GLOBALFOUNDRIES Inc. (GFS) Earnings Call Transcript & Summary

Good morning, everyone. Thank you for coming to GlobalFoundries First Capital Markets Day event here in New York City. We're extremely happy to have you here. Thank you for showing up in such good numbers. I'm your host, Sukhi Nagesh, for today. As you can see from the agenda, we have a very exciting lineup today. We have broken the agenda into 2 portions, equally timed, if you will. It will start up with our CEO, Dr. Thomas Caulfield, giving you the vision of GF, followed by GF executives who are going to be talking to you about our end markets. We'll be talking about our technology and platforms. We'll be talking to you about our commercial and go-to-market strategies. We'll also have our Head of Manufacturing talk about our footprint and the manufacturing. We'll touch upon ESG, which is a very important topic in the investment world these days. Talk about our talent. And finally, our CFO, David Reeder, will give you the financial plan for us as a newly public company. right? So for -- we have a pretty exciting lineup. For those of you who joining in person, we request that you turn off your phones. And for those of you who are online, you're welcome to send in your questions at any given point of time for the Q&A session. But before we begin, just want to just make sure you guys -- remind you that you may hear some forward-looking statements today or details on that. Please refer to our SEC filing, most recent SEC filing, which is on our website. With that, let me turn over to the -- queue roll. [Presentation]

I'm not supposed to get emotional when I watch that, but I'm pretty proud of this stuff. So what are we doing today? Look, I know the most precious commodity, always have, is our time and the fact that you made the time to learn about GF today is greatly appreciated. So with that, let's get going. I don't usually spend a lot of time on the intro chart. But I think this sets up not only the company and the day, but really where the industry is happening, partnering to change the energy, changing the world. I don't think there's any argument that the semiconductor industry is changing the world. It's how we educate our children, run our businesses. Think about some of us that have been around a little bit longer than others, right? Color TV was a big deal. Look what we have today. And it's because of semiconductors. Now why do we need to change the industry that's changing the world? Well, for one, if you don't change, you're going to be behind. But more importantly, you change for 1 or 2 reasons. You change because you don't like the outcomes or you change because there's structural or fundamental changes going on that the industry needs to adjust. And for 2 -- for us, there's 2. We need a new innovation engine that's beyond just scaling transistor size. And the second is this industry needs to double in the next decade. It's not just its opportunity. It's its responsibility to continue to drive the digital transformation of society. And it's not going to be able to deploy that type of capital in a traditional model. And so for us, you'll hear a lot today about us partnering, and partnering to do what the industry needs, to change the industry so we can all enable the growth in a profitable way for our customers and our partners. So let me start a little bit about the industry, and then I'll get into GF. We all know in all of our favorite megatrends, smart connected devices. Every minute, 7,000 new devices are added to the Internet. In fact, just as I started to make this little factoid 500 were just added. The explosion of use cases on AR/VR. During the pandemic, we couldn't get equipment suppliers to travel, there were lockdowns. Our employees couldn't travel. We found a way to take advantage of augmented reality where equipment engineers in Singapore could walk through our engineering team in Fab 8 upstate New York and how to do equipment maintenance. AI, machine learning, explosion of opportunities. Why? Because it drives efficiency. 30% of all code written today is with AI and machine learning. And that's just the tip of the iceberg of how this industry is going to leverage it. It's all about productivity. And for me, it all comes together in the megatrend of data everywhere. It's an explosion of data, 150 gigabytes are consumed by every person on the planet every day on average, and at least 1 gigabyte per person is stored. And that's going to double by 2025. And so it really creates what I call this virtuous cycle of creating data, parsing data, transporting data, processing the data, storing it, and then the ultimate is to monetize data. And this is going to continue to drive the semiconductors, as we do all these other mega trends. But the one thing about these megatrends, they're either going to be accelerated or limited by 1 factor, the rate of innovation and pace of capacity of semiconductors. So a little bit about our industry. And the way I like to talk about it, this is an industry that grew up through a series of what we call golden ages that created new applications that drove demand. Starting with PCs, and then PCs needed to be connected to one another in networking. And then something really fundamental change, the smart mobile device, the singularity of the device that allowed us to connect and be connected to everything in the world. We finally had a reason to have our thermostat connected to the Internet because we can control us from anywhere in the world, we can open our garage doors. And what happened is we went from that time from the beginning of the smartphone to where we are today, from a compute-centric industry to the pervasive deployment of semiconductors everywhere. And what you'll see is over the previous decade, 2010 to '20, we grew at 4%, and we're going to grow 3x faster than that going forward. So there's nothing like a crisis or a shortage to create awareness. Here's an industry that has changed the way humankind has lived, and no one even knew what a semiconductor was until you couldn't buy enough washing machines or cars. Now everybody knows what semiconductor is. And fundamentally, what everybody learned is what we all knew, semiconductors are at the heart of the world economy, and foundries are the bedrock of that industry. An $89 trillion world economy that sits on a $2.2 trillion electronics industry with thousands of customers, or thousands of suppliers that exists only because there's a $600 billion semiconductor industry. And you see there's hundreds of companies, you think about what the kind of companies there, sure, you have fabless. You have hyperscalers, OEMs and what we call IDMs. And the thing about IDMs, there's no such thing as a pure-play IDM anymore. They're all hybrid. They manufacture some of their own material, they come to foundries like GF. There isn't a single IBM that doesn't use GF as well. And then it amazes me every time I say this, the entire world economy comes down to 5 companies in foundry space, serving a $96 billion market opportunity. Five, 2 are in Taiwan, 70% of the output, 1 in China, Samsung and GlobalFoundries. Now I'm generous when I say scale, right? There's more foundry scale. I give -- you just have to have $2 billion of revenue in foundry, and I say you have scale. Five. And the thing about it is the semiconductor industry will outpace the GDP. You say, how is that possible? How can you go faster than GDP? Something called market share that semiconductors will have a disproportionate role in the world economy. And because of new entrants in the market, because IDMs will outsource more of their capability. The foundry TAM will grow disproportionate. And so semiconductors will outpace GDP, and foundry will outpace semi in growth over the next 10 years. So last chart about the industry, and it was where we began this presentation on structural changes that drive the need to change. What's happening in our industry? Well, when we were a Moore's Law industry, that's what we chased, and we're highly cyclical. We were cyclical around versions of Microsoft Windows and x86 architecture, looks -- and clicked ahead. Today, with the pervasive deployment of semiconductors, much more durable demand drivers, much more market segments we serve. For me, what does it mean? We will have cycles in this industry because we're not immune to macro event, but the cycles, in our opinion, will be more muted and less in frequency. In the past, it was a little bit of build capacity, they will come. So supply was always put ahead of demand. That created really bad economics if you're a manufacturer. We all know of it. Today, independent of choppiness we'll see in the market over the next decade, for this industry to double, capacity will be trailing demand, the opposite. The demand will be ahead of capacity as we get put on. In the past, manufacturers took all the risk, build it and try to get a return on it. Today, there's no way capacity could be put on without a better, more balanced economic model. GF will never put capacity on for GF, we're not a product company. We have no use for it. We just stared at it and stood idle. We only put it on if our customers want us to put it on, need us to put it on, commit to use it and help us create that capacity with their balance sheet. And then lastly, there's a reason there's 5 foundries in the world. It's tough business. Value in an ecosystem, not only is value capture or creation is equal to value capture, and it was because there was lack of value capture in manufacturing, less and less companies did that. And we can see going forward a much more healthy and balanced ecosystem of how value is captured and how it's created. And so for that to happen, we all need a new collaborative model. Maybe you heard Matt Murphy in our sizzle reel talk about what got us here is not going to get us there. And we're going to lead and have led on how this new collaborative model could work for win-win outcomes. So let's talk about GF now. That's what's happening in the industry. And maybe just to recap a little bit of our journey and our strategic repositioning. We started as a spin out of AMD in 2009, $1 billion in revenue and 1 customer. The strategy was the right vision. The world needed a global manufacturing foundry capability. Arguably, that strategy was a little ahead of its time. But it was flawed when we found ourselves in 2017, 2018, the world had changed from a compute-centric industry to a pervasive deployment of semiconductor, and we were still chasing Moore's Law. We were playing a losing game. We had heavy investments, second source business, high customer concentration. And as you can imagine, you add all that together, risky -- I'm being kind when I say this, risky and suboptimal financial results. So we decided to pivot the company to reposition, to take advantage of the structural shift that happened in the industry when we went from a compute-centric industry to the pervasive deployment of semiconductors. And we drove a transformation. The first thing we had to do is we were thinking about taking our business a different way was we had to bring in the right technical and executive team. We needed to bring people in who understood the end markets where we could create differentiation for our customers and capture that differentiation. We need to take that insight and refocus our capital deployment and R&D investments to deliver that differentiation. To do that in the best way, you needed to reestablish better connectivity in partnership with our customers, so that they can help dictate our technology road map so we can specialize for them. And as a result, we created a lot of single-source differentiated business that helped us drive margin expansion and earnings growth. Last August, just ahead of the IPO, I had my team together in Texas. And we didn't think we had to have a new play. We were in the middle of our transformation. All of a sudden, we were going to go to the capital markets. And the question was, what are we going to do next? And we gave a lot of thought about that. And it turns out, what's best for us is the realization of this business, the realization of GF, continue to gain market share in the end markets that we want to play, continue to drive the innovation in what we call purpose-built platforms with a mission in mind, reinvent and redefine capital efficiency and expansion through partnerships. And then as a result of all of that, our goal is to deliver the best-in-class financials as a foundry in the industry as we become the world's leading manufacturing of specialty feature-rich technology. So how does GF win? How do we go do all that? Well, we do it the following way. We focus on platform solutions addressing attractive secular end markets. We drive this portfolio through purpose-built innovation and platforms. We forge deep customer partnerships not only on technology innovation, but on how we build capacity together. We leverage our global manufacturing footprint. Where others will spend the better part of the next decade building a global footprint, we already have one, and we'll be able to leverage that in our -- and of course, our customers and our investors want to know we're going to be around to serve their needs, and so we're going to do it with a resilient business model that includes strong visibility to our financials. So with that, I'm going to go through each one of these in high-level detail. The rest of the agenda is -- our executive team is going to come up and take each one of these pillars and drive it much deeper. So let's start with the markets we choose to play in. Smart mobile devices, home and industrial IoT, automotive, comms infrastructure and data center. Smart mobile devices, if you think of the amount of silicon by area, in an iPhone X to iPhone 13, it's grown 60%, right in our sweet spot. If you look at the amount of silicon just in the front-end module in that same iPhone X to iPhone 12, it's another 30% within that. This is how you could hear what we just talked about, just about market share. Handsets down first half of this year versus first half of last year, yet, GF grew 14% revenue. Because we're winning market share intercepting trends, the trend of 5G transition, and winning more applications in the smart mobile devices. Home and industrial IoT. Today, for every human in the U.S., there's 5 connected devices, right? And that's going to double in the next -- by 2025, and 5x growth by 2030. Automotive, we're going from internal combustion engine, ICE, to what we call autonomous connected electrification, ACE. Level 2, EV, 3x more semi than an internal combustion car. If it's Level 4 autonomous driving, it's 6x more content of semiconductor. And then comms and infrastructure, 8% CAGR for us. So when you -- I'm sorry, comms and infrastructure, only 40% of the world's population today is connected to the Internet. So for us, when we think about our SAM, there's overall foundry TAM, our SAM is 12-nanometer and below. It grows at an 8% CAGR, and it becomes $111 billion. So Ed Kaste will go in a lot more detail about these end markets. The takeaway is this industry will double over the next decade. The SAM we play in is over $100 billion, and we will be very focused in this big blue ocean of where we play to do the following: create value for our customers and capture value for ourselves. So how do we create value? We create value through our innovation. When we think about our innovation, we do it through market-driven, purpose-built platforms, 7 of them, but we partition them the following way. We think about first our CMOS and FinFET technology. We make them feature-rich. We had embedded memory for security. We're -- had BiCMOS for precision battery management. we had high voltage for display driver. We had analog digital capability to take the analog world into our digital world. The next set is probably the best example we have about purpose-built platforms. It's all about power, connectivity with the right amount of digital performance. FDX, 22FDX, our platform, highly differentiated. Only GF has it. It's the most built-out platform we have in terms of IP and ecosystem partners. The applications range in 5G millimeter wave for handsets, to RADAR and automotive, to Wi-Fi in your household. The next one over, RF SOI, another amazing platform for front-end modules. We went from 7S technology to 8SW, 8SW+, 9SW. Every year, we innovate on that platform. As the communication standards continue to become more stringent and tight, we stay ahead of that game working with our customers. Silicon photonics, addressing the biggest issue in data center's power. The only platform in the industry that combines RF SOI, CMOS and photonics in a single monolithic chip. We take the discipline of microelectronics and bring it to photonics, the discipline of microelectronics to photonics. Silicon germanium. This is for really high-speed, high-power, high-frequency data communications. And then lastly, what's next beyond this? What's beyond silicon? It's wide bandgap compound semiconductors. And we have a program in GaN that will be the next generation of power amplifiers and helping drive the electrification of automobiles. I think the point I'd leave you with is our industry has painted itself in the corner where we think innovation is only driven by the size of our transistors. The fact of the matter is, there's many elements of innovation, and we are innovating beyond transistor sizes because someone's got to do it. The thing about being a foundry is everybody says, I manufacture, their form of foundry. Ain't so. You cannot do this alone. Your customers have no way to design into your foundry manufacturing service if you do not enable them. You enable them on feature-rich platforms with PDKs, standard cells, libraries, complex IP, foundational IP, and you do this for a broad range of applications. And you just can't do it alone. You need partners in the industry. You need ecosystem partners. We spent the better part of the decade creating this ecosystem. We have 100 partners. We have 4,500 IP titles supporting, another 950 in flight that will be deployed. Now the thing about what investors like to talk about is the moats around your business, right? What are the barriers? Well, semiconductor manufacturing, and it says is a moat, the complexity. Who's going to wake up tomorrow morning and say, I want to get into this business. Take a decade to get there. Well, you wake up tomorrow morning and you just happen to be already a manufacturing and you want to become a foundry, this is your next barrier. You got to get over this. And this doesn't happen overnight. It happens over the course of a decade in the types of investments and partnerships you need to establish. So in the end, we focus our innovation to go capture 75% of that broader foundry TAM. We talked about it's a little over -- growing to a little over $100 billion. We innovate beyond the transistor size, but focused on 3 dimensions: lowest power, best conductivity and the right amount of digital performance. And we do it -- we don't do it alone, we do it with ecosystem partners. This brings me to our customers, broad and deep customer base. 2009, 1 customer, AMD. Today, we have over 200 customers. And they're the who's who of customers. They range from fabless to IDMs to hyperscalers to OEMs. And we do work with them in a partnership way, not only do we align our technology road maps to their needs, we create capacity with their needs. We do it with 3 things in mind for both of us: Certainty, durability and profitability. The certainty is we will only add capacity if we know our customers are going to use it and they're obligated to use it. We have more opportunity than the ability to satisfy that opportunity. So we do it in durable markets. Markets that we see have durable trends where our differentiation does matter and we can capture that differentiation, and with customers who win in that space. And then we do it profitably. If we don't get the right economics, we can't add capacity. And the right economics for us are accretive to our business model and essentially very simple. If we're expanding capacity in an existing facility, it pays back in 3 years or less. If we have to build a new facility, new brick-and-mortar, it has to pay back in 5.5 [ hours ] or less. So we want to have returns on our invested capital in the high teens. And so to that end, we have over 35 long-term agreements covering $27 billion of revenue. That's a 30% increase since our IPO. In fact, just yesterday, we announced an extension of a very important LTA with Qualcomm. And our customers are participating and committing to that capacity by putting their balance sheet, $3.6 billion, as part of the ways to create that capacity. The last thing I want to make is a very important point. We meet our customers where they are. What does that mean? That diverse group of customers need different level of engagement. Some of them are very mature, and they need just good PDKs that they could count on, good IP. Others need help actually doing design. They would like us to have reference designs that then they can tailor to their needs. Others, small in nature, want a turnkey solution. They would like us to not only provide the manufacturing service to the wafers, but give them a turnkey solution to the module level. The key is we do all of this. We meet our customers to their needs. So broad customer base, partnership and innovation, partnership and how we add capacity. Balance sheet of our customers is a proof point of the determination and their desire to make sure that they have capacity set aside for them. And as we go forward, and we'll talk a little bit more like what we've done in France, we're not done. We will continue to innovate on economic models that bring win-win outcomes for both us and our customers. So this brings me to the very heart of this company, the pulse of GF, our manufacturing. We have a global manufacturing footprint. We have 3 facilities in the United States, a 200-millimeter facility in Burlington, 300-millimeter, our most advanced, in Malta, New York, and of course, our Fab 10 facility in East Fishkill. Fab 10 facility, at the end of this year, will transition to ON Semi. This was a facility that was subscale for the type of foundry business we do, but it's super scale for Hassane and his team and the kinds of power devices they make. Great partnership, transitioning a manufacturing asset, repurposing it to be a profitable business. We have 2 facilities now in Germany, our Dresden facility, and now -- and we'll talk a little bit more about what we announced and what we're going to be doing in Crolles. And of course, Singapore, 200-millimeter giga fab, 300-millimeter -- and 300-millimeter extension. This is truly becoming our first mega fab. KC Ang will take you through a lot more detail, but let me say the following: For the better part of the last 2 to 3 decades, supply chains were managed to efficiency and cost. Maybe in an ideal world that can work out, but it's clear we don't live in one, and you're seeing how much companies that have overfocused on single points of failure, or businesses, suffering. The very nature of GF, the very nature is we have a global footprint, we have a global supply chain. We actually could share supply in parts of the world where maybe things are tight because we have a supply chain established for manufacturing in, say, Singapore or Europe or U.S., we can share. We add capacity to 2 dimensions, capital efficiency and economies of scale. You can rest assured you'll never wake up one morning and see an article or press release that says -- as long as I'm the CEO, that GlobalFoundries just added capacity in the middle of nowhere. There's 0 likelihood of that because of the economies of scale and the capital efficiency. We will always build out on existing campuses. We spent a decade to create a global footprint. If others want one, they're going to have to go through that journey. And the Crolles is a perfect example of that. We didn't go and break ground on our own facility. We did it in partnership with STMicro. And then the last thing I would tell you on this supply chain security is our customers, as this world goes from globalization to more regionalization, want to make sure that we can supply their products in a more globally diverse way. So we go and work with them and plan our footprint where we can do qualify technologies in our global footprint. So 1 part number from a customer can be sourced at 2 locations. And last that we talked about is our capital-efficient model. It's with partnerships, it's with customers, it's with governments, it's with GF to create the economic outcomes that make sense for us, which you will see from 2020 through 2025, the investments we have in flight. We will increase our manufacturing output 50%. Let me talk for a moment on our French expansion -- France expansion. So how do we think about it? It has to be consistent with our business model. First, we want to make sure we're putting capacity attracting -- towards attractive markets. You could see that this facility or this capacity is for IoT and auto. These are growing secular markets. This is where we want to grow. We've put that kind of capacity, and most of the capacity we're adding in our footprint is our differentiated capability. We're not going to put second source capacity online. All of this capacity is on our 22FDX proprietary technology. Not only did we get significant government co-investments from the French government and the European Union, the part that GF owns, 80% of our -- 80% of the cash required are customers who are putting their cash on the line. When we're all said and done in this module expansion in Europe, what we have in flight in Germany today, plus when we -- the full buildout of this French facility at STMicro, we'll have 3x the output in our European footprint. And this deal checked the boxes. It checked the box and accretive to our business model because we had government funding and advanced payments to go fund this expansion. I think more importantly, it's not only that we have a manufacturing footprint, but that our customers trust us. It's in our DNA. We don't compete with our customers. We're not confused about our priorities and who we serve. Our customers trust us to deliver. They trust us to deliver in the right way, and they trust us to deliver a secure supply chain. And trusting to deliver, we have to deliver on time, 99% on-time delivery. We have to deliver with high quality. And when they're introducing new products to us, we have to get it first time right for them so they could get to market. Delivering the right way is about who we are as a company. The #1 operational metric at GF, the #1 metric in any operation meeting we cite and report on is safety, with the most safe company in the semiconductor industry. We do it, and we make sure -- when we build our capacity and build our product, we do it with conflict-free materials. And we have an eye and a thrust towards ESG. We're going to reduce our carbon footprint by 2030 by 25%, as we 50% grow our capacity. Now we're going to do -- this will come for free. We're going to invest $75 million to do that. Those investment returns are consistent with any other investment we'll make in this company. And then with trusted delivery is secure a supply chain, first and foremost, that we're going to deliver the supply our customers need us to deliver. And our customers trust us enough to deliver that, that they put $3.6 billion of their balance sheet on our balance sheet. They trust us to deliver flexibility and regional ability to deliver their product, and we cross qualify many of our technologies to do that for our customers. And not only do our traditional commercial customers trust us, governments around the world trust us, we are ITAR Trusted Foundry certified for the U.S. government. We get to build the most sensitive and secure parts for our national security. So 50% increase in manufacturing footprint between 2020 and 2025. We deliver that capacity through economies of scale and capital efficiency. We use dual-site qualifications technology to create -- to take advantage of our global footprint for our customers' supply chain and security, and we will reduce our greenhouse gases 25% between now and 2030, even as we add a significant capacity. So let's get to the last one about trust. We're trusted by all our stakeholders that we have a business model with strong earnings and visibility. Our customers need to know we're going to be around for decades. Our employees need to know that we have a robust business. I think many of you in this room want to know, we have a robust business. And so let's talk about that. Let's talk about it in the way we think about our business, certainty, durability and profitability. If you look out in time, 80% of our capacity through 2025 is covered by long-term agreements. By 2024, we have long-term agreements come in 75% of our capacity, and we have committed design wins in new business. We do not want 5 years of 100% coverage on long-term agreements. We need to remix our business. We need to innovate and continue to create capacity. Durability. Durability is about attractive end markets where we can differentiate and capture value. 90% of the design wins first half of 2022, were in single-source business. If you remember from our roadshow, we told you in 2021, 80% of design wins was single-source business. And then on profitability. First half 2021, the first half 2022, 19% increase is a mix between better mix, richer products for us and higher ASPs. So David will talk a lot more about our long-term market. We're going to continue to grow in the attractive markets and grow in the high single digits to low teens as the industry grows. More importantly, we will take our gross margins to 40%, and we will continuously be disciplined leveraging our economies of scale and partnership model to add capacity in a disciplined way. Long-term model 20% of revenue and capital deployment. So last summer, our journey to IPO. We told you our story. We told you a story about 160 times. And we told you what you should look for, we could count on what we're committing. Let's do a recap. We told you we would penetrate and grow in secular end markets, the ones that are important to us, the one was where we can create differentiation. Well, auto and IoT, 65% year-on-year growth in revenue. And in the first half of '22, $3.6 billion of design wins -- I'm sorry, in the last 12 months. We told you, to get that type of single-source business, we had purpose-built platforms where we're going to innovate. Well, 90% of our design wins in first half of 2022 has been on our single-source business. We introduced 22FDX+, it took the world's lowest power platform and took 20% power and performance into it. We qualified and launched silicon photonics platform and our RF Connex platform. We told you we continue to deliver on our commercial strategy focused on certainty, durability and profitability. Today, we stand here with $37 billion of long-term agreements, 36% increase year-on-year, and now up to $3.6 billion of customer balance sheet in play. We told you we would build modular capital efficiency in every piece of capacity we're putting on, does exactly that. But it's not just putting the capacity on. All our programs are on budget, are all on time. And all of them are going to increase our wafer output in the first half, you saw 10%. And then lastly, I think this is the most important metric we told you, you could count on. It's not about the revenue growth. It's about making sure we had multiples of profitability growth for every bit of revenue growth. Year-on-year, first half to first half, 29% revenue growth, 5x plus on gross -- on earnings, gross profit growth. So let me leave you one last chart here. This is the leadership team. This is a leadership team that leads 15,000 employees worldwide, 15,000. Dedicated 15,000 strong. 15,000 diverse. 15,000 working as one GF. Later today, you're going to hear from Ed Kaste, who's substituting for Mike Hogan who couldn't be here today on our end markets. We're going to ask Juan to come up and talk about our innovation -- I'm sorry, Gregg to come up and talk about innovation; Juan to come up to talk about our partnership with our customers; Saam and Emily to talk about ESG and our talent; and Dave will finish it up with our financials. And I want to leave you with one last thought. For the better part of the last 5 years, what we talked about GF was our opportunity, the opportunity we had. As we stand here today, it is a lot less about our opportunity and much more about our responsibility. Our responsibility to all our stakeholders, our customers, the communities we operate in, our employees and now, you, our investors. So with that, I'm going to hand it over to Ed. Thank you.

Well, thank you, Tom, for getting us off to a great start and for that introduction. Good morning, everybody. It's a pleasure to be here with all of you today. My name is Ed Kaste. I'm responsible for product management at GlobalFoundries. I'll cover the end markets overview. The purpose of this session is to give you insights into how we select our end markets and how we prioritize engagements. And we'll also get into the implications of this approach as it pertains to demand quality. As Tom mentioned, we measure this across certainty, durability and profitability. That will also set up a deeper dive on demand quality in Juan's customer engagement section. We've prioritized 4 critical end markets: Smart mobile devices, requiring a broad range of connectivity solutions and rich features driving an immersive user experience. This is our largest end market, both by SAM and current revenue; home and industrial IoT, pushing the limits of what it means to have seamless connectivity and a frictionless human-machine interaction. This market is both high growth and extremely diverse; automotive, rapidly becoming a smartphone on wheels. Semiconductor content is driven by autonomy, connectivity and electrification; and lastly, communications infrastructure and data center. This is the infrastructure powering all of the above, both in the data center and in the network. But please note, I won't be covering the PC market today as it represents a much smaller portion of our overall SAM. Before we move on, I'd like to take a minute to highlight how we think about where we play in an increasingly broad semiconductor landscape. We take each of these end markets, and we break them down. Each of these end markets call for specific applications with functionality that drives device selection. We prioritize these submarkets to pick the most important to GlobalFoundries from a growth and value capture perspective. You see a couple of examples on this slide. You also see the market opportunity doubling as we get to 2030, more than doubling. We'll illustrate more of the requirements and GlobalFoundries' capabilities to address these submarkets in the upcoming slides. Starting off with the smart mobile devices. As you heard yesterday, this is currently our largest end market and accounts for about half of our revenue. Now what are the key trends? As you can see on the left-hand graph, this shows the total handset volume over time. The growth, as we know, is relatively flat. There's a 3% annual growth there. But there are dynamics that are driving a SAM acceleration for GlobalFoundries. The first one is the 5G content. That's the purple bars. That's the subset of the overall handset volumes that are 5G-enabled phones. Those are growing at a 26% growth rate. That means additional RF content, whether it's for the sub 8 gigahertz or millimeter wave-enabled phones, both areas where GlobalFoundries has a strength. Additionally, there's feature growth that's enabled by the accelerated data rates, and this drives the overall content growth in the smartphone that Tom was highlighting. As you go from a smartphone like an iPhone X, 4G-enabled, 400 square millimeters of content. That explodes to 640 millimeters, a 1.5% increase in content, driven by the RF and the additional features enabled by the data rate. And then the third factor is the companion device market. These are the wearables. They're virtually tethered to the smartphone and part of this end market. The fitness trackers, the ear pods, the watches, the emergence of AR/VR glasses. Altogether, this means a $23 billion market and expands to $40 billion by 2030, an enormous opportunity. Now we'll go into the application areas in the smart mobile devices end market. A good way to think about the smart mobile devices requirements are on the 2 sides shown on this page. Shown on the left-hand side here is the core functionality, the communications. This requires best-in-class connectivity, whether that means cellular service up to millimeter wave, or Wi-Fi. And the Wi-Fi standards are also evolving, where Wi-Fi 6 is the state-of-the-art today, and Wi-Fi 7 in the future. There are other communications not shown, but also on the left-hand side, Bluetooth Low Energy, typically used for the audio connection. Near-field communications, may be used for secure transactions. UWB, there's more. And then on the right-hand side of the plot here is the user experience. As the data rates continue to increase, it unlocks new use cases, whether that means streaming ultra high-resolution video, enabling immersive audio quality and other use cases. And the smartphone never leaves our side. It's become essential to making secure transactions such as wireless, contactless payments or identity verification in the mobile wallet. This is becoming increasingly important in the post-pandemic world. And all of this has to be enabled knowing that power matters everywhere. The expectation is still that a charge lasts the full day even though you have additional features and capabilities. All of this means sophisticated power management. Now getting into the requirements of some of those applications. The first aspect is the communications. We sometimes forget that this was the original purpose of the phone, the selfie, taking pictures of your food, that all came later. So what does it take to enable best-in-class communications? Rate. This means high-speed and high-bandwidth data transfer. Range. Mobile means on the move. And as we move around, we don't tolerate service gaps. That's a thing of the past. And power, getting all of this enabled on a single charge. GlobalFoundries continues to lead in RF solutions that enable best-in-class connectivity. For fully integrated solutions, FDX is a winning play. For specialized functions, RF SOI, silicon germanium, and soon, gallium nitride devices. GlobalFoundries is the #1 foundry supplier of RF SOI devices, given the leadership capability that it provides, particularly for the RF front end. That's the portion of the phone that connects to the cellphone towers. In fact, if you break open -- you won't do this. If you break open the front-end module from any premium smartphone, you're very likely to find GlobalFoundries content inside for the power amplifier, the low noise amplifier and the switch. That's where this capability matters the most. Second application area that we described is the user experience. Again, high data rates unlock new use cases. Think about how that occurred in 4G, and that's surely to happen as 5G deployment becomes more pervasive. Now we all demand a seamless interaction with our mobile device, and this human-to-machine interface is analog. Icons are dragged around the screen. This only works if it works, right? What differentiates here is low latency, high accuracy, haptic response that's replaced the physical buttons on the phone. And commands are given by voice. These voice commands are picked out of a noisy environment, people have different intonations in their voice, and the phone has to be able to process all of that. Image sensors are required for high resolution, high dynamic range. This gives us great photos. This gives us image classification, which enables AI, and has to work in all lighting conditions. And given the proximity to the user, all of the places we take our smartphones, trust cannot be in question. And again, since this last link in the chain is the human-to-machine interface, it calls upon feature-rich solutions. FDX technology for the fully integration -- fully integrated architectures to manage sensor interfaces combined with ultra-low power digital processing. And BCD and NVM technologies. If you heard our earnings call yesterday, Tom introduced the industry's first 55-nanometer embedded, nonvolatile memory solution for power management. Look for this feature to be deployed in the next-generation premium smartphones. A high level of integration of these features is what wins. So to wrap up this end market, we achieved $2 billion of revenue in the first half of the year with good growth. The total design win for the trailing 12 months was $3.4 billion, and a massive $16 billion worth of LTA have been signed in this end market. And this gives us some resiliency to the macro cycles and seasonality that may occur as we drive towards the longer-term growth trends. The beauty of our proven differentiation in the smart mobile device space is 2 things. One, it gives us scale because of the volumes involved in this market. But also it helps us create innovation that's directly applicable to our other priority end markets, as you'll see in the upcoming slides. Okay. Moving on to home and industrial IoT. To get started, I'll remind you that home and industrial IoT contributed 17% of our revenue in second quarter with really strong growth, 72% growth year-on-year for the quarter. Now to put that growth rate into more context, home and industrial IoT is an extremely diverse space, you see the icons on the left-hand slide, representing a very broad array of use cases. And in terms of unit volumes, the numbers are truly staggering. The number of smart devices per person, and Tom indicated, is expected to grow over 2x as we get into 2025. And when the multiplier, 2.5x of what, it's all of the human beings on the planet, all of the billions. So the numbers really add up. We're talking about 30 billion devices by 2025. And during the course of the pandemic, the number of connected IoT devices actually eclipsed all other connected devices combined for the first time. This includes smartphones, laptops, tablets, all other connected devices, and this trend is not turning back. But there's even more to it than that. The real driving force behind this growth isn't the consumer experience. It's the monetization of the data by hyperscaler, Internet companies like Google and Amazon. In this model, every new data point that's added to the database from a user transaction or some preference, it has more value than the data point that preceded it. This is the power of aggregation. And so you can see what that means is the economic value of the IoT data themselves is growing even faster than the unit volumes. And all of these things contribute to lead us to a $32 billion market opportunity by 2030. All of this content, virtually all of this content is sitting in 12-nanometers and above. Okay. Getting into the applications for IoT. You'll notice that this slide mirrors what I showed you in the smart mobile devices. The same attributes on the left-hand side of seamless connectivity, frictionless human-machine interface, and trust, they all play front and center, just like in the smart mobile device. But what's different is the huge diversity of applications implied by the right-hand side of the chart. In the smart home, it's the smart speaker, the video camera, the smart thermostat, et cetera. In the industry, it's building control and precision farming, in factory automation and safety, all of these use cases. And all of this implies a very wide range of products, a wide range of form factors, different environmental conditions that these products are deployed into. All of this creates an extremely attractive market for feature-rich solutions. And this part number diversity means that the barrier to move those products into single-digit nanometer is even greater because of the design costs and the mass costs and the cycle times to validate those products. It's really prohibitive for this market because of this -- the broad use cases. Again, feature-rich solutions will win over brute force shrink. Getting into the application areas for IoT. We start with the wireless capabilities. This is video streaming, connected camera, smart home and so forth. Capabilities required similar to mobile, but there are aspects that push them even further. Rate in this case also means seamless connectivity, right? A user buys a new device, they expect it to connect to their home network seamlessly, right? If it's hard, the data shows that the user will give up somewhere between the second and the third try and return the product. So it has to work. Range is key. Every 50% increase in range for transmit or receive translates to a 225% coverage area increase, very important in the home. You have rooms that are further from the Wi-Fi router, harder to reach, and so range is key. And then power is the ultimate scarcity in IoT. Compared to mobile devices, the battery is much smaller. AA battery or a coin cell battery or printed energy harvesting solution. Yet the expectation between charges is not days or months. It's years, right? Think about a smart home deployment to charge one of those devices. First, you find the adapter, right, from the drawer, that's not easy. Then you get out the ladder, and how many times do you want to do that, right? They can't be more than every couple of years. So this is an ultra-constraint problem. What wins in this market? The best wireless. The best range at the lowest possible power. Integration is also a key differentiator. If you can combine efficient communications with sensor connectivity and ultra low-power edge processing, it delivers the longer battery life and the smaller form factor for more IoT deployments. All of these things cry out for FDX. This is the industry's strongest solution to integrate these capabilities onto a single chip. That's the reason why we've shipped over 1 billion FDX chips to the field. It's also the reason why we're providing the solution to the #1 Wi-Fi provider for IoT in the industry. Moving on to the second application area for IoT, the human-machine interaction. The frictionless interaction with our devices is critical. This is because IoT devices don't ship in a box with a keyboard and a mouse to set up, right? The interaction has to be easier and better. It can't be more complicated. Otherwise, the devices don't serve their purpose. They live in the junk drawer. And this requires solving that last interface between the digital world of computing and the real world, which is analog. And analog functions don't scale well into single-digit nanometers. That's what makes this such a target-rich environment for the broad array of GlobalFoundries solutions shown here. The other aspect of the synergy is that winning in wireless, which we showed you on the previous page, either helps us win the integrated solution in FDX, or it gives us an opportunity to present companion technologies like feature-rich CMOS, BCD and NBM to win in that product space. And if any of you have ever given a voice command to Alexa, it's extremely likely that both the voice processing and the data communications were both done through GlobalFoundries silicon. So to wrap this section up, and as mentioned yesterday, the revenue in this end market for GlobalFoundries, $669 million for the first half of the year at very strong growth rate. This puts us on track for this to be our fastest-growing end market for revenue contribution. It's supported by over $2 billion of new design wins in the last 12 months, and a healthy $3.9 billion of signed LTAs. We're winning with the winners in IoT. Notice among these logos, you have the #1 smart speaker brand, the #1 DTV brand, several of the leaders in contactless pay and transactions, and other #1s in their respective markets. Now let's take a look at how feature-rich silicon applies to another critical end market, automotive. So to frame this next section, remember that the automotive market contribution to our revenue today is modest at 4%, but again, with very strong growth. The second quarter results represented a 34% year-on-year improvement. In the second half of the year, as Dave mentioned yesterday, we expect an additional 25% increase relative to first half of the year as critical capacity comes online to serve the automotive market. And again, you'll see some parallels to the previous section. After all, what does the automobile become? A very large smart connected device, right? And inside the automobile is actually a mini IoT network. So these are the parallels that create a perfect extension from everything that we just discussed into this market. On top of that, there's electrification. And this is very important, similar to the mobile discussion, because the growth rate of electric vehicles is much faster than the automotive market at large. That's shown in the left-hand chart, where the overall auto shipments per year are expected to grow at about 4%. But if you look at the purple bars, that's the EV contribution, growing much faster at 15%, and leading to a point where by 2030, we expect about 1/3 of the new vehicles to be electric vehicles. And then there's the content story. Just like in the mobile phone, where if you start with an internal combustion engine vehicle, Level 1, that's basically cruise control capability, that has $500 of semiconductor content per vehicle. And then this steps up as you go from ICE to an electric vehicle, and then you step up the level of ADAS from Level 2 to Level 4, you get to $3,000 of semiconductor content per vehicle. That's a sixfold increase. All of this means that the industry is going through unprecedented change. And if you combine that with the very well-publicized supply chain crisis, which we'll get into some more, this really creates a perfect storm-sized opportunity for GlobalFoundries. So getting into the application areas for this end market for GlobalFoundries. Again, the new automobile experience is no longer defined by the internal combustion engine, or ICE. It's now defined by ACE. That's autonomy, connectivity and electrification. On the -- starting on the upper left-hand side, the key to autonomy is precision sensing. These are capabilities like RADAR and LiDAR. Automotive RADAR is designed in at extremely high frequencies, 77 gigahertz. This is well above the frequency of even 5G millimeter wave communications. And our portfolio has real strength here, as evidenced by the announcement that we did last year with Bosch around their selection of a next-generation RADAR platform. Moving on to connectivity, on the upper right. The connectivity includes both the car to the network, which could be using a 4G or a 5G type of connectivity, as well as networking within the cabin. And again, this mirrors what an IoT network looks like in a smart home. And then on the bottom left, electrification, which requires efficient power conversion and delivery to ensure that most drives can be accomplished on a single charge and more. As the vehicle becomes more self-sufficient, the passengers are expecting an expansion of features for comfort and convenience and entertainment. And so for us, and as Tom mentioned, this makes the right analogy, the smartphone on wheels, right? Others may have called it the data center on wheels, which for us, puts a lot of emphasis just on the central computing function. But if you think about all of the features to create the experience in the automobile, that's why we think the smartphone on wheels is the right reference. Let's dig in further. So starting out with the application area of sensing, which is really fundamental to all levels of ADAS. We get a lot of questions about autonomous driving, the technology readiness, regulations that might be required and so forth. But one area where there is no debate whatsoever is over the critical function of the sensors and the radars in all levels of autonomy. The capabilities required for automotive sensing are just like the human eye or the human senses that the machine is supplementing and then ultimately replacing. The key questions are things like how far can you see? How wide can you see? How well can you resolve images in a wide variety of backgrounds and lighting conditions and environments? And since the world is very dynamic, being able to do the sensing in parallel with some level of edge computing for intelligence in the sensing is essential. And again, that all must be done at the lowest possible power. So our portfolio over on the right-hand side includes FDX, silicon germanium and other technologies that are delivering these requirements with leadership performance in elements such as the highest channel densities for RADAR that are available. This is the reason why we are the foundry supplier to the #1 RADAR solution provider in this market. The second application area, as I described, is power, and it's really become the new currency in the automobile. Why? Well, range anxiety is very real, right? It's actually limiting the uptake of electric vehicles, with 300 miles being a critical threshold to provide enough coverage for most daily drives. Just like in the smartphone, this enables a full day of use on a single charge, and then a full recovery overnight so that you're ready to start a new day. Also, like the smartphone, there's a conflict, right? The thirst for new features within the automobile is in direct conflict with the primary function of the automobile, which is the transportation. This all means that the creation, the conversion and management of power is more important than ever. And GlobalFoundries' solutions, like BCD, have industry-leading figures of merit for this space measured in attributes like Rsp, Ron. They also support the widest range of operating voltages and a road map to gallium nitride power devices. This is the reason why we're with the #1 battery management supplier to the #1 OEM in North America. And the last point I'll make on this is that you can't just become a great automotive manufacturer overnight. This takes decades of investment, manufacturing systems improvement, quality control systems improvement. And at GlobalFoundries, all of this is encapsulated in a service called AutoPro. We have 1 additional slide for the automotive section just to talk about an important topic, which is this, the automotive supply chain crisis. Of course, this is an ongoing issue that was initiated sometime during the pandemic when the surge in demand for laptops and equipment to work from home and school from home caught most people off guard and especially the automotive industry who is lagging to place their upside in orders. And we wanted to just take -- spend a minute describing what were the factors that broke down. In the automotive market, there were issues described on the left-hand side of the page, tiers of intermediaries, indirect communication. This led to reduced visibility and control in the supply chain. It led to opaque planning and forecasting. The third item here is it led the separation of R&D. This led to a technology latency and lagging requirements, mismatching to requirements. Think about how fast the automotive market is now trying to evolve, and the R&D was out of sync with that and now playing catch up. And overall, these relationships were characterized as very tactical, meaning they lack strategic planning, and underinvestment was the result. The good news is on the right-hand side of the chart, we're seeing nearly all players in the industry now acting with urgency to make sure that we don't repeat this type of a mistake. And GlobalFoundries as a beneficiary of this environment is already implementing and taking part in structural changes to the industry. The one -- the direct foundry to automotive OEM relationship as well as better relationships through the supply chain have created more direct working exchange. That's been helpful towards planning near-term supply, but also getting road maps aligned for capabilities that are required in the future. And better partnerships, right, which includes investments. I'm sure you took note of the announcement of an MOU signing that we had with Ford. And in fact, earlier this week, Tom was with the CEOs of both Ford and GM and other folks in the industry at our CEO Summit just on Monday. So surely, this is a key focus area for GlobalFoundries, and you can expect more good things to come. So to wrap up the automotive section. In the first half of the year, we achieved $163 million of revenue in this end market. That's a 79% year-on-year growth. The leading indicators are very solid. So the design win total for this end market, $1.5 billion. To put that into more context, enterprise-wide over that period, we had about $10 billion of new design wins. So the automotive portion of that was 15%, much larger than the 4% current revenue share. So this indicates a remix of the portfolio and growth of the automotive content in our portfolio. And then lastly, $2 billion of long-term agreements have been signed in this end market. And honestly, this is an end market where it's really most important for the customer. Once the products are designed in, in automotive to a foundry solution, it's pretty difficult for them to be -- to move tactically. And so once the investment has been made to design into the technology, it's in everybody's best interest to make sure that the capacity is there long term. These products have a very long run time. As an example of that, the state-of-the-art today in automotive microcontrollers is in 40 nanometers, 40 nanometers with embedded nonvolatile solutions. And in that space, we are the sole supplier to 2 of the top 3 automotive MCU providers. It's all of these elements that give us confidence that automotive will be an $800 million run rate business for GlobalFoundries by the fourth quarter of 2023. Okay. Closing out in the last end market. This is communications infrastructure and data center. The last 3 end markets were all about massive data generation and consumption at the edge. This last one is all about getting that enormous amount of data to the cloud where it can be managed and returning insights from the data. As reported yesterday, this end market contributed 17% of our revenue in the second quarter. Again, with very high growth, 50% growth year-on-year. And this growth should come as no surprise given the topics that we've just discussed earlier. As the number of connected devices per person grows, right, by 2030, 3x growth. The data generated by that grows as an exponential. So 9x more data generated. And then the corresponding Internet traffic generated from that, a 20x increase. And so these are the factors contributing to a projected 4x growth in the cloud computing market by 2030. But again, it's not just the data center. This brings new requirements for infrastructure to pass around all of this data. 5G infrastructure will help, and it requires a much greater tower density in 5G to achieve the same coverage area. So you see there are 4x growth in tower deployment. It also leads to the implementation of new networking types such as low earth -- networking from low earth orbit satellites. And you can see that's a very high-growth area with 6x growth projected in 20,000 deployments by 2030. So these are the trends that are expected to sustain a 9% growth rate and place a heavy demand on feature-rich solutions. Now getting into the application areas for this end market. There are really opportunities for GlobalFoundries in 4 major categories. On the left, the first is chiplets, where GlobalFoundries has a strength. This is, for example, the interface devices that a company single-digit nanometer cores in the data center. And these interface devices are a fairly large area. They place a heavy demand on the analog functionality. And again, the analog functionality doesn't see a strong benefit to moving to smaller geometries, and you've seen some details of our ongoing engagement with AMD in this area, including an LTA extension that was publicized at the end of last year. In wired networking, at the top of the page, there's this exciting transition underway from electrons to photons and the rise of optical networking. This is a key shift expected to improve data rates, improved reach and most importantly, deliver very significant power reductions. Because of those benefits, we feel this transition is inevitable, and we'll cover the details in the upcoming slide. And then wireless communications on the right-hand side. This is where there are parallels in our mobility business. This covers the base station side of 5G build-out as well as emerging applications like satellite networking. And as before, on the bottom, each of these applications calls for efficient power delivery and management. We know every jewel matters in the data center as this drives the total cost of ownership. But it's more than just the cause of ownership or it could be more than the cost of ownership. It's sustainability. On its current trajectory, the way the data center is going, it will eventually overrun the Power Grid's ability to supply the power that's required to sustain the virtuous cycle. So innovation is a must-have in this area, and that's where GlobalFoundries fits in. So getting into this application area of the 5G infrastructure portion, this is a natural growth area for GlobalFoundries just as 5G delivers a great improvement or step-up in data rate it comes with a handicap. There's an inherently shorter range with 5G because of the frequencies that are involved. They can't penetrate through solid objects. So that's not a good trade. Most of us wouldn't trade better data rates if it meant more coverage gaps, right? So to solve that problem, and we see investment in more infrastructure. This takes the form of new base stations and other networking types that I mentioned. And there's a strong overlap of requirements for this space with our mobile business that is the best wireless wins. And depending on the architectures that are chosen, FDX technologies provides leadership capabilities in an integrated solution and for the ultimate in low noise and higher output power, it's silicon germanium to provide the best range and lower operating costs. And then the strength of this product line has made us a perennial supplier to the top cellular infrastructure brands in Northern Europe, for example. And then the last application area to dig in. We saved one of the best for last year. This is all of the excitement around the emergence of optical network and the GlobalFoundries photonics solution. This technology breakthrough leads to a number of things, higher bandwidth. We're talking about multiple terabit per second data links. Low latency. This allows for memory access in real time and more efficient machines. The reach, that's enabled with optical network, nearly infinite, right, with very low losses that allows for the disaggregation of the data center. And then the energy savings speak for themselves. For the key figure of merit, which is a picojoule-per-bit transferred, we're expecting a 4 to 5x reduction in that key figure of Myriad, driving down to a single picojoule-per-bit. We have the leading -- the industry's leading solution for this application space with our GF photonics. It's the only -- it's the world's only 300-millimeter fully integrated CMOS RF silicon photonics platform. It's targeting that next generation of architectures at the 1.6 terabit per second point, and we expect great growth here. The revenue contribution is modest today. We're projecting about $100 million for the year 2022, but that represents just under half of the total market share for optical networking, and we expect that market to grow to over $1 billion opportunity by 2026. Throughout that time, we expect to make steady market share improvements as growing along with the market, based on the strength of the clients that we're engaged in with in this space. This opportunity has been a couple of decades in the making, GlobalFoundries is extremely well positioned with the likes of NVIDIA, Marvell, Broadcom, Cisco, others, including some very innovative startups. So to wrap up this section, both the revenue and growth rates for this end market remain compelling for GlobalFoundries. $2 billion of design wins, $3.2 billion of LTAs have been signed. And given the strong engagements that we have in this space, with, again, AMD, Marvell, Cisco, Microchip, Broadcom, all of those listed here. The future of this end market remains very bright for GlobalFoundries. So to wrap up, hopefully, this section delivered as advertised some insights into the dynamics of the end markets that we prioritized and implications on GlobalFoundries domain quality. With that, I'd like to welcome to the stage, Gregg Bartlett, our Chief Technology Officer, for a deeper dive into our portfolio and road map. And I'd like to thank you all for your time and attention.

Good morning, everybody. Chief Technology Officer. And guess what? I'm going to talk about technology. I actually want to start the technology talk by using the exact same slide that Ed used to summarize the discussion about the end markets because that's how we develop technology. It starts with the diversity of the end market applications that Ed just spent time describing to you, and us formulating solutions for these very attractive markets with great value capture opportunities for us. For GlobalFoundries, when we develop a technology solution, it's because we understand that end market and that it doesn't matter if we think we have a differentiated technology. It's when the customer has a set of features that allows them to put a superior product in the end market. But because -- and if you added it up, Ed showed more than 200 end market applications or subsegments. So that's a very broad area of diverse technologies. So when we think about creating an end market technology portfolio aligned to that, there are some very important considerations. I want to highlight 2 of us. First and foremost, what is the efficient way of using your R&D dollars to maximize your coverage across those markets. It's by identifying common technology features, common traits and attributes and using those to comprehensively deploy those throughout the portfolio. That's number one. Number 2 is the intimate understanding of how that product is differentiated in the end market and making sure those features are there in a way that differentiates that end customer. So in the next 20 minutes or so, I want to dive into both of those topics. And let me actually start with the same picture of these end markets. There are 4 trends that are critical as we think about developing our technology portfolio for it. And the first one is it's all about the features. And that doesn't mean that the underlying purpose-built platforms that Tom described this morning, or the digital CMOS, ultra-low power platforms, aren't important. They are. We continue to invest in those. But once a customer has picked a platform, what they need to know is that the feature that will differentiate their product against their competitors is part of that technology, and it's a feature. In a minute, I'm going to show you a really busy chart showing just how many features we're putting into our technology portfolio. You heard Ed say it a couple of times, best wireless wins, in fact, depending upon the application, best wireless meaning best range coverage because you've got high dB power output from a power amplifier. You've got very high frequency, allowing system-level designs for very high-frequency switching in your applications. It's low on over F noise so that you have a good signal to noise in your communication circuits. So there are very few applications in which the best RF is not the ultimate decider of success in the marketplace, only possible exception is low-end smart mobile device, where cost is a preeminent consideration. Global Foundries concentrates in developing our premier RF technology solutions. You heard a lot about automotive and the diversity of the applications. The vast majority of the technologies that GlobalFoundries creates ultimately finds its way into the automotive marketplace. Sometimes purpose-built. Sometimes a decade later, somebody starts putting what was originally an RF SOI technology for a phone into a vehicle because you have to have a cellular transceiver with a front-end module component in it. So we'll talk about just how pervasive and deep our technology portfolio is in addressing automotive. And the ultimate differentiation is always around power whether that is a continued improvement in the digital platforms for battery-powered IoT devices or something as disruptive that Ed just talked about, the migration away from copper and electrons to glass fibers and photons. I'll talk about that as well. So let's talk about the scope of the portfolio of technologies we create. On the left-hand side, you see an array that, quite frankly, probably looks like hyroglifics to many of you. The core to this is, in fact, that the teal colors are representative of the new features and platforms we're building in our portfolio of technologies today. So to the point about -- it's all about the features, you see the representation of the concentration we're putting into delivering those features. In fact, on January 1 of this year, we had 120 qualified technology solutions in our portfolio. And to be very clear, each of those technology solutions has dozens and dozens of configurable elements, meaning that customers get to pick the SRAMs they want to have, how much copper interconnect, what kind of resistors and capacitors, et cetera, that they want to have to differentiate it. It literally creates thousands of configurations for our customers in the end markets. In 2022, we're qualifying 16 additional technology solutions in that portfolio. In the earnings call yesterday, Tom said that 9 of those were actually in the second quarter of the year. I'm going to talk about 2 of those that are really instrumental to our technology portfolio. We're also in the process of developing 6 new technology platforms, each of which will have dozens of features over the years and decades to come. And then something that Tom and Ed already highlighted is the introduction of our First Beyond silicon solution, the introduction of a wide band gap material for GlobalFoundries, that's gallium nitride on silicon. What do you want to know about wide band gap materials, not about the band gap structure of electrons in that material. It's about the fact that, that material allows you to have very high voltages and power conversion, 650 volts operate at very high temperatures and very high power, things silicon technology simply cannot do. We see good market applications for those, specifically very high performance and high switching RF components like an RF GaN PA or 650-volt power conversion for electrification of powertrain. So that's new to our portfolio, and we're excited about the growth opportunity that represents. The second topic was best RF wins. In fact, if you look across the 7 technology families represented here, you see that we have RF technology solutions in all 7 of those families. I want to talk about a couple of those though that were of note. On Monday, the $4 billion extension of our long-term agreement with Qualcomm was all about 12 LPRF. You'd say, well, what's so special about that Qualcomm would sign an agreement extending their contract out through 2028. First and foremost, it's because the RF performance on the 12 LP, which is our FinFET platform, is a very good RF technology. We have further optimized that technology for Qualcomm and a select other few customers to precisely meet the needs that they have for their end markets. But why does it stay so long? As it turns out the circuits you designed in RF and analog features don't really like to scale very well into single-digit nanometer technologies, which is why Qualcomm in partnership with us extended to a material amount of demand through 2028. One of the points I made earlier is about looking across various markets and identifying common attributes that can serve multiple markets. 22FDX, first and foremost, as an ultra-low power platform, but with RF content, exemplifies exactly this point. Three different market applications. Smart mobile device, Qualcomm has selected this for the 5G millimeter wave front-end module as well as their envelope tracker, which has to switch it, very high frequency for optimized power management. In addition, as Ed mentioned on the Alexa devices and many other IoT areas, that ultra-low power platform with embedded RF is the perfect solution for WiFi 6, WiFi 6E and all of the other connected devices that go along with it. And then thirdly, because of the very high-performance RF on this platform, it's also been selected by a large number of automotive manufacturers for the 77 to 81 gigahertz automotive RADAR capabilities. The last area I want to highlight is, again, another technology reuse feature across our portfolio. Several years ago, we created a technology for 5G infrastructure called 45 RF SOI. Said, well, isn't 45 kind of a strange node. Actually, if you research the physics between RF performance, to find out that it's a sweet spot for the maximum frequency you can get in a digital platform. 400 gigahertz is available on this platform today. As it turns out, because of the very high performance in switching of the drivers and trends impedance amplifiers and silicon photonics, you want that same very high frequency RF in silicon photonics. So we have lifted all of the knowledge garnered from deploying the 5G infrastructure piece with 45 RF SOI and brought that into the silicon photonics platform. Okay. Let me address automotive. You heard a lot about the diversity of the applications there. It is a smartphone on wheels, as Ed mentioned. He also indicated that it's an IoT network in cabin. All of those things are true. And with that, scales the breadth and depth of the technology solutions to put vehicle-to-vehicle RF communications in place, vehicle to infrastructure for autonomous driving, certainly, in-cabin LED lighting for power management devices, powertrain solutions and, of course, increased amount of MCUs with embedded nonvolatile memories on there. As such, our portfolio of automotive technology spans 6 of the 7 portfolios. The seventh one, of course, is silicon photonics, and we do ultimately believe it will be a very good LiDAR solution, but it's a little early in the market for that. The second point Ed made that I want to reinforce as a technology guy is it is not a casual thing to wake up one morning and say, I want to be -- start manufacturing automotive semiconductors. Let me give you an example. An automotive microcontroller for automotive grade 0 requires the following: 175-degree C operating reliability qualification. Compare that to your consumer mobile device that all of you have, that's qualified at 85 degrees C. When you go to 175, physics starts breaking down left, right and center, and you have to spend years learning how to solve problems to achieve the qualification at 175 degrees C. It also, because it's used for life safety applications like autonomous driving and braking, needs to be better than 1 part per million. Now this is on top of 175 degrees C, better than 1 part per million. We do that by having all 5 of our manufacturing facilities adhere to TS 16949, which is an automotive standard for semiconductor manufacturing. There'll be a quiz on that later. All of our factories are qualified to that because of the precise controls you have to exercise and quite frankly, the improved level and use of metrology in the line to look for in-line defects to make sure that they are weeded out from those products. So with that as context, I want to highlight 3 really important technologies in our automotive portfolio. The first one is 130-nanometer BCD technology. So what's so special about this particular offering? Our customers know it because, in fact, it's one of the fastest-growing demand that we have in automotive applications today. What's special about it is it's 130-nanometer. It's automotive, grade 0, with embedded nonvolatile. And if that isn't enough, we added 85 voltage handling capability to it. And quite frankly, the fact that we're actually now as a consequence of discussing with automotive manufacturers and customers that they want even higher voltage. We're in the process of qualifying 120-volt capability on there. That combination for automotive grade 0 is a true differentiator for us, which is why we're seeing such a significant market traction. The second category or technology I want to address is 40 nanometer. It is a powerhouse technology node for us. Ed started by saying, when he was describing automotive, 2 of the top 3 automotive microcontroller companies are using our embedded nonvolatile. This is those guys. One is automotive grade 1. The other one is automotive grade 0. They're both in high-volume manufacturing to date in our Singapore operations, both have asked us to actually qualify a second manufacturing facility to produce those. One of those 2 guys actually happens to also be using that same 40-nanometer technology node for their millimeter wave RADAR in advance of the migration to 22FDX. To best exemplify the breadth of the portfolio, GlobalFoundries has embraced, RADAR is such a great example. On the bottom of this page, you see the system level architecture comparisons, some of our customers want to use a single integrated source that has transceivers and front-end modules and digital logic all in a single chip. 22FDX is the perfect solution for that, and you heard it described that. Others are looking for best-of-breed technology and individual die and whether it's our AXP silicon germanium or the 40-nanometer that I just mentioned, GlobalFoundries is ambivalent to how our customers want to approach it. We support both of those system-level architectures, and we do it comprehensively, 24 gigahertz for short-range curb bus sensing capabilities. short-range, medium, long range and imaging in the 77 to 81 gigahertz category and even for gesture sensing, the 60-gig in-cabin capabilities. All of those features are covered by our portfolio. It is not simple to deliver this. I've already talked about the fact that all of these, of course, are automotive grade 1 or grade 0 solutions, but we have a comprehensive portfolio. To further that, we also know that, as Ed described, the changing business model, we're no longer dealing with traditional fab or -- IDMs or fabless guys. As evidenced by the announcement 5 quarters ago with Bosch, we have gone and met them where they sit in the hierarchical perspective on automotive designs and enable them with design services and design enablement so that they knew how to design and a technology that previously they had fabless semiconductors companies or IDMs doing for that. As such, 5 quarters after that announcement was made, we have already run for silicon on Bosch's millimeter wave RADAR design, and they have silicon in hand that they're going through their characterization. The last area or category I wanted to address is, in fact, low power as the ultimate differentiator. And let me give 2 examples of that. First and foremost, Tom mentioned in the earnings call yesterday, 22FDX+ was qualified in the last quarter. He also mentioned that this is an extension on top of what is already the benchmark for low power and integration of features in our existing 22FDX. But the plus part of this is a significant improvement in the power and performance of the transistors in the platform. It did not come from technology guys reengineering the suite of transistors in there. It actually came from my colleague, Dacey Yang and his manufacturing organizations control of the distribution or variation of the transistors in the line. When we recognized how much tighter they could operate that manufacturing line, we monetize that into the hands of our customers. By giving them a 20% frequency improvement, they're able to monetize that as power reductions. How if they're operating a core at 300 megahertz in 22FDX, today, that takes a power supply at 0.8 volts. 22FDX is capable of realizing that same operating frequency at 0.65 volts. Power is a square of the voltage at which you operate. So that 150-millivolt headroom that you create that to achieve 300 megahertz is better than 25% power reduction for that circuit on the -- that matters a lot. When you add the features on top of that, it's a truly differentiated solution, coming simply as a consequence of the maturity of the 1 billion devices of 22FDX that I mentioned earlier. The second, an extremely disruptive -- I'm glad I'm not the first person today to have to talk about on picojoule-per-bit, but you've heard it now several times. Why do you care about that? We would take the energy of the sun to power the data centers if we don't do something about it, and that's what's ushering in, and quite frankly, GlobalFoundries is the first to offer that solution of better than 1 picojoule-per-bit for data transmission in a data center. And it comes as a consequence of what you can do with photons and glass fibers instead of electrons and copper wires and SerDes interfaces. That benefit comes already at 100 gigabit per second. And as Ed talked about, as it scales to 400 to 800 gigabit per second per channel. That's how you start getting into the terabit per second data center frequencies in there. That only gets better as that frequency goes up. So let me talk about the Photonics launch for a minute. This is actually an article written for the press release that we made at the Optical Fiber Communications Conference in March of this year where we announced it, and you see the who's who of the customer list, whether these are quantum computing start-up companies or the who's who of the data center space, we have very broad adoption. The question is what's so special about this technology? And I'm going to geek out here for a minute and show you some really cool images of what is unique about this technology. Remember, I talked about efficient use of R&D. We actually start with the 45-nanometer technology platform. I highlighted the fact that the best RF performance in the digital platform comes from 45-nanometer, that's where we start. We add to that an amazing suite of optical elements. In this case, it's a transmission electron microscope image of the photon detector, germanium photodiode, that's an avalanche detector for the electrooptical conversion. We add to that the silicon and silicon nitride waveguides in circular patterns running around the wafer for the phase interactions and logic computing capabilities. We add to that a truly unique solution. One of the big challenges in silicon photonics is getting light in and out of the chip. GlobalFoundries has created a very elegant solution by selectively etching silicon channels into the silicon. You can create a channel made-for-fit to the 125-micron fibers that you lay into that channel. One of the coolest scanning electron microscope pictures you'll ever see is just above there. It shows 4 fiberoptic channels with one of the fibers removed from there, and you see where the center of the fiber actually couples into the wave guide on there. This was really difficult engineering. But as a consequence, we now have the world's only monolithic platform electrical, optical, fiber attached, monolithically integrated into a single piece of silicon. But we're not done. We're in the process of today of building both the technology and the ecosystem to be able to bring the indium phosphide laser on chip. Most photonic solutions are off-chip and you have to have yet more fibers coming in to bring your laser source. This allows you to metallurgically bond that laser source, right, and they're creating -- now you know why -- there were 10 guys that signed on to our press release when we announced this at OFC because it is truly a one-of-a-kind unique one. And it does usher in the 1 picojoule-per-bit promise that is necessary for data centers going forward. By the way, those customers weren't just there for the press release. They're all actively designing products. We have 12 customers taping out products on silicon photonics right now, dozens of NPIs going into our multi facility for the ramp that will occur in 2023. So as differentiated as a solution as I just described on silicon photonics and the promise that, that brings. It gives an indication of the level of innovation and commitment for research and development that we make in this space, but it also proposes an important question. On the chart here, you see the transistor count for the 5-decade journey of Moore's Law. As you go further up to the right, fewer customers, fewer number of suppliers in there. But Moore's Law still has its champions, right? Folks that are doing component research, looking at lots of novel technologies. Of course, TSMC is the benchmark in the industry for pursuit of Moore's Law. But the question that it begs is, then who's doing the R&D for the other 75% of the market for the 200-plus applications and the 7 technology families? Who's taking the obligation to make sure they're -- now I wish we would had a spoiler during the sizzle reel because it was kind of already announced its GlobalFoundries, launching and formalizing our GF labs approach. I had the pleasure of announcing that in May at our Global Technology Symposium there. It is a focused and deliberate acceleration broadening and enriching of the pipeline of technologies for all of the markets that GlobalFoundries are serving. It's an incubator of innovation for near term and long term. Let me talk about the construct of it. There's 4 pillars that go into it. First and foremost, physical sciences, whether that's new materials, new devices and advanced tooling and process a matter of fact with Applied Materials. Since we launched GF Labs, we have 7 statements of work on advanced tooling capability, not for single-digit nanometer but for the portfolio of technologies that we have. Design innovation in the last few years, and we're expanding as part of GF Labs, we're hiring system-level architects into our company from our customers and our customers' customers. They bring the knowledge of where the market and end market applications are going and influence circuit apologies, use of AI and design and certainly heterogeneous integration solutions. We're continuing to invest in that space. The third pillar is the markets in which we play. Not surprisingly, they're very much related to what Ed just covered. 6G and 5G is in our wheelhouse already. We have multiple solutions for 5G front-end module capabilities across 3 different platforms. So GF Labs is already focused on 6G and beyond. And if you thought the challenges of coverage in 5G were challenging, get above 100 gigahertz, and it doesn't go around walls. It doesn't go through wood. So there's a lot of challenges that come with that, and we're spending time on that. AR/VR, data center, certainly, the automotive markets. The fourth pillar of this is, in fact, the ecosystem that we build. This is not the 4,500 IP titles and the 100 people that Tom described in his overview, this is research and development ecosystem that we're creating, expanded university engagements, working with federal governments on technologies of interest or national interest and certainly lab-to-lab interactions. And let me actually build out the details of that a little bit. We have more than 30 partners in our ecosystem today, starting with universities. We make the shuttles or the multi-project wafers that we launch into our facilities available to the universities. We actually cover the cost of running that silicon. Why would we do that? University graduate students and professors get access to some of the finest foundry technology available in the industry. They do all kinds of crazy things. They design new circuits. They want to use it for different applications. We get tremendous insight on what our customers may be working on 5 years from now as a consequence of the universities engaging on it. They get access to great technology. So it's a great win-win. You see a large list of universities located with us where we develop our technologies. In the R&D and consortia institutes, whether it's IME in Singapore; IMEC, Front offer and Lee in Europe; and of course, SRC here in the United States, they been working with them for years. The new part of this relevant to the signing ceremony yesterday. There's 2 new logos on this page. One is [indiscernible] that has created a National Semiconductor Technology Center. The second one is ASIC, which is the American Semiconductor Innovation Coalition centered in the Albany Nanotech location. Those 2 consortiums are actually holding steering committee meetings in a few weeks that we're participating in to work directly with the federal government for the allocation of not the $52 billion that's going to fabs, but the $200 billion for research and development. And then finally, in government partnerships. GlobalFoundries has been working with EDB in Singapore, and certainly DARPA in the U.S. We're pleased to add the French state to this chart as a consequence, not just for the subsidies that are providing us for the manufacturing corridor in coal but for their commitment for research and development around the FD-SOI ecosystem. When you're innovating like that, protection of your intellectual property is quite an important consideration on there. Last year, we hired a third-party firm to independently assess our patent portfolio. They came back with some really important conclusions; #1 value in the portfolio for RFSOI, #1 in fully depleted SOI #1 in silicon photonics and #1 in silicon germanium BiCMOS technology. We weren't resting on our laurels when they made that assessment. In orange, you actually see the number of patents we were simultaneously filing in 2021, 130 plus for both RF and FD; 20-plus for SiGe and the silicon photonics. What I'd like to do now is to summarize and bring back the 7 technology families addressing the 200 plus, and reiterate the 3 core points I wanted to cover today. First and foremost, the efficient use of our R&D capabilities to cover the 200-plus end market subsegments that Ed described. By year-end, more than 135 technology solutions leading to thousands of feature-rich technologies. And finally, with the launch of GF Labs, a long-term commitment to innovation through 2030 and decades beyond. And with that, thank you. And I'll turn it over to Sukhi.

Yes. That's quite a bit of information in the first half. That was very helpful. We're going to do about 15 minutes of Q&A and then take a 15-minute break and then come back for the remainder of the portion.

So sub-Q&A session. So Harlan, do you want to take the first one, speaking to the mic though. We'll have Tom and Dave here as well.

Yes. Harlan Sur, JPMorgan. Thanks for hosting this very informative event. So Tom, during the recent earnings season, we're seeing a dispersion of results and guidance. The GF team obviously drove very strong results, very strong guidance, great visibility. You guys are on the right side of that dispersion, right? While many other -- of your peers and maybe even some of your customers are on the other side of that dispersion. What do you attribute this to? Is it your end market diversity? Mix? Is it share gains, differentiated technology platforms, LTA visibility? A lot of the things that you guys just talked about over the last 2 hours. And then for those that say, you aren't seeing the weakness, but it will show up in the next 1 to 2 quarters, what are you guys doing to cushion the company from the potential for a broader slowdown?

Okay. There's a lot there. I'll take the first half. So look, let me tell you the characteristics of what we're seeing kind of who's on the broad side to that. First, there are a number of companies that had the demand but didn't have the supply chain. And so that's back to this idea of having a robust supply chain. I think there's going to be an industry trend less globalization, a little bit more regionalization, a little bit more balance in how you do that. GF didn't suffer from that. We didn't miss a single wafer of output. Now it doesn't mean we didn't have to scramble like everybody else, but because of the resiliency of a supply chain, we didn't have that. The second thing is really about where we play and how we play. We're pervasive. We're in every end market, you saw this morning. And so when there's pockets of more weakness than others, we can offset and pick up on that. And we had the luxury, as we talked about in earnings, started the year about 25% more demand, and we had the capability or capacity to service. So we feed into some of that. But I think the part that gets missed all the time is the power of market share. I cited before, you say semis can't grow faster than the GDP, b*******, right? You can get market share. You could have a higher content. And what happens in upmarket, it's hard for anybody to see what's happening in market share. But let me tell you, if you're losing market share in an upmarket, it comes home in spades in the down market. in the smart mobile device is 50% of our revenue. Year-on-year, handset volumes are down, GF revenues up 14%. We're in important sockets with important customers. We're part of the 5G transition. Ed said, I bet you won't find a handset that's 5G-enabled without GF silicon. I'll tell you, I'll buy the phone you took a part that doesn't have our chip in near from you because that's a pretty safe bet. And then other sockets we're winning. So I think never underestimate, even in a softening of the market, if you're gaining market share, you can still hold your own in that. And I think that's the characteristics of the market. So David, anything to add that? Or how about the second part of the question? We'll be doing in case we're just not seeing things.

Actually build a little bit on what you mentioned. I think one of the underappreciated elements of our business is our growing single-source nature of the business. So as you become more of a single source supplier, which means you have differentiated solutions, then you have customers that want certainty of supply. And of course, we want certainty of demand and that leads to more LTAs. And then those LTAs, that growing LTA business, that gives us the ability to plan better. And that planning is important. And I think you've seen us deliver on those plans. And those plans enable us to then kind of insulate our business and be able to mitigate some of the perturbations that you see in the market. So for example, on the supply chain side, what have we done? Well, as we were signing long-term supply agreements with our customers, we were concurrently going back to our supply chain in securing supply. So when you heard from others that they couldn't get supply due to shutdown in China due to disruptions due to Ukraine and the Russia conflict, -- you didn't hear that from [indiscernible] And so our ability to be able to look through the demand, work with our customers with broad end markets enabled us to plan our business better. We're executing on that plan. And I think that's probably the single best way in which we've helped mitigate some of the perturbations in the market.

Insightful.

Do you have a follow on?

Yes. I just got a quick follow-up. So I believe that the recent Qualcomm LTA extension by $4 billion is, I think it's the start of an important churn, right? Because this extension was mainly focused on your CMOS FinFET 12 and 14-nanometer technologies. These are technologies that Qualcomm could have got in from. They're getting from TSMC today. They're getting from Samsung today and others, but yet they've chosen to drive $4 billion of their supply to you. It seems like semi companies are focusing more and more on business continuity planning, supply chain diversification. So similar to Qualcomm, are more of your customers looking at GlobalFoundries, not only as a supplier of feature-rich technologies, but also as a part of their diversification strategy for their mainstream CMOS technologies as well? And is this another leg of growth for the GF team?

Yes. I think it's about market opportunity. I think there are a lot of reasons Qualcomm signed up to do more. They're already designed in this. As much as you say, they could source it from others. Yes, they can go and try to redesign on TSMC's FinFET or Samsung, even though -- the Samsung GF technology was a joint launch to the industry, we've diverged in how we've added specialty to it. But I think your meta point or a higher level point is really there is the world cannot continue to have such high level of concentration, independent of geopolitical issues. There's a host of other issues. And where they have the opportunity to get a better balance in the supply chain and all things being equal, they're doing that. And I think Qualcomm, they were a leader in securing their supply chain early in 2021 with us. They're a leader in getting ahead of the curve on having a high degree of balance of where they supply. And when you're that big, you cannot be so dependent on anything. And I think that plays our strength. And I think that's the -- just like they were the harbinger of a lot of LTAs that we signed in 2021. I think they are the harbinger of what's going to happen with where another deciding factor on sourcing is can I get it in a more localized way with a foundry that gets me some diversity. David, anything?

No. Well said, Tom.

Raji? Oh, Mehdi.

Yes. It's maybe on silicon. For the team, and it's very interesting since IPO. You have been able to deliver and execute, especially on the margin. What we have heard this morning is feature-rich portfolio of services products. So why not put the focus on your core competency and tell us how the revenue mix is driven by these feature-rich or platforms rather than focusing on auto or smartphone. I think you have a lot of core competency captured in a single source. So what doesn't that make sense? We focus on feature rich mix and the details rather than debating smartphone units. And I have a follow-up.

Sure. I'll take that.

Yes. Take that, David. I mean we're not debating. We're choosing to report our business. Go ahead, David.

Yes. Let me maybe take a stab at that one. So it wasn't that long ago that we went IPO, and we actually had the same debate internally. And as we were having this debate, the reason we wanted to report by end market is because everyone develops their forecast, and it starts with the end market. So what's happening with the underlying market and the different segments -- and so you can build up your forecast that way. And as you saw from Ed's presentation, you take that forecast, you figure out which portions of those markets or which applications in those markets do you want to play in? What technologies do they require? What can you build? And your product suite to be able to be differentiated and then you report, of course, back out on those end market applications. Now I don't want to spoil the second half of the presentation, but we are going to give you a view of how our business is becoming more diversified and how it's becoming more differentiated over time. And so you're going to see some pretty good nuggets here in the second half of this presentation. I'll address exactly that one.

A quick follow-up. And maybe this is covered in the afternoon. But in terms of the wafer supply, are you going to continue to buy the wafer with the substrate? Or is there a different strategy looking longer term, especially as you add the silicon photonics in Germany and so forth?

Sure. So I think the question is around, are we planning maybe to be more vertically integrated with regards to our wafer supply. So right now, we -- as I mentioned, we're growing our wafer output from about 2 million wafers in 2020. We're increasing capacity to where we can deliver more than 3 million wafers in 2025. A good mix of those are associated with RF SOI, FD SOI and increasingly SiGe and GaN and other technologies. Right now, we have great partners with long-term supply agreements across all of those substrates. And as I mentioned, as we were negotiating those LTAs and with our customers, we were concurrently going back to our supply chain to secure that supply.

And just let me build on that. A great example of that is global wafers, an SOI wafer supplier for us. They're expanding their campus, investing about $600 million in St. Louis. $200 million of that to help fund that is prepayments from us. So we're doing the same type of thing with them. And our strategy right now is to work with those partners like our customers work with us and secure our pipeline and not try to get ahead of ourselves with integrating more elements.

That's exactly right. They're going to be a big RF SOI provider for us in the future exactly out of that facility.

There's one from Raji here.

You took the wrong seat.

Raji from Needham & Company. Just a question on your long-term supply agreements. It was discussed a little bit on the earnings call. But there's been several companies that have quoted charges from breaking long-term supply agreements. We had Qorvo last week. We had NVIDIA talked about even larger LTA charge. And I wanted to get a sense from you some of the mechanics of the LTAs, if a customer comes to you and says, we want to reduce the volume for whatever reason. How does GF react to that situation? How are you able to swap out that order? Can you swap it out on a one-to-one basis? Can you shift capacity either within existing customers or shift capacity to new customers? So I'm just curious on your fungibility because I think that's going to be an increasingly important topic if there is a deceleration in demand in some of these kind of key markets.

Yes. So we always start with the fundamentals. When we said about these long-term agreements, it wasn't to shift all the risk from one party to the other. It was about partnership. The foundries didn't want to go into the next softening, and have to bear the brunt of all that. And in these contracts -- and by the way, you notice people doing that, they take these contracts pretty serious when they're taking accruals for it. And so as a customer comes in, say, they say, look -- and we haven't seen a lot of this. Hey, I'm a little bit soft for a couple of quarters. How can you help us? We have a rule. We call it do no harm. If we have someone else who could take that capacity and the economics are the same for us, do no harm. The partnership agreement with them wasn't, hey, let's sell 1 wafer twice, right? It was about making sure we had certainty and our customers had certainty. So the first is do no harm, and that's mostly what we have done. We found ways. The way you do that is what you're talking about is how fungible it's easy when you can just apply to another customer in that same node. But what if it's -- you need more 40-nanometer instead of 55-nanometer. The good news we have is we have pretty good fungibility, but it's not 100% sometimes you may have to use 1.2 wafers to create 1 wafer. And then they do no harm, then we'd just have to be kept economically whole on that. And so there's a myriad of different scenarios that play out on what's the [indiscernible] and how we do this. The overarching is do no harm. If we could sell it somewhere else, we do that. If we have to fund it, and we don't get the same economics, we can reduce the kind of the damage of the cost by that, just what is it to keep us neutral economically. What would you add to that, David?

I think maybe just a couple of details building on top of that. So number one, we built our manufacturing to have fungibility, to have flexibility, not only within a site, but also across sites. So if you look at Dresden versus Singapore, we have overlap on technology platforms that are dual-qualified. And so you can actually shift production from one facility to the other, so not only within a facility within a corridor, but also across sites. And then at a high level, kind of big, rough and tough number, you're talking about flexibility of probably like 25%. So about 1/4 can be kind of flexible between a customer that wants to wiggle a little bit with their demand and maybe replace it with another customer in an end market. So we've got some inherent design by design built-in flexibility. And then the second thing that I would say is that when you think about some of the movements that you've seen in the marketplace, maybe some that have been publicly announced, you're talking about movements that are less than 10% of an LTA. You're not talking about a -- the entire LTA, you're talking about customers that are trying to move a small little portion of that LTA. And when you're moving a small portion of and your oversubscribed, but we're going to show you for '22 and '23, and you have built in flexibility, you have tremendous opportunity to actually remix your business to a higher more accretive margin to deliver better than planned, and that's exactly what's happening.

And look, the positive of this for those customers, they're becoming a lot better at planning their business, right? In the past, they didn't have to worry about it. In fact, forecast was whatever. Now they're becoming a lot more educated on their forecast, they're going to become better companies because of this because they want to make sure that they're...

And just remind me, again, if...

Use the microphone, please.

Sorry. Just remind me again, if the company were to break the supply agreement, there's a charge, right, that they have to pay you. And are you -- do you have that in a separate bucket in terms of how you're recognizing that revenue, if there are charges that...

Yes. So let me give you some more details. Look, the contracts are fixed price, fixed volume, take-or-pay agreements, and nearly all of them -- all the ones I can certainly think of actually come with advanced funding. So that's either access fees or prepayments. Those will sit on our balance sheet. They'll be ratably recognized based upon either just the passage of time and unit or based upon some milestone. We don't currently have any material agreement. In fact, I can't think of an agreement off the top of my head that we actually have a customer in breach. So we haven't actually recognized anything through our P&L associated with any of the changes that you're talking about. So I wanted to be clear on that point. Now to the extent that we would renegotiate an LTA in a big way or that there would be a breach and it was material, we would spike that out for you. We'd be very transparent in our financials. But currently sitting on the balance sheet.

Sorry, your last question before the break. We'll take it from Vivek.

Vivek Arya from Bank of America. I had 2 questions. First is the dependence on the mobile industry, right, I mean it's obviously -- it's the most pervasive electronics device, right, a lot of benefits and built in kind of upgrade potential content, I get all that. But when I look at your sales, it's almost half exposed to mobile devices. If I just do a simple ballpark of your LTSA, right, 15.6 billion. So that's over 60% of your long-term supply agreements. And when I look at the contract you signed with Qualcomm, they will be over 20% of those LTSAs, right? And they were 15% of sales last year. So I just kind of want to get your perspective that this dependence on the mobile industry. What are kind of the pros and cons of that? And does it expose you to more kind of seasonal variations, customer concentration risks, et cetera?

Yes. So I'll make a couple of points and then, David, I'm sure will make even smarter points. So first, when you look at our revenue as a percentage of the buy of foundry on handsets, we're not disproportionate on that. You can't be a big player if you're not in that market. I think for us, it's -- as we grow these other end markets and they grow, you'll see the percentage of revenue in that smart mobile device will come down even as we grow. And the key is not taking down what we do in smart mobile devices, but growing those other markets. Now later in the presentation, you'll see some of that remixing of our business. For us, if there's differentiation in a smartphone that needs software, we think that's still a strong market to be in. And as long as we're winning new sockets, we're not leery of being in that business. David, what would you put in that?

Yes. Let me just throw out some more details. We talked a little bit about this in the call yesterday, the earnings call yesterday. We grew our smart mobile business 14% year-over-year. But if you actually break that down and you deconstruct it, what you'll see underneath is that front-end module was relatively flat. In fact, we expect the front-end module -- portion of that, which is about half of the smart mobile business, we expect that to be pretty flat sequentially Q-to-Q to Q-to-Q. And so what that means is -- and how is that happening? What that means is that we're growing in the 5G segment, which is accretive to us. So 5G is still growing year-over-year. In fact, 5G year-over-year is growing to the tune of more than 30%. I think the last number I looked at was around 35%. And so that growth in the 5G portion of the total market, that growth, which is heavy GF attach, heavy GF accretion, that's offsetting what you're seeing in the 4G, 3G and other markets. And so that's keeping the front-end module flat. So then you say, well, how do you grow 14% year-over-year? Well, that's in everything else around it. We're winning share. So this goes back to Tom's comment earlier, which is we're winning share in all the other sockets around it. And so as we win that share, we can produce growth. And so RF front-end module relatively flat, the portions around it, gaining share, which means that we can actually still grow year-over-year, which is our expectation throughout the entire second half of this year.

And then the last point, you talked about Qualcomm, you said it's all smart mobile device, it's not. Our long-term agreement with them covers much broader market segments than just the smart handsets.

Makes sense. And just a quick follow-up. The U.S. and the EU CHIPS Acts, what do they mean for GlobalFoundries? And I think I sort of get the positive aspects of them, and they are very important. I wanted to understand the other side of it. Does it -- can it create more competition for you? If Intel starts to get a lot more money, they start to establish foundries, which they then have to fill up, they are trying to buy tower semi. So do you think that the competitive landscape outside of the China-Taiwan region will be different in the next 3 to 5 years than what it is today?

Three to 5 years, no. I mean, I can't say it's going to happen a decade from now. We spent 10 years to create a global footprint. You heard it here first, you're going to see how difficult it is to create in a new part of the world, everything you've taken forever to build somewhere else. You'll see it in Arizona, you'll see it elsewhere. The tower acquisition by Intel is still far from very little intersection point from where we'll kind of bump into each other and our customers' lobbies. By far, Intel's thrust is going to be in single-digit nanometer. That's what they're investing in Ohio. That's what they're investing in Germany, I think they can do. What I really think is going to happen is that the same capacity we could be put on, but we'll be putting on in a different distribution in the world. These co-investments are just really to create more supply chain stability. And our play in France was, again, economies of scale, capital efficiency, build at a site. Our Dresden location, we may someday announce, we'll add to that with the European Chip Act, but we'll only do it if customers need it, it's on our differentiated technology, and we believe there's long-term viability to that. So I think for us, it just creates the opportunity to be able to invest in our global footprint in a more healthy way, but never ahead of ourselves.

Thank you, Vivek. We're going to take a break guys now. We'll come back exactly in 15 minutes, about 10:55. [Break]

All right. Well, hopefully, everybody has a seat here. We're ready to get started here. Well, let's start the second session here. We're going to have -- start off with -- we have Chief Commercial Officer here, Juan Cordovez. He's going to give you a very good insight into all of our customer activities that he and his team run and we'll go from there. Juan?

Excellent. So thank you again for joining us today, and welcome to the second half of our first Capital Markets Day. It's my pleasure to address you today. As Sukhi mentioned, my name is Juan Cordovez. I'm the Chief Commercial Officer for GlobalFoundries, and honored to be here. So to launch the second half of the day, I want to go ahead and anchor on our customers, how we create these customer relationships to evolve into partnerships.

[indiscernible].

Not weird at all. At all. Can you just put a little music until we wait. Keep on going buddy? Or hold?

Keep going.

Keep going? You guys feel good? All right. Sounds good. So we'll obviously be paying attention to this presentation as well to that speaker phone. Excellent. All right. As I described earlier, we will launch by anchoring on our customers. We design our go-to-market strategy, the journey our customers go through, by really understanding our customers' unique needs. And because these needs are unique, we have to first think and act differently to serve these unique needs. So with that said, allow me to give you a brief view of how we act and think differently and serve our customers. As you see from the left, one of the core design points of our go-to-market, as Tom mentioned earlier, is the fact that we serve our customers locally across all the end markets and across the all fabs. That means that they have this continuous interface, no matter where they're an automotive player, an FDX designer, they know that they'll have a continuous end-to-end integrated journey with GlobalFoundries. And how do you get that going in a way that's fast, agile and insightful? We call this the concept of ownership at the edge. What does that mean? Let me do that through example, if you need 500 extra wafers for automotive at GlobalFoundries, you do not need to fly 8,000 miles to Hsinchu and ask for that favor. If you're working a commercial deal and you want to understand whether we can arrange a specific agreement, you don't have to set up your alarm clock to wait until Seoul wakes up and gives you a finding. By creating that empowerment at the edge, with the right commercial resources and the right technical resources, you can really create that higher degree of intimacy, that higher degree of connectivity, and then we leverage that to make customers become partners. This concept of integrated delivery, both commercial and technical at the edge is, I think, one of the differentiators, which has allowed us to develop those partnerships into the results you saw earlier. Now another element to consider is the fact that who buys semiconductors and the value chain overall has evolved. If anything we've learned over the last 2 years is that people have to connect in a more deep way in order to build value together. No longer can you have the silo environments where you just have a supplier-customer relationship. Not only will that expose you to chance, it also doesn't allow you to really fulfill the opportunity that exists by partnering as one. And we'll talk a little bit more about this concept of purpose-built engagement models coming up. This approach through the value chain creates our go-to-market results, which we will detail in proof points coming up. But before we move on, I want to direct you to the right-hand side of our chart. At the end of the day, we are the partnerships we develop. The idea that our work ultimately comes about from these customers trusting us with the future of their company and designing sole source and single-source products with GlobalFoundries. We take that responsibility very, very -- not lightly, but with tremendous honor and respect. Now let's talk a little bit more about our customers and where we serve them. As described earlier by Tom, a big part of what we do and a big part of our value is our global footprint. That global footprint is enabled and strengthened by a globally diverse set of customers. And what you see in these charts is how we serve them across the different end markets that we focus on. And what you see here is that we have a broad deployment of field sites, again, with ownership at the edge that serve these customers. So that when somebody says, "Hey, let's solve this design closure issue, design timing issue, DRC issue," instead of launching Zoom, you can meet at their desk, at their site, whiteboard it and solve it. The speed and intimacy that you can derive from that type of interaction is a differentiator. A second element that's also highly important is the idea that these local regions and ecosystems have their unique strengths. So you want to exploit those strengths and you want to build your go-to-market focus exploit those strengths and bring value. You'll see that in Europe, as we know. A big part of automotive, IoT is centered as an excellence center in Europe. You'll see that across automotive and IoT, 70% of our business is in Europe. We all know that the U.S. is a smart and mobile device center of excellence, the flagship of the world. You see that similarly, we have commensurate market share in that market. And in Asia, you see a really great story of diversity. About half of our business, be it smart mobile, the other half being IoT, community infrastructure and data center. That balance gives us strength. And by exporting and investing on those local ecosystems, we can develop the intimacy that will make us much more able to translate the market need, the customer need, the technology requirements that then through Gregg's organization and feature-rich solutions ultimately creates the value that GF builds. Now one of the key areas that I want to describe, and you'll see this in the bottom row of the slide, is this idea of ownership at the edge and intimacy at the edge. We serve our customers in their local language. As a person born in South America, who came to the U.S. at 19 and had to, like, look at an SAT in English for the first time, I can tell you that having discussions in a foreign language about technical topics is tough. I translate Spanish to English, a counterpart translates English to Japanese, and then act. It's very, very difficult. If you create the right technical expertise locally, you can drive power. And how do we do that? Something I'm quite proud of. If you look at the median distance between our field site and expertise to our customers, sub 7 miles for our top 12 customers. That's a big deal. We serve automotive out of Munich, not Amsterdam. Israel out of Israel, not Amsterdam, as examples. So now you're likely asking what exactly does this mean in terms of the motions that you carry across the sales horizon. Let me describe that in a little bit. What you see here is our mission and our path forward. One of the reasons my job, my team's job is really fun is because you get to work with customers across a variety of sales horizons. From that really critical and important PO chase, fulfill the fabs with the highest quality demand that's ultimately super tactical, to working with the customer, their end OEM integrator and creating new value through new technology partnerships through new capacity enabled with customer funding with, in cases, governmental support, big ticket items that drive a lot of our business that require innovation and creativity, our team manages all of that. Now how do you do that in a way.

[indiscernible].

Okay. Good to know either way. All right. So kind of going back to the topic at hand. The idea is that as you drive these different types of sales delivery motions, how do you ensure that you have the precision of action, the precision of outcome so that you can drive with strength and deliver the right rigor that is needed to ensure that our current quarter financials are delivered through the right demand capture, the next-generation business is captured and so on and so forth. So we're going to find these sales motions that you see on the right-hand side, right? On the near-term horizon, laser focus on creating a superior demand capture that includes fulfilling our LTAs. And as we drive in a currently oversubscribed environment, how do you select that right business, the right end markets, those right partnerships, to offer your precious capacity to? Remixing your capacity, so you create more value for your customers for the industry and for ourselves. As you track to that next horizon, how do you ensure that you're being [indiscernible] accretive business into your factories, right? Selecting those high-growth secular, strong end markets that give us a diversity that makes us stronger, alluding to the question earlier. And then finally, how do you do all of that with the right level of demand certainty where your customers are betting their future on you, you are betting your future on them by putting the right investments in place, the right capacity access, and doing so in a way that not only binds you in a contractual way, but that forces you to develop that next layer of intimacy to what is the end market, to what are the different customers and end sockets, OEMs. And through that dialogue, you get a level of precision of what your partnership has to deliver, but really, really strengthens both your technical and commercial collaboration. Now you do that well enough, not only will you have full fabs, but you'll be looking at how do you build new capacity, again, in partnership with our customers. So now that we've seen the selling motion, let's look at hard data to put all of this into context and show its value. As both Tom and Ed alluded to, we drive our business with the concept of creating sustainability through profitability, certainty and durability. These are the 3 tenants that, in our view, if our customers and ourselves join in these parameters, we can create win-win outcomes. What you see here, first on profitability, is that we have been able to renew a large set of LTAs with our customers. And we're doing so at about 20% on average price improvement over the LTAs that we're happy to show you last year during the IPO roadshow. And beyond that, the broader design with motion is delivering over 5 points of improvement of gross margin above our long-term model. Moving on to certainty. As alluded before, 75% of 2024 -- 2023 oversubscribed, 2024 at 75%. And one of the beauties of this is that our customers are eager to load us and bring more capacity for these markets that are enjoying disruptive secular growth. They are putting their balance sheets and their money worth their mouths by supporting through contribution into this capacity build. $1 billion since the IPO roadshow, almost $4 billion overall. And then finally, let's talk a little bit about the concept of durability, remixing our business so that we serve those right markets. You'll see almost $4 billion of auto and IoT design wins and they're effectively all sole sourced -- or sourced. This is an example of delivering a well-balanced go-to-market strategy in partnership with our customers that gives them, ourselves and, frankly, our investors, strength, visibility and resilience. Now with that in mind, a lot of you have asked questions about LTAs. And I'd like to give you a little bit more of a view. Now many of you have asked, "Hey, I've seen a lot of LTAs done last year. Is this a one-off? Or is this part of a new approach that we're going to be using to create a more repeatable and more predictable way of doing business?" So let's have the data that the talking. What you see here is that we see tremendous, continued momentum on LTAs. You'll see on the far left where we announced during IPO roadshow about $20 billion of committed demand. And the net is the following: Without a doubt, we're seeing tremendous customer appetite for more LTAs as they see them as valuable as we do. They see this as the mechanism by which we can grow our partnership and they can grow as a business. These LTAs and the certainty of supply they offer, coupled with the right technology, are the currency for growth. And what you see in the chart is that the rate at which we're extending the LTA size is higher than the shipped revenue, meaning that the total amount of net demand that is committed to GlobalFoundries is increasing. We're very, very proud of this. And again, I want to emphasize that not only do you have this kind of binding partnerships as linked through an LTA, how it pushes and forces you to collaborate more deeply across supply chain, technology interlock, market dynamics are the real value creator that not only gives you that certainty of supply that I think it gives you the intimacy to collaborate and adjust. I think that is a very critical element. Now as we looked at some of these selling motions, what I want to do is take a pause and anchor back into the customer. We talked about how different customers are coming into light as the value chain of technology delivery evolves. How are we evolving with it? How are we supporting acceleration of that evolution? This takes us to the concept of different agents in the semiconductor value chain exists. How do we create purpose-built models to ensure that they are successful, to ensure that they have the tools that they need and, more importantly, to create opportunity for us, for them and for our end customers. What you see here is the 3 key traditional customer types, fabless, IDM and OEM. But hey, they are evolving. And I want to describe in some detail what GF is doing to accelerate our value capture and value creation. Let's start with fabless, right? The fabless were very, very interesting. And what you see now is that based on the GlobalFoundries' feature-rich technology strategy, we're creating opportunity for fabless players which traditionally cannot compete for feature-rich solutions with IDMs because of technology. We have unique technology that allows them to go head-to-head and open up sockets and open up share gate that was traditionally not possible. Now for that technology to really create disruption and enablement, you have to provide the right design enablement and design support for these companies who don't have 2,000, 3,000 people waiting to do design automation, to consume that technology, to unlock its value and to fulfill their design ideas into [indiscernible]. The concept of ownership and support of the edge helps us. And our rich history of design enablement expertise comes to bear to provide that acceleration. Now technology, design, all this stuff without access to the currency of capacity and capacity certainly, nothing happens. So it's not a 1,2 punch, it's a 1,2,3 punch that I believe, again, makes GlobalFoundries unique and the feature-rich design point of our go-to-market ultimately unlocks value creation and open it up sockets that traditionally fabless cannot access. Now let's talk about IDMs. Tom talked about the fact that the traditional IDM model where 100% of what they built was done inside is dead. IDMs are dead but long-live IDM. The hybrid model has been adopted. Now this is another environment where we're seeing evolution. IDMs have been sourcing from foundry, nanoscale CMOS or, call it, scale CMOS for a long time. This is not a new thing. But what's happening now is rather unique. I think what they're seeing now is at some of us, like GlobalFoundries, have created sufficient technology differentiation where for the first time, they're saying, "Hey, I actually am going to use that specialty technology as opposed to building in-house, as opposed to doing feature-rich content." They are seeing the right differentiated technology as an ability to complement their road maps and not have to do it all in doors. That is new. And more important than that, what they're seeing is that if you can then partner with a company that has a global manufacturing footprint, you can then have both that complements your road map, that access to scale, but you can have access to scale where it counts. You can use as a global footprint of GlobalFoundries as a competitive advantage. You think that all these European companies are building automotive are going to depend on something that's being built 4 miles away from military fire drills? Probably not. They're going to be using global footprint from companies that are established that know how to multisource our technologies within our global footprint, to provide that resilience, that predictability and that strength. And because GlobalFoundries is uniquely positioned across these 3 geos, we can do so for every IDM in the world, and we're doing so, like that. Now I'd like to talk to you a little bit about OEM. OEM really, really fun. Similarly to IDM, they've been doing scaled CMOS and nanoscale forever with foundry. But what's happening now, and they learned this last year, is that the feature rich content is the one that's enabling AR, VR, display sensors, and that technology is not something they've traditionally been focused on through their own design, and they know they have to get into that. They recognized that you're going to have a far higher degree of technology alignment with these feature-rich solutions to allow them to architect their end user experience in a much stronger way, to unlock the value of that. And when you do that, you also similarly will [indiscernible] of having the access to the right geos and to having the right access to the end-to-end supply visibility and resilience that is required so that 2021, '22 does not happen again. Now with that in mind, let's talk of some real proof points to give you a sense for what we're up to. I have 3 here that I'll show you. One for each of the customer segmentations models. I'll talk about Qualcomm a little bit. I know we've talked about it a lot, but as a person that kind of sort of was involved in that deal, I'm very proud of it. And I want to talk to you a little bit about it. So it's very simple. Qualcomm is a world leader in 5G. They are doing really great work to diversify and you see what they're doing with connectivity, both in the car as well as in IoT. And they dance with a lot of people. The size and scale that they have, as Tom alluded to, requires that they ensure they have the right partnerships for people that have stability, people who can execute and the people who are willing to partner with them as they drive their story. The GlobalFoundries value proposition and the Qualcomm needs are an ideal fit. Our commercial strategy and their business priorities are an ideal fit. And as a result of that, we were able to make this agreement, which includes them supporting our expansion of our U.S. multi fab. I'm very, very proud of that. Now let's move into IDM. I love this example because it gives you a sense for the other selling motions in play and specifically to the question asked earlier, how do we ensure that we have the right diversification of customers. You'll see $0.5 billion reservation. And the more important element is that it's 7 years long. It tackles market segments that are connected by one thing. They require security, embedded processing and geo footprint. And that MCU technology that this company has awarded to us as a single source, allows for work in data center with embedded controllers, in the data center around secure boot, secure boot of trust, stuff that you're not putting in a conflict zone is built by GF. And the extension of that MCU product line because that connects MCU creates tremendous opportunity in IoT. This is almost 36 products, 3 dozen products, that have been taped out to the company. And to give you a sense for that broad and pervasive deployment of semiconductors and our participation and enabling in traditional IDM to work with a foundry for the right technology, for the right diverse product group, serving the right end customers. Finally, I want to talk a little bit about OEM. This one is a really fun one. Last year, we saw a lot of supplies issues. And we had a very clear choice. Do you get super tactical in filling gaps? Or do you get strategic and fill the gaps that talk about a broader, deeper, more insightful collaboration? Do you use the crisis as an opportunity to develop intimacy? And I believe we did that. With this very, very large OEM who prefers to not be named, we worked the deal from a tactical basis, increases units incremental to their baseline of 300 million, which is a lot of millions. But the more interesting piece here is what we did on the strategic side. We recognized that, that company was awarding designs that were critical and single source on a yearly basis. But ultimately, it's linked together with GlobalFoundries and our differentiated solutions. Instead of that process that often talked about gaps of understanding, gaps in capacity, gaps in alignment we created a partnership deal with them so that our capacity for these differentiation solutions is available to them. And then in turn, they partition that capacity to their suppliers in a way that completely takes away that completely obfuscated communication chain between the selection of the foundry, design of product and the award at the OEMs socket. The level of transparency to the product, to the socket, to the timing and its allocation has now been completely closed. We're very, very proud of that. And they have that access for multiple technologies across multiple geos. So it's not a onetime deal, but really a platform to scale. And in my view, this is a new engagement that will be a proof point of how the world and the new value chain will scale into the future. I'm incredibly proud of it. Now to give you a bit more perspective, allow me to give you an example of a specific customer where we go a little more into our commercial strategy, our go-to-market and see how we've realized the partnership. But before I do that, I want to give you some proof points on the specific elements we just discussed around customer concentration, customer type and end market. This will give you a sense of a report card of how we're doing. You'll see that in customer concentration, from 2020, '22, '25, 20% improvement meaning that we're less reliant on the top customers. You'll see on customer type that similarly, we have a far improved balance with a high growth of OEM, albeit starting from a relatively modest side. And then finally, you'll see on the market size that we are no longer as reliant on markets that are saturated. You'll also see that while we did land the 5G transition, huge growth in automotive, huge growth in a home industrial IoT of about 50% and a doubling of our communications and data center business. Now if you look at a specific example of a customer, you'll see the type of intimacy we've been able to develop with such a partnership. The shared investment of capacity expansion, which is the new economic model Tom talked about, you see tremendous skin in the game from the customer, both in investment and in demand certainty. You'll see that now we're working with OEMs and this customer to ensure that, knowing that we wouldn't win the OEM together. They probably wouldn't win the OEM together. When you work as one, you succeed. It drives that to the diversity of end markets that we've discussed and through geo footprint. As you see on the right, 8x growth in auto from 2020 to 2025 forecast, a huge growth in industrial and all of that done from what started as a single geo relationship now to a 3 geo relationship. This is a hallmark of the way that we do business. Now you will ask, "That's one example. Tell me how you are doing?" And I'm very proud to show you how we are winning with the winners. You'll see that for these key end markets that we tackle, we are working with not the top 3, but the top player in their segments. You'll see that these relationships are also a single source. You'll see how we are the primary foundry for either the battleground or overall. And not only that, we ensure that we underwrite those partnerships with the right long-term agreements and, when it's needed, the right co-invention and technology co-optimization. This gives you a sense for partnership across the tactical elements like PO capture, all the way to the more valuable end market alignment, socket alignment, technology alignment. This is the strength that we can provide through the partnerships we've built. To leave you, I want to give you a few notes. Who are we? We are a team that's deployed in the field with ownership at the edge to drive that intimacy and clarity. We are a team that is driven by win-win outcomes through profitability, certainty and durability. And we are a team that believes that customers exist like partnerships are rare. It is our goal to deliver those partnerships across the value chain. This is who we are. Thank you very much.

It is my distinct pleasure to invite the world famous KC Ang to stage, our Chief Manufacturing Officer.

Thank you, Juan. He just announced me that -- yes, I'm the man that deliver quality wafers to our customer on time. Okay. Thank you, Juan. Yes. In this section, I really like to cover 5 main points with all of you. First, I will talk about the global manufacturing footprint, which includes the latest collaboration that Tom talked about in the morning with ST in France, our manufacturing performance, our cost reduction efforts and the productivity improvements and also our health and safety report. But also I'll show you what is the progress on Singapore side on new track construction with a short video. So with that, let me go to this point here. This is our current manufacturing footprint. And as you can see from the map that we have 5 sites strategically located in U.S., Europe and Asia. And let me start with the U.S. fab. So Malta is our most advanced fab running [indiscernible] fab, RF and our proprietary silicon photonic process. This fab is currently capable of 400,000 wafers per annum, and Burlington is one of our 200-millimeter fab, lots of innovation from this fab. Major processes of high-performance, CG and RF processes. We have made significant improvement in productivity and yield in this fab. And currently, the capacity of Burlington is capable of 250,000 wafers, 300-millimeter equivalent. Next is our East Fishkill fab, which has 135,000 wafers capacity with major production in our HSW and high-performance CMOS. This fab, as mentioned by Tom earlier, that is in the transition to become on semi by end of 2022. So all the 3 U.S. fab across the 5 trusted foundries where we have business relationship with U.S. government, they are long-term loading commitment. So let me talk about Singapore site. Singapore, which is our largest site of production today, Singapore has both 300-millimeter and 200-millimeter with a combined 1.1 million wafers per year output. And Singapore site have many feature-rich technology in the area of RF, BCD, non-volatile memory, low power to support a wide range of customers and products that Juan talked about this morning in the market segment of automotive industry, IoT, medical, mobile and PC. Finally, our Dresden fab, which is our fastest-growing fab. Since last year, we have been ramping from 500,000 wafers per year to 680,000 this year. And Dresden is where our FDSOI processes was first developed. And now the fab also running 28-nanometer technology and 14-nanometer as well as 55 BCDL in conjunction with the Fab 7 in Singapore site. So let me conclude this point by saying that we executed our manufacturing strategy very well. In the last few years, our modular capacity expansion is also working very well. As most of our business, you heard this is all about single source, lower sourcing from different region of GF fabs is extremely important for our customers' business continuity planning. So our customer can now get the same product, same quality, same [indiscernible] from Singapore and Dresden fab, especially automotive customer that they love to get from Singapore and Dresden for business continuity. So we can now provide Singapore-Dresden, Singapore-Burlington, Singapore-East Fishkill for the dual sourcing capacity for our customer. And we also created a global center of excellence where we can benefit the cross-sharing and global resources to improve our excellence in all areas. So let me move to how our global manufacturing footprint looks like beyond 2023. As you can see in this page, we added growth and then removed East Fishkill since they become on semiconductor starting next year. But despite its on semiconductor, we continue to get 120,000 wafers from them in RF technology and high-performance CG. Malta and Burlington will grow the capacity, respectively, and Malta will still focus on job impact with more feature-rich technology in embedded memory, RF SOI and significant volume of silicon photonic by then. Burlington fab, as Gregg pointed out earlier, this fab were focusing on the wide band gap technology development for gallium nitride technology. Our Singapore site will grow with the additional capacity coming out from our new fab, and this will bring Singapore site to more than 1.5 million of 300-millimeter equivalent. And Singapore will continue to provide wide range of product and customer with a feature-rich technology. Dresden will grow to 850,000 wafers per year, with more 2 FDX, 28-nanometer, at the same time doing the 40 [indiscernible] and 55-nanometer product. So we now added growth, as you can see on the chart there. And this slide will focus on 22-nanometer with a full build-out capacity of 360,000 wafers per year. I will further explain growth back in detail in my next page. So we will grow our total capacity to more than 50% through 2025, and we will continue to grow our global footprint by utilizing global talents, dual sourcing offering to our customer for the business continuity planning. So besides Singapore-Dresden, Singapore-Burlington, Singapore on semi, we will expand Dresden growth on 22FDX, Dresden-Malta, and in future Singapore-Malta. So let me move to Crolles fab. Now as you can see from this point about the layout of the Crolles fab. But here, I want to repeat what Tom say this morning. Two points are here. So first, we added our capacity in the capital addition way. We have grants from the French government and customer prepayment. And with this additional capacity plus the expansion of Dresden, we will be growing triple capacity from 2020 to 2028. So let me explain this collaboration model from the operation and execution point of view. Here, you can see the diagram of our existing ST facility in growth in the red shaded area, which is readily [indiscernible], we are going to bring in 20 unit too to form a pilot line for us to do technology transfer of 22FDX starting next year. So then when the blue area, which is the [indiscernible] by module expand, we will build capacity and start production towards the end of 2024 and then ramp in 2025 to the maximum capacity of 360,000 wafers per year. So we are leveraging the existing facility and adopting the modular capacity expansion approach. The facility will be co-managed by ST. And it's actually effectively to me, is the extension of our Dresden fab, where GF is responsible for 22FDX production ramp and new improvement. Therefore, the on-site management team will be supplemented by the Dresden expertise. On the other big component is the cost saving with ST on depreciation fixed and variable costs, including the power usage. So let me move to the capacity expansion road map and this kind of summarize what I said earlier. Let me explain the color. The purple color is our existing capacity. The blue color is the capacity expansion within the 4 walls and the green is under construction. As you can see from the chart here, Singapore will grow to 1.2 million wafer, coupled with the 200-millimeter of [ 3 70 ] Singapore will have volume growth to -- volume over 1.5 million wafers per year. Dresden site will be 850,000 wafers per year. And Crolles as I mentioned earlier that we'll go to [ 3 60 ]. All this capacity do not include the future expansion plan, such as Malta fab expansion and further Dresden expansion which we mentioned earlier. I just want to emphasize the investment criteria here that we put together. It's customer commitment, government grants and copy-smart-technology approach that allow us to ramp the fab in a fast manner. So let me talk about the manufacturing KPI. Here, if you see from the chart here, we are actually doing very well in our manufacturing KPI. On the customer side, we continue to execute well in the new prototyping of product. Our first time right success of more than 95% helps the product to launch to the market on time. And our yield KPI is excellent at 99%. As Gregg talked about the automotive this morning, we provide automotive quality to our auto customer. And it's not easy because all our fabs are TS 16949 certified, and all the fabs are subject to VDA 6.3, a very stringent audit from our auto customer. And we also executed our dual sourcing strategy very well, where we have already end customer product qualified in Singapore-Dresden, Singapore-East Fishkill and Singapore-Burlington. On the business side, I have to say that I'm very proud that manufacturing did not miss a bit on production. Shipment to customer in 2021 was 2.4 million. And this year, we probably will get to 2.6 million and our on-time delivery is 99%. And we also drive very hard on maintaining our cost by achieving $250 million saving per year, which I will have a separate chart to show you on the next page. Now as you can see from this chart, the scale drive costs lower. So here, you can see that we increased our shipment by 28% from 2018 to 2021, while the cost of goods sold per wafer reduced by 35%. As we achieved the scale, we will continue to increase our capacity at the rate of 10% year-on-year in a capital-efficient way of expansion, and we'll continue our productivity improvement. On the cost side, we will continue to push for cost saving activities through the global pricing, global benchmarking and all the engineering projects. So now let me talk about the productivity. We think there is a big opportunity to improve our further capacity by improving the equipment performance. Equipment performance uptime to our 1G app across site as well as external benchmarking. So this productivity improvement across the process optimization we think we can easily get additional -- 5% to 8% additional capacity. For the input side, let me focus on 2 biggest spend in our factory. One is the equipment maintenance and another one is our indirect material such as gases and chemical. On the equipment maintenance, we have already moved from the time-based PM to the usage-based PM. And we'd like to progress and implement more and more predictive PM using big data, FDC and AI. The predictive maintenance will give us a better equipment uptime, better usage of equipment, consumables, spare parts and, hence, will lower our maintenance cost. On the material side, we will continue to source for alternate chemical, gases standardization, optimization the usage across all the sites. We also plan to do a lot more chemical on-site generation and branding chemical like ammonia hydroxide, auto developers to save our cost. More importantly, we are using big data and AI to optimize our usage. So with the productivity improvement and cost reduction, I'm very confident that we will continue to drive towards a competitive cost despite the headwind that we see in all areas. So let me go to the new fab in Singapore. If you look at the chart and you compare the actual building to the architect drawing on the bottom right corner, you can see that the admin office is still under construction. But more importantly, the 2 building, utility building and the fab, inside the [indiscernible] is pretty much done that -- right now, we have power, water, chemical gases are all delivered in there. Despite the COVID-19 pandemic, where many countries were locked down and the numerous supplies and logistic issue, we managed to move the first two into the [indiscernible] 1 year after our groundbreaking. I tell you, I built so many fabs. This is the -- one of the best I've ever seen that they have great safety and quality record, within the budget and, most importantly, meet the schedule. So now let me show you a video on the progress of this fab. So video, please. [Presentation]

So as of yesterday, I checked, we have 95 tools under installation and setup. So by end of this year, and hopefully, by next year, we should be able to run production. Now let me talk about our health and safety. As Tom mentioned earlier, our #1 metric for our company is safety. So if you look at the data, we have a very respectable TRIR, which is the total recordable injury rate, 0.13, which is the lowest in our history and way below our industry standard of 0.7. We are, I think, on track also to meet the target of 2022. And GF also won the award as one of the America's safest company. As you can see, production ramped up, we continue to maintain our safety. Now let me talk about COVID. To keep the production running, our GF safety committee has adopted a situation that we just dynamically on the health protocol, safety protocol, just to keep the production running. So with that, I appreciate very much on [indiscernible] guidance as well as the cooperation from all the employees. So let me summarize what I've mentioned to everybody earlier. First, our safety always, right? We maintain our world-class safety performance. All the fabs are running at maximum utilization with excellent manufacturing KPI. So we continue to stay cost competitive by rigorous cost reduction and productivity improvement, and we achieved our equipment move in into the Singapore new fab on 23rd of June, and we are now focusing on how do we start up the equipment and go into production starting January 2023. Last but not the least, we continue to expand our global manufacturing footprint in a capital-efficient manner to satisfy our customer long-term agreement. So with that, I thank you for your attention. And let me take the opportunity to introduce our Chief Legal Officer, Saam.