 Good afternoon. I am Yongjun Sun, James J. Saurberg, head and Rensberg professor of the School of Industrial Engineering. I would like to welcome you to Purdue Engineering Frontiers lecture series co-hosted by College of Engineering and School of Industrial Engineering. So now I'm going to invite Dr. Mark Lundstrom, everyone knows Dr. Lundstrom, who is the chief semiconductor officer of Purdue University, and Don Carroll Cipher's distinguished professor of ECE and a former internship dean of College of Engineering, whom I worked very closely several, last several months, and I want to invite Dr. Lundstrom to the podium so that he can introduce today's distinguished seminar speaker. Mark? So thank you, Sun, and good afternoon, everyone. You know, we're really fortunate today to have one of the nation's leaders in semiconductor manufacturing here to talk to us about perhaps the most important, certainly the most important immediate challenge that the nation faces. You know, I've worked on semiconductor technology development and manufacturing and research for my entire career, but most of my friends and my family had no idea what a semiconductor was or why chips were so important. All that has changed, and now we all understand how critical this technology is to our nation's defense and to our economic future. Companies like Intel will play a critical role in addressing this challenge, and I sleep a lot better knowing that Intel has leaders like our speaker, K-Bahn Esfajarni, a Purdue Engineering Industrial Engineering alumnus. So K-Bahn is executive vice president and chief global operations officer and general manager of manufacturing, supply chain, and operations at Intel Corporation. He's responsible for the company's worldwide manufacturing operations as well as for corporate quality insurance, corporate services, and supply chain. Before his current position, K-Bahn led Intel's non-volatile memory worldwide manufacturing. His tenure at Intel began in 1996 as an industrial engineer after working in the automobile industry for four years. Since that time, he's held a wide variety of leadership roles in process engineering, yield in manufacturing, and also in enabling technology startups. And K-Bahn earned his bachelor's and master's degrees right here at Purdue in industrial engineering. So it's a real treat to hear from someone on the front lines of what people are referring to as the chipboard. Please join me in welcoming K-Bahn Esfajarni. Hey folks, first of all, it's such an honor pleasure to be back at home, I should say. And you know, today's session, and thank you for that very nice, you know, I would say kind, you know, description. But to really describe what I do at Intel, I'm basically in charge of, you know, making sure that, hey, the company provides the leading edge, you know, the process technologies, the chips that you all sort of know as Intel inside, and you put that into your laptops, data centers. And with that comes all the responsibility of running the factories, running the supply chain, operations, construction, and all of this. However, and it's a little over 55,000, 60,000 people that basically are in my team that they do this. But today, I don't want to talk about Intel. Today I want to talk about this tremendous and amazing sort of journey ahead of us, which I really call out. And that's why, you know, put the title of the talk as we are in the trillion era. So what does that mean trillion era? And part of that, it is becoming so critical that, you know, if there's one thing that we learned during the COVID was how important it is that the technologies like semiconductors, that is so critical to everybody, to all of our lives is so critical to make sure that we have that balance of that supply chain all around the world. Intel has been into this for the last, you know, 50 plus years. In fact, the semiconductors were pioneered, you know, here in the United States. So I want to give you a little bit of perspective of where is the state of the industry? Where is the state of the technology? And what we are doing about it? And what really the key priorities are. So I'm going to try to spend less time giving you foils and presentation. But the part that I love about these kind of engagements is questions. You know, your sort of what's in your mind, so that I will do my best to try to answer them. So let me just start with, you know, just a little bit, you know, the background I think Dr. Lance from mentions, I did my, you know, formal education here, do some of the best times and the years that I had was was here in Purdue. And I had a little bit of a stint about two, three years. I was working actually in a town very close by here, Logan Sports, Indiana, until Intel called me. And I was actually doing my PhD here. And at the time, you know, Intel said, Okay, why don't you come here for a couple of years. And then you can decide whether you want to go back. I said, No, no, no, I am going back. Okay, just just I'm telling you. And it's 27 years now. And my roles has been mostly in technology development, manufacturing, and the way Intel operates is you develop the process technology, right? And then you transfer that product process technology, and you ramp it all around the world. Okay, that that's sort of the business model. And you keep doing that every two years, right? So I'll give you a little bit of a context of that. But let me sort of start with why should we care, right? Now, many of you are too young to know that, Hey, it's not even too long ago that, you know, those those things we call them, I wish I could walk around. Can you can you hear me? Can you hear me? It's a small area so everybody can't hear. You guys can hear. Okay, good. You know, these, these these things called typewriter when I was applying to to colleges, you know, I had to type my my paper. Now, most of you know, there's a lot of young folks in the audience. You don't even have that concept. You just you just type it up and you know, in your PC and you just print it, right? You know, digital cameras, right? You used to put them in Kodachrome films, you take them to the, you know, a store and they basically develop the films for you, right? I mean, every every picture you were taking, you had to worry about how much is going to cost me to develop that thing, right? You know, we used to go, you know, to video stores and, and, you know, drive there, go select a cassette, bring it home, watch the movie, then take it back, right? And I'm sure you all have many, many examples of how technology is disrupting the old business models, right? No, it's just unthinkable. You go to some other location and get a movie, bring it home. And by the time you get there, the movie cannot be around. Look what's happening with the progression of moving away from fossil fuel cars to the electronic, you know, you know, vehicles, EVs, right? So the foundation of all this is the world is becoming more digital, right? That is what's happening. If you don't believe that, then, then, you know, look around you, right? How many of you use Uber or Lyft, right? Right? Yes, everybody. What happened to those taxis? Again, their business model got disrupted. And this is going to keep happening over and over and over again, right? So the thing to realize that because the world is going digital, what's foundation in all of that? It's semiconductors, right? Semiconductors is what powers the world of digital. It's those tiny little chips, right? And those tiny little chips, right? Is something that requires technology and innovation in order to keep providing these enabling capabilities. So I want to talk to you a little bit about, you know, and provide you the context of that, right? And at least it is our view, like Intel's view, is we are going to this trillion era. And digitization, it is moving very, very fast. How big is the, you know, the overall total available market for semiconductors today? But I don't expect you all to know. But if you do, that's great. You know, all these chips that you make, how much, you know, if I add them all up, what's the total revenue for the whole industry? It's about, yeah, 600 billion. See, Dr. Lanzron, by the way, he's one of the leaders, you know, to help us and partner with us to move this. Now in 2030 or so, that's going to be a trillion, right? So it took us 50 plus years to get to 500, 600 billion. And it's going to take us in another decade to pretty much close to double that. So you see that the move to this trend is unstoppable, right? And what is really driving that? And this is what we call the superpowers, right? Of course, compute, right? All the applications, right? All those examples I mentioned, you know, digital cameras, you know, cloud and, you know, using those e-commerce and Uber's, they need compute, right? The other one is just locally having the compute is one thing. Connectivity is critically important, right? You know, cars talking to each other, you know, systems talking to each other, right? IoT devices talking to each other, right? And then also really moving the data center all the way to the edge where you're actually taking those applications all the way from cloud to the actual end use cases, right? And then sensors are going to be everywhere, right? And whether it's in the manufacturing lines, whether it's in your, you know, cars, whether it's in the refrigerator, whether it's in your, you know, heart monitoring machine, its sensors is explosion of growth. The problem today is, yes, sensors are available. And there is enormous amount of data that gets generated. But guess what? We don't even use 0.2% of that data that gets generated. There's globs of data. This is where AI, this is where some of those, you know, predictive analytical approach that comes in that uses that data to really make use of them, filtering noise from signal, right? And then lastly, you know, the explosion, the whole industry, the world is going to change around AI. What do they all need as you're moving into this sort of digital world? They need semiconductors. Thank you. He's listening. Thank you. So we are having an absolutely critical juncture. So and this industry has been thriving on a simple concept, which was invented by Gordon Moore. Have you heard about Moore's law? Yes, most of you, right? So Gordon Moore, who was one of the founders of Intel, he predicted that in every two years, the number of transistors on a chip is going to double, right? And most of the engineering team right here, that's a, that's a logarithmic scale, a straight line of logarithmic scale. And this is these dots on this picture here shows the actual number of transistors basically per package. And package could be a chip, right? Basically a chip or it could be, you know, as we move forward, we're going to have number of packages come together. But why is this important? Is it just, hey, it's just an ego, let's just do fun science projects. Why do we care about the number of transistors on a chip? Keep increasing. Why do we care about what's called transistor density? Is it just for the fun of it? You get speed, right? It's the performance. What else? Cost, economics, right? It's like, think about a very expensive real estate that that chip that silicon is very, very expensive. The more transistors you can put into it, right? What happens? Your dollar per transistor keeps going down pretty rapidly. What else? I think you mentioned power with with with more transistor density, you need less power. You can use those those that you care about your battery life on your phone. Yes, you do, right? And of course, your aerial density. So this is not just a, hey, let's, it's a nice to have. It's economics, right? And 30 years ago, there were, let's say, 20, 30 semiconductor companies, right? Today, it's much more consolidated. Why? For those that are in the leading edge, right, which I said the Moore's law, which you have to keep shrinking the transistors, right? If you're not on this sort of curve, by the way, getting to those transistor densities means you need to do technology development, you need to do innovation, right? What if you get it wrong? But if you don't get to in two years, you're going to be doubling the number of transistors. What happens? All those things we just talked about, you won't have, you don't have power, you don't have performance, you don't have aerial density, you don't have economic goodness. Then what happens? You get what you deserve, you're out. You're out of the business. Done. Today, who are the companies that are in the leading edge semiconductor design, development, manufacturing? Intel? Very good. What else? TSMC, we're manufacturing. There's a lot of people that do Fabless. So AMD is Fabless, they don't manufacture. Micron is making memory. We're talking about processors, microprocessors, right? Apple doesn't produce the chips, they have somebody else produce them. And video is Fabless. You see, these are all great suggestions. But remember, you hear all these names, but none of them produce these chips themselves. They have somebody else make it for them. This is an important concept for this industry as we're moving forward. They're really, yes, they're three companies, right? Samsung, Korea, TSMC, Taiwan, and Intel, United States. Three companies in leading edge of the technology, developing what I just said, right? Performance, power, aerial density. And what are those applications for? For artificial intelligence, for high performance computing, for taking globs of data and processing them. For all these things, those are the three companies that do that. Now, let's take a step back. Is this thing important? Absolutely. Now, what if one company controls the world on this whole area? Not a good thing. What if this whole thing is moved away from, let's say, one region of the world to another region of the world? Not a good thing. So that's why you hear about this. And one of the things I also want to highlight is, when I said you get what you deserve, is this is a race, right? Because, you know, Intel, we have had our own issues, right? We have had a situation where we did not stay on the, you know, we sort of tried. It's all about being at the leading edge of pushing limits of science and technology. We're talking about atomic level sort of transistors, right? The equipments you use, the material you use, right? If you don't get it right, if you're late, what happens? The other guy gets ahead of you. And now, we had this issue four or five years ago. Now, you're coming back out of it, right? So if you look at, if you hear, when you hear from companies like Intel or others, they are basically going to be on the different nodes, right? For Intel, Intel 7, you see a bunch of products that we make on the PCs as well as data center, Intel 4, Intel 3, 20, 18, this is progression of moving to these new nodes, right? And with that, you're producing products, right? So how does this, how does this machine operate? You develop technology, process technology, you start with design and process technology, you develop the product, you transfer that technology to mass volume production, and then you do it again in two years, right? This is the business model, right? And it's very, very intensive. And I'll talk a little bit about that. Now, let's go back to, okay, where are these things produced? And this is where the call to action comes in. And if you look at the left side of this graph, right? It shows you back in the days, let's say in the 90s, right? This shows you the share of the production of the semiconductors. They were largely in US and Europe. Now, what happened? Do you see how US and Europe have shrunk and Asia have significantly increased? Do you think this is okay? As a global sort of, you know, presence, those three companies making this stuff for global presence of customers, you have a COVID happens, or you have some disruption, right? Geopolitics, right? Is this okay? It's not okay. It's totally not okay. It even goes beyond Intel. It goes to national security issues. It comes down to if you cannot have access to what basically is a part of our everyday life, whether you're in the hospital using that heart monitor, or you're in the MRI machine, or the cars, or TV you're buying, et cetera, et cetera. Would it be a good world to say, hey, you know what? Sorry, I'm going to allocate you what I think you need for your chips. I'm going to keep the rest for myself. That is not okay. So this is where the call to action is. This is why you keep hearing the chips acts. This is why you hear the bipartisan support of basically the administration, of Republicans, them or everybody else rallying around, the importance of why this is such a critical issue. And why? Because we just got the wake-up call during the COVID. Many of you went to, I don't know, your car dealership. What did the dealer say? Sorry, I cannot give you the car until six months from now. Why? Because we don't have these tiny chips. A $10 part was holding up the line. So it is call to action, and we're actually extremely encouraged by the everybody, especially in U.S. as well as Europe, really coming to realization that, hey, there needs to be a public and private partnership to really solve this problem, right? Because building these chips is not just, hey, it's not just a cave-on. Hey, cave-on, go build more chips. No. The infrastructure you need for this requires support beyond what Intel can do. So I just mentioned this whole industry thrives on what? On technology. Technology just doesn't happen by wish and hope, right? Who does technology? It's the people, right? It's the talent. It's all of you here who are going to be graduating with your baseline technical degrees, and you're going to go into companies like Intel or other leading-edge companies, and you're going to solve technical problems. So the partnership between universities, academia, government, infrastructure, all these things is critically important to change that trajectory around. And we know exactly the playbook. Why did these countries have the share of semiconductors? It's because they had policies, they had incentives, they had favorable environments, right? Where semiconductor manufacturing could basically operate and be very, very productive and profitable. That's going to change, right? I'm extremely encouraged by certainly President Mengcheng's, you know, vision of this whole industry. He's been absolutely putting his foot forward, probably the most famous university president in Washington D.C., if I can imagine. You know, he was directly reaching out to R.C.O. Pat Gildinger and me and just say, okay, let's sit around the table. This is so, so important. It's not just words, actions. And I'm extremely encouraged that, you know, Purdue lead the way for semiconductor degree program to really help to address this issue. And, you know, my friend Dr. Lonstrom here is leading this initiative and he's not just on Purdue, it's all of us as an industry. We have to come together. So hopefully you understand that why this is important. And, you know, there's some, you know, big, dollars that are, that is being, you know, put in place here. You see the CHIPS Act is about a 52 billion dollar type of an incentive support. Why? Because for someone like, you know, Intel or others who are going to produce these, these, these factories, you need to get the incentives that's going to be required to, to really get, get over the hump. Otherwise, it's a, it becomes very, very, you know, non-economical. And there's structural cost gaps between the United States versus places like Taiwan or Korea or China. And this is where this policy is going to help to level, help to level the playing field. So very, very encouraging. Now, I also want to give you a little bit of a feel for what are we talking about? Why is it such a problem? Just, just go, just go build this stuff. Right? So semiconductor manufacturing is not like any other manufacturing. We're talking about angstrom scale, right? Most people here are engineers, right? What's angstrom? 10 to the minus 10 or, or 0.1 billionth, right? That's, that's the kind of transistor dimensions we're talking about. That's, that's, that's Moore's law over 50 years on logarithmic scale. That's what happens when you just keep staying on that, right? It also requires extremely precise equipment to be producing this stuff, right? To be at that level of angstrom level for the kind of, you know, dimensions we're talking about, each of our equipment. For example, in lithography, there's a company called ASML that makes these equipment. How much you think is one of their equipment? One equipment, 100 million? Since you're recording me, I say it's too expensive, but it's about two, three hundred million dollars, right? So, yeah. So it's, it's just one equipment. And, and of course, what do they do? They are thriving on innovation too. Similarly, you need the position of gases, all these things that you put on those wafers, you need to etch the wafers, right? You know, basically put it simplistically, you know, these chips are basically produced on a wafer of silicon. You take this, you know, you do all the design, all the patterning, all the process and mechanical and chemical activities on the, on the wafer. You take it, once it's done through the fabrication process on the wafer, you take it to what we call the back end of the process and you dice it up into individual chips and then you package them, right? That's essentially how you produce these chips. And in order to produce them, right, you need humongous infrastructure, you have utilities, you have to have a lot of, you know, what matters in producing, you know, you're making that wafer, you have individual chips. How many good chips I get on out of that wafer? Do you think that matters? Hugely matters. That's yield. Do you think I want to know about how much dye fallout, each one of those chips fall out? Do I want to know at the end of the line or do I want to know it right when it happens? When it happens, because I don't want to throw too much money at it and then know about it when I'm done, right? So how would I know about it when it happens? That's when software, automation, analytical tools comes in, right? And by the way, making this stuff takes all kinds of minerals and chemicals, so supply chain is critically important. If something gets stuck in one part of the world, right? Then what? You bring your semiconductor fab down? I don't think so. I mean, it's just unthinkable for someone like me to think like even, you know, five minutes of a downtime on a fabrication facility. It's just not acceptable. You lose hundreds of millions of dollars. If just five minutes, you basically shut the operation down. Now, to build these things, right? Even though it's tiny chips, it takes huge infrastructure, like I was mentioning. Today, how much does a leading semiconductor fabrication facility cost? Just one. 20 billion. Actually, it's closer to 25. That's what my boss would say, K-1. You got to get it down to 20. But that's what it is. And you need to build these like cookie cutters, right? In order to upkeep to the trillion era, you need a lot of cash flow to do this. And to build it, you can sort of see the structural steels you need. You need all kinds of, you know, it's a humongous infrastructure. And you see that there's a, you know, as you see there, we have construction in Intel. We have these very similar of these 20 billion dollar type of things. We have them going in Ireland. We have them in Israel. We have them in Arizona. We have them in Ohio. Soon we're going to have them in Germany, right? And we also have them in Israel. So balancing supply chain is not just a talk. We are all in. We are all in, right? And in order to be all in, you also have to be leadership in technology. You have to deliver products that many of those companies that you said who don't make this stuff, right? They can basically count on you to deliver them the leadership performance that they are looking for. But it's not just, hey, let's just go build these things. Let's go, you know, pour huge amount of concrete and structure steel and all this stuff. You have to do it sustainably. This is why environment is hugely important in this process, right? And it starts with the value that you stand behind as a company, right? This is not something just, we're just thinking about. It's been decades that Intel has been sort of focused on, hey, sustainability. In fact, producing these chips requires a lot of water, you know, who wants to waste a lot of water, right? Nobody. So you have to have innovation where you're recycling. You're reducing your water consumption, right? Also, you make waste, right? Once you polish things or once you do all kinds of processes in your manufacturing, right? How do you dump that stuff? You don't want to dump them into landfills or ocean, right? You want to reduce the waste. And lastly, this is a grandiose mission. This is the big, hairy, audacious goal of being net zero. What does that mean? That means from all these operations that I said, all around the globe, right? We want to have zero carbon emission. But it's easy to say, but how do you get there? You have to innovate. You have to invest. You have to come up with alternative materials and processes, right? So this is why this is such an exciting journey ahead, right? Yes, the world is moving to digital era. Yes, it's going to be unstoppable trend, trend, but it takes innovation to get there, right? And that innovation never stops. So what are the kind of, your modern, what kind of people are you looking for? Well, it's not just one discipline, right? What it takes to win in this trillion era, right? It takes all the disciplines, right? Chemical engineers, you know, double E's, you know, mechanical engineers, you know, industrial engineers, you know, software, you know, usually important, you know, AI, you know, construction, environmental engineering, civil, you sort of see it's not just one discipline. So in my team, all of these elements are there. If one of them is missing, you know, the bus doesn't have a wheel, right? So and every one of these areas have to keep reinventing, have to keep raising the bar. And Intel and Purdue have a tremendous partnership. In fact, even even since, you know, President, you know, Chen and as well as Dr. Launstrom, you know, we're even up leveling that, right? You can sort of see on the research, you know, one and a half million of Intel funded investments with the university, you know, we have all kind of activities with SRC, $5 million there, you know, we, you know, rising star, I mean, number of these programs, lots of focus. My personal passion is also about talent development, right? And that's about, hey, some of the programs for the summer, you know, semi-one-on-one, semi-degree programming, these things are like role model for the entire nation, right? And then of course, many of the talent developments as well. So this is a journey. And as I mentioned, this is something that we as, you know, one of the leaders in the industry look upon ourselves to make sure that we are preparing the future generations to come. And I always say, with leadership comes responsibility. And Purdue is a leader in engineering, right? Everybody thinks of Purdue, they think of engineering, right? And with that comes responsibility. This is not just, hey, let's just listen to him and let's go away. No, what I want all of you to think about is what does this mean to me, right? Now, if you care about this topic and if you know, there is a role for you to play. I assure you because the future is critically important to transition into this digital era and it's going to require innovation. And that's where you're all sitting here today for. So with that, let me turn it over to you for any questions and you have. So I don't know how am I doing time wise? Okay, good, good. I held my promise. So I love to see any questions you have, anything that you want to ask me about, I'm going to be very transparent with you. As long as it's not about the stock price, as long as it's not about, you know, we are in a quiet period, so I cannot talk about market stuff. But rather than that, anything I talked about, I'm happy to answer. Hi, so my name is Francis Stollers. I'm an industrial engineer. On behalf of all I ease, it is a pleasure and a huge inspiration to see a fellow Purdue industrial engineer up on the stage. My question is about the crisis, semiconductor crisis after COVID. And besides international dependencies, what is the most important lessons to be learned from this crisis? You mean crisis? What are the crisis? Well, I think the first thing first is, it's just the human nature that you don't want to, you know, accept the reality. You know that, hey, this graph was, we were looking at this, everybody was looking at it. Hey, you're so dependent on Asia. You know, what if Shanghai Airport closes down? What if, you know, Swiss Callen closes down? But you're like, well, that means I got to go invest. I got to go put in hundreds of millions of dollars of this, I have to do this or that. So the human nature is like, well, that's never going to happen. You know, oh, I have all these other problems. Let me not worry about that. But when COVID happens, you realize, holy moly, this is important. Now, the good news for us in Intel was that we already had a very good resiliency across the world. You know, I mentioned we have operations all around the world because we thought about these things, right? But many of our suppliers didn't, right? So this is where, you know, when COVID happened, right? Many of you started learning from home. You started working from home. Imagine, I would have said, you know, world, sorry, everything shut down. You cannot work from home. You cannot learn from home because you cannot, you know, many of you and your family members, you had to buy laptops, right? All these things, right? Imagine you didn't do that. That's supply chain resiliency, right? We were able to sort of predict ahead. But we were so short. We were so short of supply because our suppliers, they did not, they were not able to accommodate, you know, the raw material. So we had enough inventory so we could, we could sort of respond to this crisis. So the lessons learned is you need to sort of not just think about what you need. You need to think about what your supplier needs. And you think about what their sub-suppliers need. And this is not something a human brain can do. This is where you have to use decision support tools. You have to use analytical models in order to help you understand where those vulnerabilities are, right? I call it that $100 million equipment. They cannot ship it to me. Why? I'm waiting for cable. Where the hell is the cable? Oh, it's in this little city in China. What the hell? What were you thinking? I didn't even know I have that cable coming from there. These were the tough discussions. But, but moving forward, we have learned our lessons. And this is why programs like Chips Act is trying to address, mitigate some of those risks. It's a good question. Other questions? Thank you so much for a great presentation. One of the charts you showed, you mentioned there's imbalance now versus balance 20 or 30 years ago. The way I looked at the chart, it seemed like there are five pieces of 20% each, roughly 50-50 Eastern Hemisphere and Western Hemisphere manufacturing. Do you see a future where that is seen as a balance perhaps with Intel having fabs all around the world or many corporations from around the world having fabs that are distributed where the customer is? It's a good question. Our view is, by the way, the manufacturing of semiconductors are not all created equal. There are mature notes that are still using 65 nanometer, 100 micron. These things are okay to be, and they're products that you have abundant availability. Where it becomes rather critically important is places where you are depending on these key products for future advancement of, let's say, military applications, surveillance, intelligence. If you're blocked from having access to those things, then people's lives are going to change. Now, is it a 20% or 30%? What is more critical is you have access to that capability, and it's not just manufacturing. It's the R&D. Remember, it's the foundational capabilities of those technologies is what you don't want to lose. Because manufacturing going elsewhere, it's okay. If you can make products and it's more efficient, let's say in one part of the world versus other part of the world, you get what you deserve. That's good for them. But if you lose the foundational technology capability, that is not okay. So I would not read too much into the percentages. What I would read more about is, do you have the foundational capability and access to those capabilities? The moment you lose that, game is over. Good question. Other questions? Yes? Yes, we have about two dozen folks online. So I have a few questions there. First one is, have you read the book, Chippor? And if so, what are your thoughts on it? Yes, I've not only read it, I've met Chris Miller, who's the author. He's actually, those of you who haven't, I would strongly recommend you to read it. I think he is totally echoing exactly what I just talked about. But he looks at it from a historical perspective. Why was semiconductors invented? It started with critical government requirements. It was about national security. It talks about how the market evolved. What happened with many of these companies that basically either did not innovate, either they kept their eyes off the ball. And lastly, it also talks about why the public and private partnership is critically important and talks about the overall infrastructure that's needed there. He's quite a thought leader. He teaches at Tufts University and be absolutely support and encourage. His thesis is very much aligned to what Intel is trying to do. He writes the book, we commit and put our neck on the line with hundreds of billions of dollars. So that's the name of the game. Other questions? Hi, my name is Steve. I'm doing my master's in engineering management. First of all, thank you for your talk. So my question is you mentioned how, you know, when COVID happened, you had to change how you think about things. And you also mentioned how you have a particular way, your business model is a certain way. When this change happened suddenly and you realized, sure you had it in the back of your mind, but you realized you have to change a lot of things now that things are starting to accelerate. How did you go about integrating that into your business model? How do you go about changing something that has been robust, but you need to change it fast without breaking other things in the network effects? Yeah, it starts with the expectation setting, right? It starts with, you know, for operation like the one I run, I have about 16,000 plus suppliers, right? And you have to sort of work with them to say, okay, you know, they sell you products and you tell them, okay, this is what we expect. So it starts with, okay, we would ask, and it's not, you know, with the button over people's head, you just say, hey, in a partnering way, like, look, in our, you know, let's say quarterly review of your business performance, we want to understand where your vulnerabilities are, you know, how you're doing against your lead times, what is your constraint, you know, some basic back to the basics of constraint management, right? That's how you basically really drive the entire ecosystem into, hey, how do you, you know, really drive a lot more efficient and I would say less risky of a supply chain. Now, it's not an overnight kind of a thing, right? The way you approach these problems is one way is just duplicate, you know, hey, if everything is somewhere else, let's just duplicate the thing. Well, that can be very expensive, some of these products, you don't have enough margins to just, let's just go build a new factory, right? You lose money, who would want to do this and lose money? So places where you can, you duplicate. There are places where you establish alliances, right? That's where the role of the governments come into play, right? Certain countries, they have certain resources, they have certain minerals, they have certain supply chain capability, so you build alliances. You can do regulatory measures, right? Where you say, you know what, certain things cannot be shipped from one country to the next, right? These are actions that when you hear the news, these are all surrounded around, hey, how do we create a more balanced supply chain, right? Good question. Other, why don't, why don't we go here? First of all, thank you for your wonderful talk. I'm just one, and my question is probably going to take you back to your time here at Purdue. I'm doing my masters in industrial engineering here, and I read in the speaker notes that you fondly recall your simulation course here by Professor Schmeiser. And I'm currently taking that course with Professor Hunter, who's in the same academy as you. And I was wondering, were your early years in Intel as an industrial engineer related to simulation work? If yes, then how? And if not, then maybe you could shed some light about how simulation is used in Intel. Any interesting? Yeah, good question. Yeah, so even to the state folks, you know, even though you hear me big responsibility, but I still have my college books, you know, in my office once in a while, I even opened them. Because in the real world, things don't happen exactly as you think, right? You, you know, I plan billions of dollars of capital, right? If everything went exactly as I thought, it would be amazing. Everything works the way. The world works on a basis of, it's very stochastic, right? You know, and that's where these, you know, models comes in. It's like, hey, you use some baseline of data, you model the probability of certain events happening, right? And then based on that, you plan your, you know, operation. In our world, in this, in highly capital intensive business, deploying your capital and how you use that, that capital asset utilization is a huge, huge deal, right? And this is where modeling like simulation is very helpful to really understand, hey, you don't want to plan for over capacity. So, so you want to be always planning for all these unexpected because when you plan for over capacity, what happens? You buy too much capital. Too much capital means you drag your margin down, but you also don't want to under plan because then these, you know, variability happens, whatever it could be, right? It could be a yield problem. It could be supply chain problem. It could be a, you know, some defect that's in the water line, right? These things, if it creates a shock into the system, you want to be able to plan for them. So, this is just one example of where you can apply simulation in a highly capital intensive business like what we do, and it's critically important. There's, we're talking with Dr. Son about, you know, like digital twin, right? Now, it's easier said than done because, you know, until you really come and work in the environment, you cannot oversimplify, you know, the complexity involved in running these operations, but it's a great start. Those are great concepts to have. Other questions? Yes? Oh, actually, go ahead first and then Dr. Son. All right. And finally, I'm currently a PhD student in industrial engineering, and I'm interested in semiconductor supply chain, especially the 2021 auto chip shortage. We all know it's a consequence from the interaction from different supply chains like the auto industry, like the consumer electronics, and the semiconductor supply chain is at the center of these different supply chains. So, I'm curious about, like, how to, what's the current practice of the semiconductor coordinate between, coordinate among the different supply chains, and maybe designing a new contract to, like, do the signals and improve the social welfare of the different supply chains. I wish I had a simple answer for that. It's a very complicated equation, you know, and if the semiconductor industry has many, many tiers of variability, just like we're talking about the operation, but even in the market, right? There are many, many elements that that affects that, right? We just saw one shock in the system, which was COVID, you know, another shock in the system could be, you know, some, some, you know, natural disaster in one part of the world where you're counting on, you know, earthquakes and things like that. You know, having said that, I think the best you can do is to know what your vulnerabilities are, and you always want to have essentially a contingency plan, right? And, and you also, that contingency plan, that means you have to qualify those capabilities, right? And that goes all the way down to every tier, down to the mineral, right? Even where that tungsten is being produced, right? Where the mine is, right? Those things matter. And you have to use, you know, analytical models in order to achieve that. And there's many tools coming about with the AI and so much data available. This is how you can bring and link these systems together, because you cannot manage this on locally optimizing this, but that from an industry perspective, there are ways that you can approach that. And, you know, we're in a journey. This might take a decade or two to get to the vision. But, but the, you know, anybody who tells you in this industry that what happens six months from now, don't believe them. Okay? That's just here from me. So including myself. Okay. Other questions? Yes. Dr. Langstrom. Yeah. Well, thanks again for a great talk and a good discussion too. So my question sort of gets at Jay's. You know, I'm, I'm old enough to remember Sputnik and the race to the moon. And, you know, there we had a clear objective. And when we landed one of our alumni on the moon, we knew we had succeeded. Now, when we think about the chip war, whatever we call it, is there a clear objective? How will we know when we have succeeded? And the second part of the question is help calibrate us on the magnitude of the challenge. Are we talking about five years of intense effort or will it take longer? I think you would know that, you know, I love to have a simple metric, like you say, like, let's get a man to the moon and bring them back safely. At this stage, right, I would say is you want to get to a point where semiconductors are no longer the constraint, right? Semiconductors are enabling capabilities that creates all these new business model that makes the people's lives better, right? And that, to me, is very simply measured. If you want to go innovate, if you want to go deal this big humongous AI, you know, transition, right? You know, you hear these things like chat, GPT, all these, they need what? What do they need? They need semiconductors. What do you think is the issue? If the whole world wants to go use it, you're back to square zero, you are going to be constrained. So to me, is, you know, you want to get to a point where semiconductors are no longer the constraint, and you're also no longer in a position that one region of the world is controlling this, where the innovation occurs. We need to see that balance, right? And, but then ultimately, you know, the, you know, big hairy audacious goal is stay on the Moore's law. You know, that is, that is, you know, as long as you stay on that, and it may not be the good traditional way of staying on Moore's law. It's not going to be, hey, you just, you know, maybe the new era is going to be in packaging. Maybe that's how you keep staying on the Moore's law. But those are to me simple metrics. Is A, you're not constrained, and B, you just keep staying on the Moore's law. But at least for the next couple of decades, maybe until I retire, right, Moore's law is alive and well, you know, so that is, that is the mission. We have to stay on that. Good question. I think they have something online. Do you want to? Yes. Or let's do it with you first and then let's go online. Okay. Hi, my name is Namisa, and I'm a software industrial engineering, and I'm actually participating in semiconductor research here as well. And I wanted to ask, I know with the CHIPS Act, it's all about how to reduce dependency that we have with other global countries, but how do you know how to draw the boundary between dependency and also global collaboration? Because it's also important to have connections with other countries and moving forward. How is it going to work? I think it's an excellent point, as I made the remark earlier, right? The way you solve this is not just by just let's just keep building factories, right? You solve this by three things, right? Number one is you want to have enough scale and manufacturing presence so that you keep the innovation, you keep the learning, you keep the talent, right? That's one piece. And where it makes sense, you replicate that. Then there's also alliances, right? You hear like a Japan, you know, Netherlands, you hear many of the European, you know, EU, they are all in alliance in order to figure out how do we all rally around the same common goal, right? That's where the alliances comes in. And then there's also regulations, right? That you have to put certain regulations. And regulations can be around how do you use these materials? Certain materials can be not, you know, you can have this humongous growth, but certain materials may actually not be as environmentally friendly. So regulations is also critically important, you know, where the leading edge technology goes, where the heck is that going to be used for, right? Is the leading edge technology going to be used to kill people? That's not okay. So there are regulations for, and I think those three things is how you can approach these problems. And again, it's going to require public and private collaboration to get there. But the goal, the common goal is the same. When you have common goal folks, amazing things happen, right? Even within your own teamwork, right? I always encourage you, before everybody gets together to team up, right? The best teams are which one? Where you have a common goal. You start with, what's the end in mind? Like Dr. Lancer said, what is the goal? Once you establish that, amazing things happen, right? This is no different. I just keep going until you tell me stop, because I don't... Okay, no worries. Actually, this question kind of builds on that. The US does not have a good track record in government backed industrial policy. How do you think we can do things differently this time? Okay. So let me start by saying I'm definitely not in politics. I'm not going to be able to comment on how political issues are going to be solved. So I will start with that. But I would say, certainly from my lens of basically leading a semiconductor operation, I have been very encouraged to see the level of progress in both US and European Union. Basically, we meet very closely, very regularly, with the leaders of both regions. And I would say more than ever, the level of urgency and the level of everybody rallying around this common goal is something that's actually quite refreshing, right? So as you're hearing the news, you saw the President Biden passing of the CHIPS Act. These things are happening, right? Because there's a bit of a forcing function and similarly in Europe as well. Because it's such a critical problem that unless you intervene, the problem is not going to go away, right? Because it's just structural differences between various GEOs are just so vastly different that it's just not going to work, right? So this is why the common goal around this so far is everybody understands. It's basically the graph that I just showed, right? Yes. Hi, everyone. Hey, Vijay. Good to see you. Great talk. Thank you for your leadership of the Intel Purdue partnership. The question I had was, I think there's one school of thought that says that some of the supply chain imbalance has really been driven by a cost issue, right? That supply chains tend to get biased towards regions where cost is lower. And as you just put it, I think, really well, we've kind of imposed a forcing function on that because it is so critical, I guess, from a larger perspective. So in some sense, can that same school of thought say as by if this forcing function succeeds, the cost of semiconductors is perhaps going to go up, right? So my question to you is, do you agree with that school of thought? And B, if so, is this essentially some kind of a national security tax essentially that everybody will have to pay for their semiconductors, which may be a maybe an okay thing? Well, first of all, I don't think it's going to harm the customers. I don't think it's going to create average selling price increases. And remember, why that degradation of the share of the world supply went from US and Europe to other parts of the world? Why do you think that is? Is labor cost? I'm investing $20 billion, $10 billion, let's say five years ago, 10 years ago on a FAP. How much of that do you think is labor cost? Very, very small. Make no mistake, folks. It's not about labor cost. This is when you're talking about $20, $30 billion. What happens? Supply chain, it comes down to favorable policies, right? They voted for it. The United States did not. That is the issue. So what do we mean by that? What is industrial policy? It's taxation, right? Mr. Corporation, come. You operate here. I give you tax incentives. Mr. Corporation, come here. I give you rebate on your capital investment. Mr. Corporation, come here. I will give you all the infrastructure you need, whether it's your water, whether it's your power, whether it's your land. You have a free land. I give you free building. And with that, Mr. Corporation, you're going to generate lots of jobs. You're going to generate lots of talent. You're going to bring all that innovation here. That's what's worth to me. So they give you billions in policy incentives and you generate them. More than billions of return. It's not an intel thing. Go read some of the third-party studies, consulting companies. I encourage you. There is 10x, sometimes 15x multiplier in a semiconductor investment. For every dollar you put in, there's 15 times more. Because why do you need supply chain? You need the infrastructure. You need the people that are going to deliver the stuff. You need the warehouse. You need the parts. You need the people running the water plant, right? You need the hotels. You need restaurants, right? Building these factories is like you're building a city, right? So they had a policy that they voted on. We didn't. That's what happened. And the thinking that, well, you're just going somewhere because it's cheap labor, that's fallacy. Good question. I know time is up, but I agreed to take the last question of today. Yes, thank you. My name is Michael. Thank you for your time in the presentation. We talked a lot today about the leading edge of semiconductors. What is the gap between the leading edge of the technology versus what is manufacturable and scalable on a widespread scale to take so many global customers? Yeah. It's an excellent question. It's actually a good place to end. The semiconductor industry has got a wide spectrum of applications, right? I talked about leading edge because that is what enables these high-performance computing, AI, and all these things you hear about. But there's a huge market for lagging what we call mature nodes, right? Mature nodes are your phone. They don't need your high-performance computing, right? They might need a power module IC. That's on older nodes. It might need a radio frequency module. It might need a signal processing. These things, you don't need to be on angstrom type of era, right? Now, these things come together. What controls your USB on your computer is a chip. That comes together with this leading edge. This is where packaging comes in. We used to put all of that because of monolithic integration. We put all of that onto a chip, right? But when you're buying equipments at 200, 300 million, you're building this humongous infrastructure, what do you think you want to do? You don't want to use that real estate for this mature stuff, right? You want to use your leading edge for that real estate is important. Then you bring those mature chips like somebody mentioned about memory, right? DRAM, NAND, high bandwidth memory. These things, you basically bring all those things together and you put them into a package. That package has a huge innovation that's required. This is those of you who are mechanical engineering, right? The tensile strength of this thing. The warpage. How do you connect the chips together? How do you make sure there's overlay and alignment? This is a whole new era that's going to really drive this industry forward. But there's also, there's a huge trend also. You hear Intel inside, right? We build chips for PCs and data centers. Well, guess where the world is going to, right? They're like, while I appreciate your chip, Mr. Kvan, but I'm going to be using the chip for a car application, for EV application. I'm going to use the chip for a graphics application. I'm going to use it for a wireless application, right? So I want to be able to get that chip that's customized to my needs. That's where you hear the name of companies like NVIDIA, Apple, AMD, Qualcomm. What do they do? They design their chips, and they're going to come and say, hey, Mr. TSMC, Mr. Kvan, go produce this for me. So that's a whole new era, right? And this is what Intel hasn't done, but Intel is getting into that space as well. So with that, folks, I want to sincerely thank everybody for your time. It was really honor and very, very relevant questions. Hopefully, this was useful for you. And with that, I'm going to turn it back over to Dr. Sun here. Thank you. Well, Kvan, it's great to have you back to our campus. And let's give Kvan a big round of applause, okay? So thank you for coming today's talk. And I want to thank the College of Engineering team for preparing this outstanding talk series, okay? So have a great day.