going to talk about 2 things. First, why is generative AI transitioning to the edge? There's lots of reasons for that but I want to make sure that we all understand why is that happening? And second, why is Qualcomm uniquely positioned to drive this transition and drive generative AI inference at the edge? Two concepts, we'll go over it step by step. First things first. AI is fundamentally changing how we use our devices today, whether it's handsets, automotive, PCs, XR, networking, all kinds of devices, consumer and industrial IoT devices. And there's lots of reasons why this is happening. But the first question that comes to mind is why? Why do you want to run generative AI inference on a device? Surely, you can just run it on the cloud, nothing wrong with that. Well, the first level answer to that is because we can. If you just take a look at the enormous computational horsepower that we have in devices today, it's capable of running some very serious generative AI inference workloads in a very energy-efficient and high-performance computing manner. The other part is when you have a device with you, whether it's a PC or an IoT device, you want to make sure that you have the same kind of a response or immediacy regardless of connectivity to the cloud. You don't know how the connectivity is going to be but you want to make sure that it's agnostic to that. Increasingly, enterprises and consumers are interested in a far more personalized AI experience, relying upon data that's located within the device in itself, which basically leads to concerns and more importantly, making sure that we have the right level of privacy and security with the data that's located there. So the generative AI can tap into that data and provide a more personalized data experience. The other part is as you kind of think through and I'll talk a little bit more about how the user interface itself is changing in a large number of these devices. And of course, every consumer, every enterprise wants to keep a tight leash on the cost in itself. We're going to go into a few of these KPIs in more detail. Before that, I want to actually step back in time. November 30, 2022, 2 years back. That's when ChatGPT was introduced. And I think that was about the time that everyone started waking up to a new reality where there was this almost human-like responses coming back from what seemed like a chat bot. I think by Christmas, everyone was talking about it. And that was based on a model called GPT-3, which was further refined in March of 2023, and it had a certain model quality but the definition of a large language model, that's a large part, it was 175 billion parameters. Here's a fun fact. In about 18 months, we have a model that was introduced. It was in the Llama family of models. It's 120th the size, 120th, 8 billion parameters and its model quality is better than what we saw just the year before. In summary, what run -- what can run on the cloud last year, you can run on a device this year. Is it just a one-time trend? Not really. This year, a new model was introduced, GPT-4o. Now this is multimodal. So it's not just text. It's a combination of text and voice and all the other modalities. Already in just a few months, another model was introduced, Llama 3.2, which is now beginning to be not yet in the same performance, but it's getting there. Give it a few months, we'll get there. The trend line continues. Now we don't have to call it as an AI law, but there is a clear trend line over here. The model sizes are continuously decreasing, while the quality is improving. Why does that matter? Because you can run all of these models on devices that you and I have in their pockets today. That's a pretty important trend. And when you put it all together across a very large number of model creators, not just we talked about GPT and the Llama family, Google Nano, another model, Ministral, which is introduced by Mistral family, the monotonic increase in the model quality continues. What that really means is that we can build some compelling use cases with these small models on devices with the model quality that you expect as to what is running on the cloud. Now there's a second KPI that we should take a look at. Let's say that all we do is run generative AI inference solely on the cloud. Third-party reports have indicated that the cost of running generative AI inference just on the cloud alone, it's becoming cost prohibitive. Increasingly, it's becoming cost prohibitive. It's kind of hard to scale in this way. There are a few numbers that are listed out here. These are from third parties. But if you just think about what we said in the previous 2 slides, surely, there's a better way of doing things. Instead of running just on the cloud, we can take generative AI inference distributed across the network and run it on devices. What this really translates to is a much smaller marginal cost of running generative AI inference on devices but it also allows a new paradigm. Train on cloud, run on Qualcomm. And what kind of models are we looking at? There's a very large diverse set of models that are available today. You can go to Hugging Face and pick any one of the millions of open source models that are available there. I think it just crossed about 1 million models. There are proprietary models coming in from different domains, from different OEMs. There are models, which are coming in from different regions. Every region is investing in their own language models, preserving the culture and heritage of that region. And there's a lot of classical AI models that are still out there. From Qualcomm's perspective, we are in the business of running every single generative AI model on the planet on our platforms. We want to make it easy to be able to do that. That's the job, that's ahead of us, and that's what we are in the midst of. Finally, when you kind of take a look at all these models and we talked about the model quality, different kinds of devices have different capabilities in terms of the size of the models that they can run but there's more that's coming in from a complexity standpoint. Much larger context length. What that means is you want to give more thoughtful responses, thought answers to anything that might be running. In addition to these thoughtful responses, the number of modalities goes beyond just text. We talked about text, voice, images, video, radar, LiDAR, infrared sensors, all these modalities coming together and a far more personalized experience. In a nutshell, there's a lot going on within this technology itself but we are in a position to be able to do this on devices today. What kind of applications are we looking at? Well, there's quite a few of them. There's consumer-centric applications, which are like content creation, image editing and so on but enterprise applications, code generation, e-mail generation, document translation. There's a lot of these key use cases that are beginning to emerge in PCs, in automotive, in XR devices, in handsets and in IoT devices. In addition to all of these use cases, something else is happening. We are in the midst of creating a new user interface to all these devices that we know very well already. And when you got to think about user interface, just go back in time, in the '70s, command line interface. That was the user interface back then. In the '80s, we used a mouse and a graphical user interface. By the late 2000s, we started getting into the concept of apps, all of us using tactile fingers with a clutter of apps that are there in all of our devices. But now we're in a very different era where AI is now emerging as the new user interface to all the devices around us. How does it work really actually? Well, it starts with the fact that take any device that you want, and there's a lot of input that comes in, whether it's you talking to your device, whether it's an IoT or a PC device or a handset, along with the local context that comes in from the sensors. The first top, your first interface to that is an AI agent. It's the 1 single entity that's inside your device that's processing all of this information before you even take a next step. First, it looks into a personal knowledge graph, which defines you as the end consumer or the end enterprise user, knowing all of your preferences, tapping into the local company confidential information or your own private information. Then picking the right set of models, there's so many to pick from. But for a given test, there are some specific models that might make sense, and in the end, on the back end, tap into the applications. With this, you can easily see how AI is now that AI agent is the user interface that most consumers, if not all, will be able to see. With that said, I want to go into the second part of the conversation, and that's about what is Qualcomm doing in this space. Now we have scale at the edge across all the platforms that was mentioned earlier. But what we have done is build up a large amount of foundational technology that allows us to drive Gen AI at the edge. We'll talk about hardware, the right software stack, AI agents, all the tools, making it easy for developers to build applications while at the same time, making sure that we are as agnostic, completely flexible so that we can utilize the same stack across all these different product categories. Let's take a look at hardware first. Very critical to have the right kind of CPU, GPU and NPU running generative AI. We have built -- this is based on custom IP, the best-in-class high-performance, energy-efficient computing with our CPU, very well-tailored for immediacy when you have a very low latency. Our GPU that is used for any kind of a graphics-based application. And our NPU, which is custom designed from ground up for generative AI applications. It's very unique in the industry in that sense. Well, is that it on hardware? There's a bit more. In fact, there's a lot more that is needed for sustained leadership for Gen AI at the edge, and we've built up that IP over the years. Thermal design, packaging technologies. You got to be good, not just at one of the processes that was listed earlier but all 3 of them and be in a position to harness the power of all those processes together, which we call as heterogeneous computing, so that we can drive these Gen AI applications. Chiplet integration and memory bandwidth, high memory bandwidth, these are some key IPs that are absolutely critical. Not all of them might be relevant in the cloud but at the edge, you need them to lead, and that's what Qualcomm has built. A second important part in terms of building up Gen AI applications is the software. Now I said earlier that we provide our software in such a way that we can scale across all the platforms. That means pick your framework, any framework that you want, PyTorch or whether it's TensorFlow, Keras, ONNX, pick any runtime. We provide all the libraries, all the tools that are necessary for that. And at the same time, this is a system-level software so that you can build solutions on top of it. So it makes it very easy for OEMs and developers to train on the cloud and run on Qualcomm. We talked about our AI agent earlier. Qualcomm is building our own agent AI orchestrator, which relies upon those 3 key pillars: A personalized knowledge graph; the ability to tap into the rest of the ecosystem; the applications and all the partners and at the same time, pick from the right set of models to build a rich and compelling use case. This agent AI orchestrator is part of our software solutions and tailored and customized for any of our customers out there. A few months back, we introduced the Qualcomm AI Hub. Now this is a very unique offering in the industry. There's nothing that's anything like this in the industry. It's a place to be if you're a developer for Edge AI applications. There's a rich set of tools, software, a device farm that's available, which makes it easy for developers to go from ground up all the way to applications running. In the first step, you can come in and pick a model. If you don't like one of the models that's already out there, bring your own model, we curate it for you, so it runs efficiently on Snapdragon. If you don't have a model, bring your data, we'll create the model for you, giving you full flexibility. Pick a device and as you start writing -- those devices can be PC, XR, IoT, automotive, handsets, as you write your application, we'll give you cloud native access to a device firm where you can test these applications on physical devices, not emulation, physical devices and deploy it anywhere. And when we take a look at what has happened for AI Hub over the last few months, more than 1,000 companies have been using the AI Hub tens of thousands of jobs that are running in parallel. We work with everyone in the ecosystem. You heard earlier about how we are making it easy to use AWS SageMaker to train and run on Qualcomm and working with all the other model creators everywhere on the planet. In effect, it's a one-stop solution for any generative AI edge running -- applications at the edge. In summary, I want to come back to where we started from. Edge AI is inevitable, and it's driving -- it's going towards the edge. And Qualcomm is leading Gen AI and making it very easy for any developer to train on the cloud, run on Qualcomm. Thank you. And before I leave, I want to welcome to stage, Alex Katouzian. And just before that, we're going to see some demos.

Hello. It's great to be with you all today. We're excited to be working with Cristiano and the entire team at Qualcomm as we embark on the AI revolution. We're entering a whole new era of productivity and efficiency, and it's reshaping the PC experience. Our strategic collaboration with Qualcomm is powering our exceptional Snapdragon X Series AI PCs. Together, we're delivering extraordinary speed, groundbreaking battery life and incredible AI performance in the workplace and the edge. We're making the AI PC a true digital partner for everyone from SMBs to creators, to developers and designers and enterprises and everyday users. The PC was made for this moment, and we are leading the way in our partnership with Qualcomm. Thank you.

Thank you for inviting me to share a few thoughts to your event. Our partnership with Qualcomm has already brought so many remarkable technologies to market from PCs, tablets and smartphones to mix the reality platforms, smart collaboration, edge computing and more. This year, Lenovo teamed up with Qualcomm to launch a range of Copilot+ PCs powered by Snapdragon Series processors. These amazing products are accelerating the AI device transformation in our industry. Even more, with Lenovo AI PCs powered by Snapdragon, we are unlocking new growth opportunities globally across consumer, enterprise and the SMB sectors to deliver the next level of performance, productivity, personalization and data protection to our customers. And we are particularly excited to partner with Qualcomm to create this new category of products in China market. Finally, I want to thank you for your support in Lenovo's Tech World event. Together, let's continue to build our future of smarter AI for all.

Hello, everybody, and thank you for inviting me to join Qualcomm's Investor Day event. It has been a milestone year for our partnership. We launched a new generation of AI-powered PCs built on the Snapdragon X Elite processing. Together, we are not just driving innovation. We are restoring leadership to the Windows ecosystem. Looking ahead, we see great potential, especially in the future of work. Today, workers want more fulfillment and flexibility, while businesses need to innovate, grow and stay profitable. Bridging this gap with smart technology creates a unique market opportunity and a way to redefine how people relate to work. Our Copilot+ PCs powered by Snapdragon X Series processors are already making a significant impact. These devices are more than just PCs. They are powerful tools with impressive battery life and on-device AI capabilities. This enables our customers to perform their work with immediacy and security, ultimately, amplifying productivity and creativity. And this is just the beginning. We share an exciting vision for the next generation of computer. As AI evolves, we see even more ways to partner across consumer and commercial applications pushing the boundaries of powerful seamless experiences. We are incredibly grateful to have Qualcomm as a strategic partner and look forward to all that's next.

One of those great testimonials. I really love that. Thank you. One of those great testimonials. I really love that. Hello, everyone. I'm Alex Katouzian. I'm the Group General Manager for our Mobile, Compute, XR and Wearables businesses here at Qualcomm. It's great to be here with you today. I'm very excited to talk about our success in both the PC and the XR markets and talk about our strategies that will help us scale in both of these growth businesses. So let's start with the PC. So PCs are undergoing a significant transition. There are 2 major anchor points. As you heard, on-device AI is transforming the human-to-machine interface. Multiple models pervasively are running in the background from a more intuitive and interactive agenetic experience. With our 15-plus years of product experience, our Hexagon NPU is leading this change. High performance and long battery life are top requirements for all consumers. And with over 25 years of experience of developing high-performance compute cores at extremely low power, our custom Oryon CPU is leading in this market. Furthermore, all the best attributes of smartphones have already transitioned to the PC and no one, no one knows how to do this better than Qualcomm. The AI PC has arrived and is based on the Snapdragon X Series combined with Microsoft's Copilot+. Microsoft trusted us to restore leadership back to the Windows PC and help them lead in this transition. Copilot+ was first compiled and tested on Snapdragon. That means all user experiences, including passive pervasive on-device AI with over 40 different models as well as performance and power efficiencies were all optimized on the Snapdragon X Series. Microsoft announced their 2024 surface Copilot+ lineup that were powered by the X Elite has the highest performance, longest battery life laptops they've ever made. We predict that by 2029, there'll be 100 million notebooks, greater than $500 per year that will be Copilot+. This is a great market for us to expand and scale into. The X Series platforms are engineered for leading performance per watt. Multiple high-performance computing cores at lowest powers have been integrated into these X Series SoCs. But we also designed multiple peripheral chips that go around these SoCs. For example, all wireless communication solutions, charging ICs. And most importantly, the power management ICs that are tightly coupled to these SoCs to design the most optimal power distribution network for the entire system. This is resulting in examples such as 22 hours of video playback on 1 charge or 15 hours of continuous browsing, all on thin and light designs. On-device AI models are running pervasively but they're only taking minutes of battery life for the entire day of use. We now have a broad road map of category-leading platforms. We entered this market in June of 2024 at the premium tier. What that means is our OEMs and our channel partners and the ecosystem trusted us to differentiate their products at above $1,000. No one has ever done that in the PC market before. Our roadmap lineup will soon cover tiers from $600 and up. We've now scaled up to 58 designs in production or development for 2025. And we're going to have 100-plus designs allowing us to access 70% of the TAM by 2026. This entire lineup is pin and software compatible. What that means is it allows us our OEM and ODM partners to quickly scale into multiple tiers. Every device has the exact same AI user experience because we use the same NPU core across all of these tiers. Performance per watt leadership now belongs to Snapdragon. We introduced our X Series in the fall of 2023 and at that point, we handily beat our competitors in performance and power dissipation. Up and to the left is what's better on this chart. So we waited an entire year for these competitors to show up, and they're still not good enough. Please note that this Intel device is designed in 3-nanometer process technology, and it is good enough to beat AMD's 4-nanometer devices, but still not good enough against our 4-nanometer X Elite. We have the best performance per watt CPU architecture and design regardless of the process technology. This benchmark was run with the PC plugged in, the chassis open, air blowing on the circuit to provide the least thermal resistance with the highest performance. But the whole point of a new AI PC is that you want to be unplugged, is you want to work from anywhere and be free of the charging cord. So what happens to the performance when running the same benchmark unplugged? The performance of our competitors fall between 30% to 45%. And there's absolutely no drop on Snapdragon. That means you get the same exact performance and experience when you're on the go with Snapdragon. We improved the performance and battery life of our cores every year. And now we have advanced to our second generation Oryon CPU. We introduced the Oryon Gen 2 in our premium mobile platform, let me repeat, in our mobile platform in October 2024. We have improved 30% in performance and 57% in power. And obviously, this core in mobile is much higher performance than our competitors in the PC. In our next product family of PC system solutions, we're going to be moving to the Oryon Gen 3 with even better performance and better battery life. And is currently in development and will be introduced in 2025. Remember, there are 2 main anchor points for the PC inflection. High-performance and long battery life and on-device AI performance and user experience. When we run this benchmark, unplugged, our competitor cannot match our performance. Their highest performance point is below our lowest but at 311% more power. Now imagine the battery drain and the user experience degradation with pervasive AI models. I'll wait till you take your pictures. Go ahead. But you have a picture of me now against the slide. Okay. Let's talk about software compatibility, which is one of the most important requirements for consumers. Based on Microsoft's telemetry data, which is extremely accurate, 90% time that you're in front of your PCs, you're in apps that are already native on Snapdragon. What that essentially means that every app that you need is running native with great performance on Snapdragon. We've built our graphics subsystem for best-in-class productivity PCs and casual gaming for thin and light PCs. Even if you've been productive 10 to 12 hours in the day, you still have ample battery life left over for gaming. We have over 1,300 games running and over 700 of them at 1080p 60 frames per second, which is a great user experience. This roster of games grows every day. Let's now talk about our go-to-market focus areas. We just established that Snapdragon has the best technology to disrupt the new AI PC. Our partnership with Microsoft will allow us to lead and scale Copilot+. So let's now talk about how we get this technology into the hands of the customers with our focus on marketing and brand investments and our engagement with retail, commercial and enterprise partners. Snapdragon is the #1 brand in consumer preference for Android smartphones globally. We are extending that investment to build more brand awareness and preference in PCs. We are partnered with extremely popular franchises such as Manchester United and Formula 1 who have billions of fans. The Snapdragon brand is prominently placed for all viewers. We're also closely partnering with Microsoft to co-promote the Snapdragon brand alongside with Copilot+. What this is creating is billions of impressions for consumer brand awareness. We're investing in educating the consumer through key opinion leaders and influencers, targeted educational videos and content and carefully placed ads. And we also have campaigns to increase Snapdragon preference in PCs. We're also working very closely with our partners in the channels to draw consumers in and help them buy Snapdragon PCs. Let me talk through our retail channel strategies. We're concentrating our investments in regions that sell the most amount of PCs. In these regions, we have 95% coverage of all our products today across 9,300 stores. We've trained over 100,000 people in those stores on the values of Snapdragon and how to sell them in those stores. We're partnered with key market makers in these regions that have a domino effect on other parts of the retail channel. These partners include Best Buy, Costco, MediaMarkt, Currys, JD.com and many more. Let's go through in detail about everything we're doing with one of the big retail partners Best Buy. We are now in 100% of their stores in the U.S. We've trained over 30,000 employees and Best Buy is now carrying the largest assortment of Snapdragon devices. Activities in-store, online and training include weekly dedicated sales training, targeted advertising to Best Buy customers and back-to-school and holiday campaigns. Let's look at our commercial channel strategies and activity. We are working with the top 40 commercial channel partners in the world. These partners are also market makers that have a domino effect on many other companies in the channel. We're also partnering up with 500-plus top global enterprises that are currently testing Snapdragon or deploying Snapdragon. We take multiple examples off of these enterprises and feed them back into the commercial channel partners to help promote Snapdragon across smaller, medium-sized businesses and other enterprises. So let's look in detail of what we're doing with one of our key commercial partners insight. We've now trained over 500 salespeople at Insight. In turn, they're engaging with over 2,000 businesses. We have activities across sales and technical team training and Snapdragon campaign activations. This includes joint customer engagements with top enterprises, awareness and preference through digital media, and ongoing regional customer events with OEMs. These are some of great testimonials from our enterprise engagements that we're feeding back into the commercial channel. We used Insight as a large enterprise and their Chief Digital Officer, Rob Green describes how performance, battery life and on-device AI of the X Series is driving employee efficiency and productivity. Citibank's CIO, Balaji Kumar, is highlighting how performance and secure on-device Gen AI is saving them time and money with cogeneration on the device and offloading their cloud infrastructure costs. So let me summarize my key points. Snapdragon is the platform of choice for the new AI PC. We have sustained differentiation in performance and battery life. We have proven software and app compatibility. We're establishing our brand with investments across the marketing funnel. And we're scaling volumes with lighthouse retail, commercial and enterprise partners. Now let's take a look at a quick video of our key go-to-market partners. Please enjoy.

The greatest part about this collaboration is that I think all of us started with the customer. And that's not always the case. Sometimes it becomes very technology specific or it becomes very specific around how something looks this time, I believe strongly we started with what will matter to consumers over time. And the irony of this technology is it's actually going to make all of our hardware feel more human. And that's actually something we all can agree on. And what's beautiful is where we take it from here, over time, this is just technology that gets smarter and more personalized and does exactly what you want it to. And I think that is a rallying cry for all of our teams because I just firmly believe over the next 18 to 36 months, this completely changes the way we interact with our computers, and I love to be on that journey with Qualcomm. It was amazing to me how quickly Qualcomm and our teams were able to collaborate together to bring this new technology to life. And I think it really bodes well for the future and what we're going to see from here.

Good afternoon. It's wonderful to be part of Qualcomm's Investor Day. At Citi, we're proud to have partnered with Qualcomm Technologies for nearly a decade. Now this is a relationship that takes various forms, where Citi acts as both a global financial services provider and of course, as a customer of your technology. And it's this facet of our partnership that's really been integral to our efforts to modernize our bank's operations and ensure we have the agility to deliver for our clients in the digital age. With the Snapdragon X Series platform, Qualcomm's approach to PCs is much more aligned to the way we actually work as a global bank, always on with all-day battery life and capable of handling a variety of workloads. The Snapdragon X Series laptops are also playing an important role in our strategy to make our data centers much more efficient. Through our partnership, we are transforming the way people work and the quality of service that we're bringing to our clients. So thank you, Qualcomm.

Another great set of testimonials. As Cristiano said, we do it the Qualcomm way with all these videos. So let's turn to XR right now. XR will be the new computing platform and as personal smart device. For many of us, the PC was our first computing platform, and we're still very productive on them. Smartphones were introduced and became another computing platform. We do many tasks on our smartphones. But still, we use our PCs. And now XR will be a new computing platform and another personal smart device. Let me walk you through the 3 categories of extended reality or XR for short. VR or virtual reality is all about immersion. It will make you feel like you're actually in the app itself. For example, in a game or watching a concert or a sporting event would like the best seat in the house, personal tutors and trainers right in front of you. With mixed reality or MR, you have the immersion advantage, but you also get to view what is around you. So that way, you can interact and collaborate with others in the environment. AR or augmented reality is a perfect way to interact with the real world. The glasses see what you see. They hear what you hear. And it's a very familiar form factor that lets you naturally interact with the environment around you. With integrated AI, we believe augmented reality will power many use cases that are very intuitive and personal. Let me walk you through some of these cases. You're a busy person on a daily basis. You want to know where -- what your next schedule looks like, where is your next meeting. You simply have to ask and I'll display it for you. You're riding a bike. You feel like having some coffee. Simply ask, "Take me to the nearest coffee shop, show me the shortest route to get there". You're on vacation in a historical place. You look at a building, you simply ask, "Tell me about the history of this building". Or you're an avid runner, you're on a business trip, you feel like exercising and ask, "show me a 5-mile route near me and display my health metrics while I'm running." And there's hundreds more applications like this. And with so many practical examples, we believe people will carry AR devices with them every day. We also think that on-device Gen AI will scale MR and VR. In health and fitness, imagine you have a personal coach who understands your needs and trains you exactly the way you want it. Imagine you're playing a game or inside a game. But you're not familiar with the terrain. You're not familiar with how to play it well. You call a non-player character or NPC, as an assistant. And you ask, how do I enter this building? Show me the terrain. Where do I pick up more swords? How can I put some more skins on these swords? All done through on device AI. A personal tutor ask them to walk you through a problem or see a 3D model that you would never otherwise experience in a book. In the workplace, you can actually work and design and collaborate around 3D models that you can modify and improve in real time. So we think on device, Gen AI will scale MR and VR. We just showed our sustained performance per watt advantage in the PC market, and you already know we're leaders in smartphones. But remember, PCs are operating in tens of watts. Smartphones are operating in low single-digit watts. But in the case of XR and particularly for AR, we can only operate in the milliwatt envelope. With our proven track record, it would be extremely hard for anyone else to catch up with us in XR. We are the leader in key technologies purpose-built for XR. We have the best cores in the industry across all computing, multiple cameras, AI, computer vision and audio, which are critical for XR headsets. These cores are combined with an extensive set of perception algorithms. Today, the way you interface into your device is typing, using a mouse, clicking, expanding, pointing. But the tracking of your head, your hands, your eyes, your gaze and many more algorithms in XR devices help you interact with the environment around you and are inputs to your devices. All of these algorithms -- there are many more that we haven't listed, but all of these algorithms are developed and optimized over years at Qualcomm. Over time, many of these algorithms are hardened in silicon to get the best performance per watt experience that we can get. Why? Because we're operating in milliwatts. We have the most complete road map for VR, MR and AR. No one else has been able to duplicate this. And furthermore, our Snapdragon Spaces developer platform allows these algorithms with APIs to reach thousands of developers. This forms a continuous feedback loop so we can perfect these algorithms. Spaces works and supports multiple OSs in XR. And now Snapdragon XR has become the platform of choice in the ecosystem. Let's talk about our primary partner in XR, Meta. They have partnered with us to deliver purpose-built solutions for their entire XR product line. Together, we are co-defining multiple generations of VR, MR and AR solutions that are the most successful in the market today. And to enhance our solutions, we're also optimizing on-device AI with Meta's Llama small language models. Every big tech OEM and content company, creating and pushing XR forward chooses Snapdragon XR solutions. We are the preferred choice for the spatial computing ecosystem and we have multiple strategic long-term collaborations with key partners. So let me summarize the key points. XR is the next computing platform. Gen AI will drive scale in XR. Snapdragon is the clear leader in all XR categories. We are the long-term partner and preferred solution provider for major tech OEMs and developers. And AR is another smart device that you will carry daily. Let's watch a quick video for Mark Zuckerberg, the CEO of Meta on our XR partnership. And after that, you will hear from my partner and good friend, Nakul Duggal, about our auto and industrial IoT businesses. Thank you. [Presentation]

Good afternoon. How are you all doing? Very exciting to be here with you again. I saw you for the first time in New York City in November of '22 when we had our very first automotive Investor Day, and I hope many of you have become believers in the automotive story at Qualcomm. Thank you for attending today. I am responsible for the automotive, the industrial and embedded IoT and the cloud computing businesses at Qualcomm, so I spend a lot of time focusing on our diversification mandate. Today, I'll share with you our progress in automotive, our road map strategy, the role of Gen AI. The focus that we have had on driver assistance and automated driving, the investments that we've been making and how they will shape more than the remainder of the decade. And then in Industrial IoT, I will share with you how we are redefining our road map to target this market, which is going through a tremendous amount of disruption as AI transforms industries. From 2012 until today, as I look back, we've been very disciplined in the focus that we have had on building a long-term strategy for automotive. Our strategy was built into a market that was foreign to us. We started with a modest connectivity stack, which we built upon brick by brick. Since then, we have exceeded our financial commitments. We have re-architected the automotive compute platform. We've redefined the supply chain from being component-based to being value-based. And we've made the car a digital lifestyle product. Most importantly, we have done this by partnering and enabling the ecosystem. Tier 1s, Tier 2s, hardware and software partners and, of course, OEMs globally. We believe we've played a small part in redefining the future of the car. We established the Snapdragon Digital Chassis now ready with AI and defined it on the following key principles. The compute fabric that goes across automated driving and cockpit that tiers across all trim levels of vehicles. That has the highest standards of safety, reliability and quality. That is designed with the depth of understanding that hardware and software design are interdependent. And what it takes to support a product that has a life cycle of over 10 years that is deployed globally. Not many companies know how to do all of this. Qualcomm excels at such system-level problem statements. These require a tremendous depth of understanding of architecture, of silicon, of process nodes, of IP, of hardware software co-design, how real-time operations coexist alongside consumer-grade software. How these complex systems are designed, how they scale, how analog and digital domains interact, how many tech ecosystems have to coexist and how do you support such long-term complexity over the long run. The software-defined vehicle requires tremendous versatility, agility and options for the automaker. It creates a differentiated experience that allows automakers to upgrade, to reinvent. The chassis that we are building is now AI ready. It is open and programmable. We support every operating system flavor, every digital ecosystem, every configuration and, of course, Edge AI. We assume that the workloads that will run on these SoCs will be mix critical and independent of the underlying architecture. AI has become a key ingredient of every edge and cloud interaction. Being able to host models at the edge and orchestrate hybrid transitions between edge and cloud is essential for latency, for privacy and overall cost of inference. Finally, the pre-integration of all leading digital tech ecosystems is one of our biggest differentiators. All with deep safety and real-time software integration for certified and secure implementations that work out of the box. We have stayed consistent to the strategy that we laid out at Investor Day in '22, and the market is embracing our architecture. Last month, we announced the most advanced automotive SoC family for central compute at our Snapdragon Summit in Maui. The Snapdragon Cockpit Elite SoC for luxury in cabin experiences and the Snapdragon Ride Elite SoC for automated driving are new set points in automotive performance expected to start production in 2026. The Snapdragon Elite family is a new class of processing altogether. To design such a complex SoC in an advanced process node requires deep experience with quality, safety and reliability, something that we have uniquely mastered. The Elite Series is a foundational automotive safety fabric, which means that any workload can process and inform both non-safety as well as safety oriented domains. The Elite has massive AI capabilities. For Cockpit applications, we'll be able to support 30 billion parameter models locally. The NPU is 12x more capable than its predecessor. And the Elite Cockpit experiences are now powered by the most powerful safety-grade Oryon CPU. The Elite is designed for multimodal Edge AI, and I'll give you some examples of that. For automated driving applications, the Ride family features an architecture for real-time multisensor and planning networks that run with performance and accuracy at a power budget that is 2x better than our nearest competitor. The camera pipeline is designed for multimodality and the Elite will support over 40 concurrent multimodal sensor inputs to create an environmental model of the world. With Elite, the convergence of automated driving and in-cabin experiences will drive yet another cycle of innovation powered by Gen AI. Let me give you an example. I want to show you where the next level of Edge AI will take us. Imagine the scene, with Elite, the sensors in the car absorb the environment and process it through many layers, the automated driving domain focuses on creating a model that allows the car to determine its driving trajectory through a combination of end-to-end AI networks that have been trained on massive data lakes and optimized based on rules-based learning. The copy domain is absorbing the same environment, parsing it to multimodal inputs. Is the driver looking for parking, for coffee, a point of interest. AI models assist the car in processing the scene to enrich the experience, making navigating more intelligent. Imagine you are sightseeing in a new city, and your car is able to provide to you everything that you wanted to know. The car knows where you're headed, your location, the time of day, it can even create a curated guidebook once it assesses your interest. Multimodality in the car with AI creates infinite possibilities for innovation and creativity for automakers. How do we make all this happen? Within the car, this mesh of inputs is processed by mapping the long-term knowledge graph of the digital life of the driver in the car, to a more immediate-term input from the cabin and its occupants, their preferences as well as the world around the vehicle. The car then orchestrates based upon this real-time processing, how the world is navigated for automated driving, as well as to provide the occupants a more intelligent curated experience based upon what is around them. We are very excited to bring what we call compound AI to the Elite platform. Let's focus our attention with an update on the automated driving stack. Once again, our strategy remains unchanged here. If you may recall, I introduced to you many aspects of our automated driving strategy back in '22. We are developing a global automated driving stack in partnership with BMW for drive policy and customer functions. In parallel, we are developing our in-house computer vision stack, which continues to be a key input into this partnership that we have with BMW, both will commercialize first in 2025 and scale thereafter. The hardware/software codesign allows us to deeply understand the requirements of the stack algorithms end-to-end. The stack allows us to focus from entry-tier mandate level applications all the way up to various levels of automated driving. This strategy, coupled with our broad-based SoC road map addresses every tier of vehicle and provides us a holistic footprint for our automated driving and driver assistance strategy. The stack development has allowed us to build significant infrastructure around perception, path planning, behavior planning, data collection, annotation, differential maps, all of this has created the foundation for a data software factory or DSF, which is now a key differentiator for our right platform. We will launch in 60-plus countries next year. So we're very excited about the scale that we are about to achieve with the Qualcomm ADAS tax strategy next year. The stack is now being tested across various urban, rural and highway operational domains. And this is a true testament to silicon software advanced AI algorithms and the tremendous amount of effort that the ecosystem has put together, both at the edge and in the cloud that come together to build this end-to-end solution. We're also very proud of our partnership with BMW as we prepare for upcoming launches in 2025. Another very exciting development that I would like to share with you. We have just recently in the last month won our first design based upon the BMW stack at a new OEM. This also happens to be a Flex capable design. Flex is a new architecture that we introduced in 2022, which allows us to concurrently run driver assistance infotainment capabilities on the same SoC. We dedicate resources to both domains and that significantly improves the overall cost of ownership of the platform. We will commercialize this design in 2026, which will take our computer vision stack, the drive policy stack along with an infotainment stack on the same SoC. And this allows us to continue to push forward on the differentiation that we are driving with central compute architectures for the car. Over the past decade, we have built our business brick by brick, deeply understanding the evolution of the automobile. And developing a scalable SoC and software architecture that allows for modularity, tiering, reuse, differentiation and coexistence of multiple diverse ecosystems and requirements. Pivoting to take on challenges of automated driving and evolving the company to develop expertise with highly complex safety, computer vision and advanced AI networks that require edge and cloud infrastructure, we have scaled globally to keep extending our lead in automotive. And all the while, continuing to keep evolving our SoC architecture to keep up these innovations. Our focus has been the enablement of the ecosystem of our partners to create long-term differentiation on an open programmable platform. And now in the age of Gen AI, we once again find ourselves leading, bringing new features and functionalities to our OEM customers and the infinite possibilities that these will bring to consumers and to the automotive ecosystem. We are very thankful to our partners and our customers who continue to look to the Snapdragon Digital Chassis as the platform they trust for differentiation, for choice, for global footprint. And we appreciate your confidence and support as we've built this outstanding business over the past many years. We look forward to the remainder of this decade as we continue to accelerate our growth. To summarize, we have delivered on the commitments that we made to you in November '22. The industry continues to build on the Snapdragon Digital chassis. Gen AI will further reinforce the digital cockpit transition, and we are ready for it. And driver assistance and automated driving stack opportunities will expand for us as we get past our FY '25 commercialization with BMW. Now before I turn over to Industrial IoT, I would like you to hear from some very important customers. [Presentation]

A round of a plus for our customers. We are so honored to be partnered with Mary Ola and Chairman Wang. Thank you. for your support. As we transition to industrial IoT, I would like to first start the session off with a video to give you a sense as to the approach that we are taking as we go after this new market. Over the past 6 months, we have embedded ourselves at Aramco. Aramco is the world's largest oil and gas company. And the focus that we had was to help build along with Aramco, the largest industrial IoT cellular network for industrial operations. We work with Aramco and their ecosystem to design to modernize everything from sensors to AI at the edge. We're focused on worker productivity. We're focused with Aramco on taking their local models, training them in the cloud and then deploying them at the edge across a wide variety of end products, which I'll give you examples of. We're focused on technologies such as camera and AI that are critical in environments such as oil and gas for real-time surveillance for detection of threats, for making sure that the employee base is safe at all times. And we've been very excited to put ourselves along with Aramco in their shoes, deeply understanding and then deploying our products in their environments. As you know, we've done this successfully in the automotive space, and we're going to take a very similar strategy as we go focus on the industrial IoT ecosystem. As you will see today, we are rebooting our industrial and embedded IoT strategy. First, AI is changing how businesses look at productivity, resilience, environmental decisions and safety in a new light. To realize this, there are 3 main catalysts. Number one, connectivity for all edge endpoints; number two, intelligence at the edge. And finally, end-to-end AI thinking across the workflow from edge to cloud. This is something we are spending a lot of time on with our partners, both system integrators and customers. We are seeing patterns of solutions repeat across verticals. The use cases range from looking for anomalies, changes in trend, the detection of an alarm at an unmanned endpoint, requiring interaction with an automated workflow and in many cases, a human. Our approach has been to solve these problems with ecosystem partners at a system level. Let me tell you how we plan to do this. Let's start with a few examples. In machine-to-machine applications, a major shift is occurring. Before Edge AI, the only way to extract intelligence from data from machines, for example, information from a sensor or images expected from a camera was to upload this information and process this in the cloud, not very scalable, not very efficient. Now with AI available at the edge, the architecture is changing. First, enterprises are training and tuning models with private data that they own. This can be time series data. This can be golden data that they have collected over a period of time, anomalies that they're aware of, positive reinforcement as they learn more about their systems. And then they are deploying these models at the edge. So sensors can now in real time, run AI models at the edge to take intelligent decisions and if needed, the necessary action. This can be implemented at the far edge, which is the device itself or the near edge, which is the Edge AI box, which acts as an AI workload aggregator. This shift applies to every vertical from energy to utilities to oil and gas to retail in any form of data from time series readings in a sensor, the camera feed for an industrial inspection robot or a security camera. And by applying Gen AI on intelligently processed AI output makes it human consumable instantaneously. Edge AI makes this available universally, improving the total cost of ownership, the productivity and compliance and reducing risk for business owners. For human-to-machine interaction, it's the same analogy. AI allows you to take all the data that an enterprise has mined, train a model, which is then deployed at the edge. The industrial handheld or an intelligent kiosk can host this local model, respond to queries in real time, accurately, privately, it's all local, and it's aided by the cloud where necessary. With the ability to host large language models that range from 7 billion to 30 billion parameters locally at the edge, we have worked with customers on barcodes, on instruction manuals, on frequently asked questions, training guides. This allows vocal productivity to improve significantly as the total cost of ownership goes down. We are working with enterprises and system integrator partners to embed pretrained models to near and far edge devices for local inferencing. These are then front-ended by autonomous agents that are part of the enterprise's data brain with access to data curation, index databases and long-term knowledge graphs. We have seen this transition across every market from autonomous equipment and agriculture and manufacturing to remotely manage construction equipment from intelligent retail kiosks to computer vision-enabled vacuum cleaners, drones, line inspection, surveillance, tracking, reporting, sensing, all of this is a massive transition being driven at scale. Workloads are moving towards edge intelligent processing with real-time analytics and actuation. Now across a very wide variety of use cases that require more than just sensing. The sensing is local as is the processing and the decision-making. And this has a very broad applicability, fixed and mobile applications, consumer and safety workloads, edge-only deployments with form factor and thermal constraints. We are seeing a transition towards a more performant class of devices. Connected intelligent endpoints are redefining the industrial edge. We have all the technologies, all the enablers as well as the multi-industry experience to put this into place and build our industrial portfolio, which I will go over. As you know, we have instantiated our technologies into a product portfolio across many industries that are now at scale, and that's what we plan to do with industrial IoT. Bear with me for the next few slides as I walk you through the approach that we're going to take. We offer chipsets for every application, connectivity and compute. We support every software stack, every operating system, safety grade, commercial grade, consumer grade. We support various enabler technologies that are abstracted via SDKs. We build AI infrastructure at the edge that allows us to deploy models for accuracy and then scale them. We are building micro services that allow us to be able to take applications, connect them to each other. We are building solutions that are referenced templates for broadly used applications. And finally, we are building a location and observability cloud that allows our customers to track and manage their assets. Let's talk about our chipsets. We have a vast portfolio, as you can imagine, we have now segmented for the IoT business, this portfolio by consumer applications, commercial applications, industrial grade applications. Those that require safety. We have a large connectivity portfolio, and we are also going off of the industrial PC space. This allows us to have full coverage across every possible application processor and connectivity segment. As far as the hardware and software stack goes, we support simple to complex out-of-the-box platforms with hardware development kits, IDEs, the necessary SDKs and multi-operating systems. With the AI Hub, industrial developers can now optimize for their specific use cases. The developer selects an application for their use case, model of their choice, optimizes it with their data set and then deploys the model. Next on macro services. Today, we see app developers able to transfer their projects to our container-based macro services framework. We have developed vision and robotics macro services that are already being used by developers to get a running start on our IoT solutions platform. And finally, for the deployment of the end solution, our foundry services supports software packaging, security and the deployment support necessary. We are creating several reference solutions from complex video surveillance to robotics, for home and factory environments and simpler Gen AI applications for on-site or in office chat functions with local language support and train on proprietary data sets. I'd like to share some detail on the part of the portfolio so that you get a sense as to how expansive the applicability is. We introduced the IQ series in October of this year. And the one key opportunity that was obvious to us was that as industrial transformation occurs across all these verticals, edge processing with AI is absolutely necessary, but it has to take place in very harsh, very demanding complex and deployment environments, industrial inspection, robotics, energy and utilities, video surveillance. They all require Edge AI, but safety compliance is critical, extreme operating conditions and temperature, the full traceability for failures. The IQ Series supports the most advanced camera and AI pipelines, and it supports both industrial and consumer operating systems. Customers have embraced the family because it is highly tiered. It can apply to any vertical, and it supports multimodal AI. The same product works as a productivity platform solving Gen AI use cases, a video surveillance platform for security as well as the data analytics for edge processing for the utilities and energy industry. With Edge AI on the IQ series, we are supporting autonomous robotics and military applications, which require multisensor fusion, localization and path planning, something that we've developed already in the automotive space, and these are getting traction from customers. Customers are looking to replace power hungry expensive CPUs and GPUs that are traditionally used in these spaces with Snapdragon SoCs. As we move to the edge, it drives intelligence to the device and lower total cost of ownership for the enterprise. It's a complicated slide. Let me give you an example. Video surveillance and safety is a fundamental use case if you focus to the middle of the slide. The approach that we are taking is that by driving intelligence to the camera, if it is a smart camera or if it is a legacy camera, connecting it to the on-prem AI edge box, we are able to process the relevant images of interest that are identified by the model at the edge itself. This gets away with hours of manual processing and scanning and the intelligence on the edge has a direct relation on the total cost of operations. With this approach, we are all local, all on-prem. All the influence is happening at the device edge or the near edge on-prem. You only need to send on the cloud what you want to store for permanent storage. Everything else is happening locally. This allows us to be able to integrate this capability into existing infrastructure as well as greenfield infrastructure, working with system integrators. There are many such use cases that we can extend this capability to. Worker productivity in enterprises, automated mobile robots in food delivery, all of these benefit from AI on the edge and lower overall cost of ownership. In collaboration with our system integrator partners, intelligence at the edge and integrating into existing infrastructure has become a kiosk and they are ready to scale these solutions across their customer. Connectivity has always been critical to our road map and with industrial. Because of the breadth of environments from fixed indoors, nomadic, outdoors, high-speed, low coverage, no coverage, there is just a massive amount of complexity. We have solutions that range from electronic shelf labels and retail stores supported with Bluetooth low energy to gas meters with low power wireless access to 5G industrial gateways and everything in between. This allows us to be relevant to every industry in every market across every type of edge device. And last but not least, the Qualcomm Aware Platform adds the necessary observability and visibility as a cloud service. This allows our customers to track their assets, record their operation and behavior, get access to location through any type of connectivity. Our location cloud for Wi-Fi has over 10 billion Wi-Fi access points and it can provide cost location to any device and serves multiple millions of location requests every day. And of course, we can support fine-grained location with GPS and network assisted capabilities. And all this requires is the integration of a simple SDK that can be embedded in Qualcomm or third-party silicon or any end device. Hopefully, that gave you a sense of the way that we are approaching our industrial IoT portfolio as we are rebooting this strategy. This obviously allows us to expand the addressable market opportunity across every market type. And we are doing this organically. The breadth and applicability of our portfolio is very vast. And now with AI at the edge, we are a part of every conversation where new disruption is ongoing. Why is that? Because we are always connected. We are intelligent at the edge, and we are working with customers and system integrators to drive topologies that will change the way that AI gets deployed. Now that you have a good sense of our product strategy, let's talk a little bit about how we will take these solutions to market. With automotive, we've built quite a bit of experience in entering new markets with this strategy that is based upon understanding customer needs deeply, where the technology inflection points are and then developing scalable, repeatable templates that we can take to market with partners. For industrial IoT, we are following the same concept, and we are curating solutions by verticals. It drives our long-term strategy on being the global ecosystem company. Here are some examples. We've spent a lot of time in the energy and oil and gas space in the past many months, and we have indexed on a few operational problem statements, site operations and safety, worker productivity and asset monitoring are examples. In oil and gas, we have identified 3 problem statements. Worker safety and situational awareness, a real-time monitoring of sensors and worker productivity. As I've shared earlier, we are working with system integrators such as Accenture, Capgemini and Deloitte as well as many partners like Honeywell and others with whom we are implementing solutions. As new AI models are trained on customer proprietary data and then deployed at the edge, whether these are in a security camera or the drill bit that has a camera head in front of it or a chatbot for a help desk, our system integrators and specialist partners are helping us to scale these solutions to be deployment ready. This strategy builds on our success in automotive and mobile, where we have brought together many such ecosystem and partners together. With this approach, our ecosystem and partners are also starting to see the value of solving these problems horizontally. Another market that we've spent a lot of time on is retail. The focus has been on worker productivity, customer experience, in-store analytics and efficiency. These problem statements cut across many technologies, products and end-use models. Deploying AI at the edge, as I showed to you earlier in retail is very powerful. Store surveillance for safety of customers, loss prevention and overall situational awareness is a pain point for every retailer post-COVID. Our surveillance and situational awareness solutions have been very impactful. We are focused on real-time tracking and visibility of assets and customers with Aware and to drive and improve the store associates productivity, we are deploying models at the edge that are trained on the retailer's data, creating tremendous efficiency gains. In-store visibility on the customer's experience is tracked via camera, Bluetooth and Wi-Fi beacons. And we are able to track the entire end-to-end journey of a customer and improve store analytics. This is a market that Qualcomm has been involved in quite deeply for many years, and we've been working with retailers, big-box stores, payment processors as we deploy this fairly complex ecosystem across many product types. The new industrial IoT strategy is to build a scalable blueprint for every vertical. Look for clusters and patterns of similarity where they might exist and develop bespoke solutions as needed. It allows us to be relevant across many markets in a short period of time, and it makes our portfolio highly relevant. And of course, we will continue with our direct chipset business with our new expanded portfolio. The go-to-market is obviously the other key pillar of our strategy. We have a very broad footprint across contract manufacturers, ODMs and OEMs, module manufacturers, distributors, system integrators. And we have direct access to a multitude of end verticals who give us access to over 15,000 customers and growing. Given the breadth and depth of our new portfolio, we get to see opportunities through many, many lenses. The industrial IoT portfolio that I shared is one of the most competitive and comprehensive in the market, and we are organizing ourselves to be highly relevant in this space. I'm confident we will see the same success that we have in other markets. In summary, a big thanks to our massive customer and partner ecosystem, which continues to grow daily and adds tremendous value to our road map. The key takeaways from our perspective for this business are that the edge endpoint is now connected and intelligent. The enterprise is continuing to embrace AIs and evolving its workflows. Gen AI will be a massive excellent in industrial digital transformation. And the Qualcomm industrial IoT portfolio delivers the most comprehensive product software and solution and is a confluence for our ecosystem and our partners. I would now like to welcome our CFO and Chief Operating Officer, my friend, Akash Palkhiwala to the stage. Before Akash comes on, let's hear what a few of our customers have to say. Thank you.

Good afternoon, everyone. It's a pleasure to be part of today's Investor's Day event. At Honeywell, we greatly value our collaboration and partnership with Qualcomm Technologies. Together, we are driving industrial digital transformation, improving worker productivity and accelerating the automation to autonomy transition across our industrial portfolio from our industrial grade handheld devices to controls platform, covering up our edge devices by leveraging Qualcomm technologies across the board. Qualcomm's portfolio of low-power, AI-enabled processors, combined with Honeywell's sensing technologies, enables a new family of industrial sensors that drive smarter, more responsive operations. We're also integrating Qualcomm's connectivity and AI technologies into our field process knowledge systems to extend connectivity to remote plants and manufacturing facilities, enabling greater data capture and analytics at the edge. Our collaboration and partnership is empowering industries across the board to interact more intelligently with their environments and advancing the industrial space towards autonomous operations. At Honeywell, we know the future is what we make it. And I couldn't be more excited about the future Qualcomm and Honeywell will shape together moving forward. Thank you.

Hello, I'm Joe Ucuzoglu, the Global CEO of Deloitte. Thank you for the opportunity to say a few words at your Investor Day event. As you gather today, enterprises around the world are realizing the promise and potential of the AI revolution. AI is helping transform countless global enterprises and there has never been a greater opportunity to drive innovation and efficiency. In a world where real-time insights are essential, enterprises are searching for AI-driven solutions that enable their businesses to make smarter decisions to streamline operations and to unlock new revenue opportunities at the edge. And that's why I am incredibly excited about Deloitte's collaboration with Qualcomm on connected intelligent edge solutions that have enabled us to integrate powerful, scalable AI directly into market solutions to help clients address their challenges and reimagine their customer experience. Together, we're pushing the boundaries of possibility by combining Qualcomm's Edge and hybrid AI products with Deloitte's GenAI capabilities and industry-leading solutions, enabling companies to achieve increased agility, resilience and competitive advantage while optimizing business processes and delivering deeper insights and innovation across every level of their operations. Thank you, Qualcomm. Deloitte is proud to be working together to help global enterprises truly achieve more from their transformation journeys.

Good afternoon, everyone. And I just want to start with a big thank you for all of you for being here and actually people online as well who are watching. Thank you for coming here. Thank you for listening to our story. We're very excited as is obvious through the presentations about where we are and where we are going as a company, and we want you to be the part of this journey. It's always great to be in New York. But as I was getting prepared for this event, I realized that there are two things that bind all of us together. First one is obvious, it's a love for Qualcomm. We all love this company. The second one is almost as important, it's our mutual dislike for the L.A. Dodgers. If you're a Padres fan or a Yankees fan, you don't like them these days or ever actually. So when we step back and think about ourselves in the future, the thing we're most excited about is the unmatched technology portfolio that we have. And it's sitting at the intersection of all the transformation we discussed across happening across all these industries. So we are very fortunate to be at that position. You've heard from Cristiano about his vision for the company. You've heard from Durga why it's a no-brainer. And I know there's a bunch of financial investors in the room, cost is the main reason. It is free. You got to run it on the device. It's a no-brainer to do Edge AI on the device. Alex and Nakul made a very compelling argument as to why we have the right to win in these new markets that we are participating in. And my job here is now to bring it together in a financial framework. And so let's just get to it. Okay. So we'll start with a quick overview, backward looking, and then I'll get into four different topics. I'll talk about revenue growth and diversification, capital allocation and operating discipline. As since Cristiano became CEO, he has been very focused on changing the company to focus on this diversification journey. And the entire management team is very aligned with that while keeping operating discipline in check. So fiscal '24, we delivered very strong results, very strong execution. Revenue grew by 9%, EPS grew by 21%, and that shows the operating leverage in the business. Given the cyclicality that we've experienced over the last 2 years, we're happy to see a relatively normalized year in fiscal '24. And we're looking forward to growing from that point on. Strong year, we're happy about what this means for us in the future. We looked at fiscal '24, probably makes sense to go back further in history and look at how our track record has been. So we picked fiscal '19 as the starting point. So okay, we can get rid of the COVID, the complexity and the ups and downs, inventory build bleed that it introduced. So over the last 5 years, how have we done. Our revenues have doubled, EPS has tripled, very strong performance. This performance validates our strategy, and it provides a strong foundation to accelerate growth and diversification from this point on. Now if you look at QCT, our chip business, for the same metrics, we've more than doubled revenue there, and we've increased our operating margins from 15% to 29%, very closely aligned with the long-term target of 30% that we've set. We also saw very strong growth in our revenue streams. 31% CAGR in automotive. Double-digit growth in handset and IoT as well. In handsets, if you abstract out the share gain at Apple and look at Android, we grew low double digits in Android as well. So very strong performance across the board, across our portfolio. Now we'll transition to looking at our current business priorities going forward. As you know, we have 2 reporting segments. We have the QTL business, which is our licensing business. We have the QCT business, which is our chip business. And our priorities within the chip business are Android handsets, automotive and IoT. As Cristiano outlined, we have a very large TAM that's available for our business, approximately $900 billion. And we have all the right technologies to go and take advantage of it and exploit it. In addition, as Durga outlined, we expect Edge AI to be an accelerator in all these markets, both in terms of the size of the market and in terms of our competitive differentiation within the market. So we're very excited about it. So next couple of slides, I'll quickly address handsets and then move on to our diversification priorities. So starting with Android handsets in QCT. Android has unmatched scale in the handset industry. 82% of the phones worldwide are Android, which is approximately 1 billion units a year. It's by far the largest ecosystem in handsets. Within Android, Qualcomm has a very strong presence. If you compare us to our closest competitor, we are 2x the revenue overall, and we are more than 5x revenue in the premium tier. Snapdragon 8 chip is the performance benchmark chip in premium tier in handsets. So we're very happy about that. We'll continue to maintain that position as we go forward. There are two key drivers of revenue growth within handsets for us. First is content increase, and second is stronger product mix. If you look at the chart to the left, it shows our premium tier chip over the last 3 years. And if you do an apples-to-apples comparison, we've seen a content increase of greater than 10% CAGR over this time period. So very strong demand for more technology as consumers continue to view the handset as the primary device in their lives. What are the drivers of this content increase? More processing power, both from the CPU and GPU perspective; new camera technology; and of course, AI; and a bunch of other things. So we have a very strong road map. Looking back, we have a very strong road map looking forward. We are confident that the content will continue to increase in this tier. We've also seen the same trend in other tiers in the handset market. If you look at the same period, the last 3 years, what has happened to our Android revenues in handsets, we've grown at a CAGR of 8%. As we look forward, we expect the revenue TAM for the Android handset market, this is for the overall market, to be mid-single digits going forward. So we expect to continue to see growth in the market as more and more content gets added to the devices and the mix gets richer. Okay. So I'm transitioning to automotive now. It really is a great example of our execution and being able to take advantage of a market that is going through transformation. When we came here in the 2021 Investor Day, we were a newcomer. Most people in the industry did not believe that Qualcomm could scale so quickly in automotive. But since then, Snapdragon Digital Chassis has become the platform of choice in automotive. We expect the TAM to go very significantly over the next few years from $50 billion to $100 billion. As software-defined vehicles take over and more and more silicon content is required to manage the transitions that are happening in the automotive industry, we're going to see this TAM expand significantly, and Qualcomm has all the right technologies to be able to take advantage of it as Nakul showed. We expect this growth to be a tailwind for us going forward. On the design-win pipeline, we updated about 5 months ago, and we said the total design-win pipeline was $45 billion. When we did this Investor Day 3 years ago, it was $13 billion. 2 years ago, it was $30 billion. Now it's $45 billion. And if you look at the diversification within the pipeline, it's tremendous, not just from a product perspective because we have connectivity, cockpit and ADAS, but also from a geography perspective, we have traction -- broad traction across OEMs in Europe, in Americas and Asia. And that's one of the strengths of Qualcomm that are design-win pipeline cuts across a lot of customers. It cuts across a lot of peers, it cuts across a lot of products, and it puts us in a unique position in the industry. We expect this pipeline to continue to grow. Over the next several months, we're going to see major OEMs make decisions on their platforms, and we expect this pipeline to keep growing. So very strong position in automotive as far as design wins are concerned. Revenue growth is really the ultimate proof of your competitive differentiation. And we've been very proud as to how our revenue has grown over the last 12 months. We've continued to gain share and increase content in our devices. As a reminder, the revenue targets we've previously provided is -- greater than $4 billion in fiscal '26, greater than $9 billion in fiscal '31. We're adding a new target of $8 billion in fiscal '29 and we're going to give targets for various businesses in fiscal '29, so we're aligning to that. And the current revenue trajectory puts us in a very comfortable position against these targets. And our pipeline gives us confidence that we'll be able to achieve it. If you take the revenue over the next 5 years that is implied here, our design-win pipeline cumulative revenue over the 5-year period, our design-win pipeline covers 80% of that already. So this is something that we have significant insight into, and that gives us confidence in putting out these targets. So now I'm turning to IoT. And as we've said in the past, for us, IoT includes three things: that's consumer; networking; and industrial. In consumer, we have PC and XR that Alex presented about, but we also have other personal computing devices, wearables, hearables, tablets. In networking, we have Wi-Fi access points and 5G fixed wireless access devices as well. If you look at IoT today, it's already a scale business for us with $5.4 billion in revenue. And going forward, as digital transformation happens, it is going to drive demand for the technologies that we have, more connectivity, more processing, more AI, and so we're very excited about what we can do in this area. Now what I'll do is I'll cover in detail the 3 presentations that were previously provided on PC, XR and industrial. But before I do that, I want to just quickly remind, go back to something that Cristiano had said as to our framework for success in these areas. The way we operate is we look for a market with very large TAM that is going through an inflection point. We bring our technology to it, we build the channel and we scale revenue. We did this very successfully with automotive, with autonomy coming in, with digitization coming in. We did this very successfully with RF front end before that with 5G coming in. All these 3 markets are going through a very similar transition. There's an inflection point happening, and we're going to take advantage of that, and we've become a significant player in these industries because of that. So I'll start with the PC. But before I give financial forecast on it, I want to reiterate three key points that Alex made. First is the strength of our road map. We have the best part in the industry, and we will continue to have that. And we have a series of chips that allows us to address all different price points. The second is our partnership with Microsoft on AI and with Copilot plus PCs. We were there first, and we'll continue to be the partner because we have the right technology to enable the use cases that they are envisioning. And then finally, our design wins across OEMs. We expect that to cover 70% of the total volume of notebooks as we get into 2026. So very broad set of design wins, which positions us to then go and gain share. With the things I just outlined, we are providing a target of $4 billion in fiscal '29 for PC revenue. This target contemplates a TAM of over 200 million units and $35 billion in silicon revenue. The two key metrics for us that matter is what percent of these devices become AI PCs? And then what percent of these devices move away from the x86 ecosystem. In 5 years, we expect 90% of the devices to be AI PCs. That's where the world is going, and we have the performance advantage. And we expect 30% to 50% of the notebooks to move to non-x86 platforms. That puts us in a great place as well. So that is the basis for our forecast. And this would position us to be one of the top silicon players in the PC industry. Now I'll transition to XR. Similar to PC, I'm going to reiterate a couple of points. First, we see XR as the next computing platform. Second, GenAI is coming in and it's driving this inflection point on the device. And it's going to enable new use cases, and it's going to require the technologies that Qualcomm has. Third is we are the platform of choice in the industry. Every major ecosystem, Meta, Google and Samsung, Chinese OEMs and others, everyone is using us. We have very broad traction. So we are providing a target of greater than $2 billion in revenue in fiscal '29 in XR. And the way we thought about this is we looked at other ecosystem of devices with similar use cases and their scale in the market. So for AR, we looked at what is -- what are portable devices and what's the scale of those portable devices in the market, like smartwatch, like headphones and handheld gaming. And then for VR, we looked at stationary devices, like console gaming, PC and others. So this forecast is based on 15 million units a year for AR and VR and MR combined. Clearly, there is opportunity beyond this. If you look at what use cases can be enabled in AR, especially if you look at the view of the market that a lot of the other leading ecosystem players have, you could see significant upside versus the number we are showing here. Next, going to Industrial. I'm going to reiterate a couple of points here. First is beyond chipsets, which everyone knows we're very good at, we are building software and solutions for this market. And we're doing it for various verticals one at a time so that we're building something very specific for the end customer. Second, we are building the channel to the market. It's very important that we go into a new industry, we understand the channel, we build the channel and we operate in a way that our customer wants us to so that we can get to all of them. We're investing in both these areas, and this is what positions us to be successful. So digital transformation and AI are driving an inflection point in industrial. This happened in auto before. A significant portion of the market was MCUs. It transitioned over to processing, it transitioned over to AI and connectivity, and we were the beneficiary of it. Think of this as a parallel to what happened in automotive. There is a significant portion of the market today is MCUs and wired connectivity. The industrial market is moving independent of any of these -- any of our companies, it's going to move over to connected devices, wireless and wired, to devices that do high significant processing and have AI capability. And that moves the market to us. So we are setting a target here of $4 billion in revenue in fiscal '29 in 5 years. So now let me put all of IoT together. In addition to PC, XR and Industrial, of course, we have a very large networking business, which is already operating at scale, and we have a growing personal consumer device business as well. So you put all of these together, we expect revenue to grow from $5.4 billion today to $14 billion in 5 years. This is a 2.5x growth in our revenues. Now let's do that again with auto and IoT combined. We're talking about revenue growth from $8.3 billion to $22 billion. Again, more than 2.5x growth at a CAGR of 22%. This growth in annual revenue far exceeds the scale of the Apple chipset business revenues today. And it's highly diversified. It cuts across various end markets, various customers, various products. Okay. So now I'm going to transition to the third part of my presentation, which is really talking about operating discipline and capital allocation. This slide shows our approach towards operating expenses and how we're managing it while focusing on diversification. Over the last 5 years, we've grown revenue at a CAGR of 15%. We've grown R&D against it at a CAGR of 8% and SG&A just 2%. So we've been very careful in how we scale spending along with the revenue growth. If you look at the last 3 years, we've kept the total spend approximately flat. And within that spend, we've redirected investments from our mature businesses to the diversification priority areas. With this, we're also maintaining our target of 21% to 23% OpEx as a percent of revenue going forward. As you know, we are investing in several different areas ahead of revenue ramp. And so we are very proud to be able to manage OpEx as we go into these new opportunities and deliver on the revenue forecast we just outlined. Acquisition approach. The key message is nothing has changed, right? We've always thought of acquisitions in two buckets. First is, can we buy an asset that allows us to accelerate our diversification plan? Second is, do we buy an asset that gives us new talent, new technology that is required for our technology road map? As an example, if you think about AI, we've bought several smaller AI companies where we brought in the teams, brought in the technology. And what you see in the AI presentation earlier in the day is the culmination of those things, plus, of course, our internal development. Overall, the M&A strategy that we have has been consistent. Similarly, our framework remains consistent for capital structure and capital returns as well. We are committed to a strong balance sheet, strong investment-grade rating, consistent with where we are at today. We will return most of our free cash flow to shareholders through dividends and share repurchases. On dividends, we now plan to grow dividends low to mid-single digits going forward and will scale buybacks to align with our total capital return program. Before I complete my presentation, I want to quantify the long-term target for our diversification plan. We are targeting a mix of 50-50 by the end of the decade for handset and non-handsets. And we believe this transformation will be highly value accretive, and we look forward to giving you updates on it going forward. With that, some key takeaways, we're executing on our diversification strategy while staying committed to operating discipline. We are positioned to benefit from the proliferation of on-device AI across all Edge devices. We are committed to strong cash flow generation and capital returns. And then finally, with all these objectives, we're looking to transform the company, have much lower customer concentration risk and higher secular growth opportunity. That concludes my presentation. I would like to request the presenters and Alex Rogers to come back on stage for Q&A. Thank you.

Stacy Rasgon at Bernstein Research. I had a question on your PC targets. So you're targeting $4 billion by '29. It looked to be about 11% of the revenue TAM, if the revenue TAM was $35 billion. But if I calculate the units with the 30% to 50% range, it to be something like 60 million to 80 million units, if I sort of work through all the different percentages. At the midpoint, it would be like 60 million to 100 million. So the midpoint would be 80 million units, $4 billion would be like $50 each. It's something like 30% to 45% of the unit TAM and only 11% of the revenue TAM. Like what are you guys assuming for actual like units in '29 and ASPs? Are my numbers right? Or am I making some mistake in the math?

Stacy, the way to think about it is we're new in this market. We have a great product -- set of products. We wanted to set a target that we're very confident about. And then as we go forward, we're going to look at that target and recalibrate if needed.

I guess what I'm asking is the revenue target doesn't seem consistent with the unit numbers that you laid out. Like where is my math -- where am I getting it wrong? Or are you just being conservative on the revenue number relative to the units...

I think we're going to set a target that we are very confident with on revenue, and then we'll look to work through it over time.

Okay. And maybe one more, if I could, briefly. On the Android SAM growth, you said like mid-single digits. You don't think the content increase you can see from AI if the edge is higher than that? Or like any thoughts on share?

Yes. So I think clearly, if you contemplate AI, there could be multiple vectors of change within the handset market, right? You could see a lot more content increase. More importantly, I think competitive differentiation for us is significant. And you could also see the market expand -- the total units expand significantly as well. But we have not seen that yet. And so we thought -- think of that as upside rather than the base planning assumption.

So for your handset business on the Android, you're assuming sort of like mid-single digit and then the Apple is going to do whatever it's going to do.

Yes, that's the base market forecast. And we play in the premium tier. So we have -- usually have an advantage over that.

Chris Rolland, Susquehanna. I appreciate the long-term targets, but the next few years could be a little bit challenging with the North American handset customer transitioning out. I was wondering if you guys could provide us perhaps for the next couple of years an outlook, how to think about revenue? If not something specific, then a little more broad during this transition period?

Yes. So the way we think about it is the growth that we are forecasting in all these areas, very significant. And of course, it scales over the 5-year period. When you combine it with other factors in your forecast, we still feel like bearing any cyclicality in the overall market, we think that we have a position to continue to grow revenue every year.

Perhaps as a follow-up, you guys have had pretty incredible content growth, not just with Snapdragon, but the surrounding platform and chips. There have been some articles that suggest that you might be hitting the limits of Qualcomm as a percentage of the handset cost or total BOM. Do you guys have some pushback there? And do you think that content growth is still going to be a very significant driver for you guys from here in these platforms?

Yes, I'm happy to take that. Look, if you look at what happened on the handset on the premium tier, and I remember the conversations when everybody will basically say it's impossible for premier tier handset to get more expensive than $600. And you look where we are right now. I think the device are becoming more and more useful, and we have seen demand for more processing power in those devices. We have not seen a slowdown when we think about the road map in the next few years from a camera perspective, graphics perspective, CPU and AI. So I think there's always going to be a ceiling when you think about how much is going to be the total device ASP. But we're very confident when we look at the road map for Snapdragon 8 of a runway for content increase, especially on the processing side.

Brett Simpson, Arete. I had a question on the gross margins for auto and IoT. I mean, I guess, I'd love to get your perspective on what sort of returns you think you can make. If I look at some of the guys in ADAS, their gross margins are well in the 60s. How do you feel about the leverage in the gross margin line as you diversify the business?

Yes. So there's kind of a different answer for the different types of markets we are in, right? You think about several parts of automotive where we have very strong gross margins higher than the company average. There are areas within industrial IoT, where it's certainly a different gross margin structures, very strong as well. I'd say PC, we are starting in the market, and so that's a place where we do have slightly lower gross margins than Qualcomm average. So it's really a mixed bag between those areas. And I think the best way to think about Qualcomm gross margin is at the consolidated level. And we've said consistently that where we are at right now is a reasonable way of projecting the business going forward.

And maybe just as a follow-up. And this is linking back to your case with ARM and Android. I guess there's been talk about them potentially canceling licenses. And I'm just sort of thinking through that path. RISC-V is that an option will you consider? And where is the status of RISC-V porting to Android?

You take the first part, I'll take the second part.

Yes. I think we have broad rights under the license that we have with ARM for custom design cores, and we're very comfortable that those will be affirmed. Trial starts next month.

Maybe a comment on RISC-V. Look, there has been quite a bit of development in RISC-V. I think also there was a desire to actually drive RISC-V towards high-performance and commercialization. As a matter of fact, I think we've been very pleased that Qualcomm been elected to Chair, I think the standard body right now. We have been very active in development. There's a number of other companies and ecosystems. I think we have a joint venture called Quintauris, which is a really focus of that with European semiconductor company. It will take time, but I think it's encouraging to see development in RISC-V, especially as not a lot of love, I think, to semiconductor companies from the other ecosystem, I think that's an accelerating, I think, the R&D on RISC-V across the board.

And just on that point, Cristiano, is the Oryon core, is it capable of moving to different ISA? I mean, could you, in theory, use the NUVIA platform based on a RISC-V architecture?

Look, we are incredibly proud of our custom CPU. And I don't know if some of you pay attention when Alex Katouzian was presented, there was accelerator of Lunar Lake accelerator with TSMC 3-nanometer and was still outperforming at 4-nanometer. And I think I'll argue, we have an incredibly flexible micro architecture designed by Qualcomm. As an example, we could immediately put it on PCs, have a second generation in a short period of time on phones and have a safety-grade version, which is completely different for automotive. So I think that speaks to the flexibility and adaptability of our design.

Ross Seymore from Deutsche Bank. The IoT side, 3 years ago, you had a growth forecast that was somewhere in the upper teens as far as the CAGR that you had. And I think the CAGR over that 3-year period was a little closer to flat. And we all know there's crazy things that have happened over those 3 years cyclically. But what's your level of confidence in the future growth rate today versus what it was 3 years ago to give us the confidence that these numbers aren't more aspirational and then actually are more realistic?

It's a great question. Maybe I'll start, Akash, then you take it. Look, if you remember what happened in the past, I think PCs, I think, for us, still going to the process of maturity. There was a lot of things that need to be done on the ecosystem, especially to support the ambulator from Microsoft from 32-bit and 64-bit was only at Windows 11, the 64-bit emulation became available. And also we had the inflection point of AI that has not happened at the time. Same thing, we feel about XR. I think the inflection point was missing. We have that with GenAI on the device. On industrial, one thing that we have done because we're pursuing that organically, and we took a lot of time to develop the complete platform. Hopefully, you saw from Nakul's presentation that we understood what we needed to do, and we got busy doing it. And that's why I think we're focused on those three things. When I look at the landscape right now is, PC, we already developed, I think, the leading platform. We have the design traction. 58 is not a small number. I agree with Akash view starting from 0, almost like when you saw what happened with our auto revenues growing over time. So we put targets that we know, we feel very confident we're going to meet. Same thing for industrial. And it's not an unrealistic target when you think about what's happening for smart glasses and the demand that we started to see on AR to think about the target as well. So it's a whole different outlook. And I'll say in industrial, we mature a lot and we needed to take the time that was necessary to build the platform.

We have time for one last question.

Lou Miscioscia, Daiwa Capital Market. One of the charts you actually put up in the auto area of the pipeline, it seems that a lot of the share was digital cockpit and connectivity, which is great. A lot of vehicles are going that way. But on the ADAS and AD area, it does seem like that a couple of your competitors are having problems and maybe they're not as focused as they used to be. What can you talk about share gain in that area? And then I have a quick follow-up.

Maybe I'll start and I'll ask Nakul. One thing we said back in July, about 1/3 of the $45 billion is ADAS and autonomy. Also, we have an important milestone coming up, which is 2025 will be the joint -- of our joint development stack with BMW, and we started to see interest from other companies as well as they be able to see that stack on the road. And as far as the pipeline, as you know, we're not updating it today. But we are participating into a number of RFPs, and we're happy about our position. So I think that something as we continue to invest in the road map, you will see. I don't know if there's anything you'd like to add, Nakul.

I think one thing that is pretty unique about our portfolio is this common compute fabric that we have built that allows the customer to be able to develop for a cockpit application or an autonomous driving application that significantly reduces their overall cost of ownership. And that has -- as we have become much more relevant in the ADAS space, a lot of customers are designing with their own stacks, we will have our own stack, that underlying foundational platform becomes the chassis on top of which the industry deploys. So we have already started to win a lot, as Cristiano said, 1/3 of the pipeline is ADAS SoCs. We expect that to continue over the rest of the decade.

Okay. Great. A quick follow-up. Ignoring what we don't know yet about the outcome from the election in the sense of tariffs or other things, what can you say about China? A lot of companies that have gotten tripped up there. Very often, the governments there try to push homegrown solutions. How do you feel about your competitive situation? And is there any revenue possibly at risk if the governments there and the tariffs or the trade war gets worse?

So maybe I'll start, and I'll ask Alex to complement the answer. It's interesting. I think China is one of the most competitive markets in the world right now. And it's -- what we actually see, it's interesting, for example, in automotive, when there's a discussion about domestic developed solution, what we actually see is a Chinese EV companies want to be on Snapdragon because they want to win into domestic market as well. So what I think -- what I can basically tell you is, in parallel, as geopolitics started to become front and center in the U.S.-China conversation, the Qualcomm partnership with China has actually increased as we expand to other industry. I think that's true in automotive. That's -- that is true in PC, and it's going to be true in industrial as well. And maybe Alex can talk about our experience with the new administration.

Yes. We've had a great relationship with the Trump administration in the past. We expect a good relationship going forward. We're very positive on the recent pick for Commerce Secretary. So we were expecting to have a good relationship and to be engaged as we have been through this past administration.

I think that's it. All right. Thanks, everybody. Thank you. Appreciate it. Thank you.