The Future of Semiconductors Promises to Be a Period of Growth, Risk and Opportunity

(DESCRIPTION)
Logo, Travelers. Text, Decoding the Future of Semiconductors: Emerging Trends, Risks, and Opportunities. Amanda Bohn appears in the upper right corner of the slide presentation. She is seated and speaking to us with a Travelers umbrella logo in the background.

(SPEECH)
AMANDA BOHN: Welcome, everyone. On behalf of Travelers, thank you so much for spending some time with us as we decode the future of semiconductors, looking at emerging trends, risks, and opportunities.

(DESCRIPTION)
Text, this material does not amend, or otherwise affect, the provisions or coverages of any insurance policy or bond issued by Travelers. It is not a representation that coverage does or does not exist for any particular claim or loss under any such policy or bond. Coverage depends on the facts and circumstances involved in the claim or loss, all applicable policy or bond provisions, and any applicable law. Availability of coverage referenced in this document can depend on underwriting qualifications and state regulations.

(SPEECH)
The next decade in semiconductors promises to be a period of explosive growth, profound change, and significant opportunities. New innovations may enable a continuation of Moore's law, with profound implications for advanced electronics and software, and in the life sciences industry.

As our reliance on digital technology accelerates, the role of semiconductors becomes increasingly critical. Look around the room that you're in right now. I bet you can look and find five things that have some type of semiconductor device embedded in them.

They power our consumer electronics, our automobiles, our critical infrastructure, and nearly all military systems. To seize the opportunities, semiconductor companies must also contend with a range of geopolitical, operational, and technological risks. Let's take a closer look at how they can prepare.

(DESCRIPTION)
A slide titled, Our Panel, shows headshots of the three speakers. Text, Moderator, Amanda Bohn, Chief Underwriting Office, Technology and Life Sciences at Travelers. Panelist, Chris Miller, Associate Professor of International History at The Fletcher School, Tufts University, and author of "Chip War." Panelist, Richard Gu, Vice President of Investor Relations at Cadence Design Systems.

(SPEECH)
I'm Amanda Bohn with Travelers. And I'm so happy to have Chris Miller and Richard Gu joining me in today's discussion. Chris Miller serves as associate professor of international history at the Fletcher School at Tufts University and co-directs the school's Russia and Eurasia program. He's the author of Chip War-- The Fight for the World's Most Critical Technology, a book that explains how computer chips have made the modern world and how the US and China are struggling for control over this fundamental technology.

The future of computing, the book argues, will be determined by who controls the ability to produce the world's most advanced chips. Chip War won Financial Times Best Business Book of the Year Award and was described by The New York Times as a nonfiction thriller. Chris received his PhD and MA from Yale University and his BA in history from Harvard University.

Richard Gu is vice president of Investor Relations for Cadence Design Systems. Headquartered in San Jose, California, Cadence is a pivotal leader in electronic systems design, building upon more than 30 years of computational software expertise. The company applies its underlying intelligent system design strategy to deliver software, hardware, and intellectual property that turn design concepts into reality.

(DESCRIPTION)
Why Now? A small rectangular semiconductor chip.

(SPEECH)
The past couple of years of the semiconductor industry has seen supply chain challenges, decline in earnings, and cost-cutting measures. But

(DESCRIPTION)
Words appear above the chip photo. Text, Embracing Supply Chain Diversification, Digitizing Many of Their Processes, Bring Manufacturing to the US.

(SPEECH)
I actually think the future of this industry is pretty bright, as companies refocus by embracing supply chain diversification, digitize many of their processes, and most notably, bring manufacturing back to the United States. Travelers wants to watch how this all unfolds.

But we also want to play an active role in education and risk management. Next year, our Technology and Life Sciences practice will celebrate 40 years, 40 years of specialization in ensuring companies in the life sciences and technology sector and, specifically, the semiconductor industry. Over these 40 years, we've partnered with companies as they introduce groundbreaking product advancements, and when they've experienced devastating setbacks, like product liability claims, fires on the production floor, or cyber events that bring business to a screeching halt.

I think we all, Chris, Richard, and Travelers, have a vested interest in better understanding the challenges and the risks. Because there's so much amazing material and I have a lot of questions for our panelists, we're not going to take live questions today. But if you do have a question, put it in the chat. We'll do our best to reach out.

Let's get started. Chris, I'm going to start with you. Tell us a little bit about your background and how you got interested in the topic.

(DESCRIPTION)
Photo of Chris Miller. Text, Chris Miller, Associate Professor of International History at The Fletcher School, Tufts University, and author of "Chip War." Chris replaces Amanda in the upper right corner. He is seated in front of a bookcase and speaking to us.

(SPEECH)
CHRIS MILLER: Well, first off, Amanda, thank you for having me today. It's very fun and enjoyable to have the chance to discuss these issues with you and with Richard. My background is as a historian of Russia, which is a strange background in some ways for someone who ended up writing a book on semiconductors.

But I wanted to understand the determinants of the arms race during the Cold War between the Soviet Union and the United States. And I thought it would be a story about missile systems or fighter jets. But it turned out that, actually, all of the key defense technologies that emerged during the Cold War were fundamentally reliant on computing, sensing and communications capabilities, which, in turn, were only possible because of semiconductors.

In fact, I learned, as I did my research, that the first chips that were invented in the late 1950s and early 1960s were used in guidance computers, like the guidance computer that was in the Apollo spacecraft that sent the first astronauts to the moon, but also guidance computers in missile systems. And so, there's been a very deep relationship between advances in computing capabilities, advances in semiconductors, and the development of the US military as it sought to bring new technologies and new computing capabilities to bear. And so that was my initial source of interest in semiconductors.

But as I dug into the industry, I quickly realized that, as you alluded to in your introduction, modern life simply can't function without chips. And so, it seemed to me that, in fact, whether you're trying to understand the military dynamics in the world today, which was my initial interest, but also if you're looking to understand the shape of globalization, you could understand it without looking at semiconductors. And today, just to give you one data point on that front, China spends as much money each year importing chips as it spends importing oil. In other words, the world's second largest economy, its international trade flows can only be understood by reference to semiconductors. And so everywhere you look, literally in terms of around your office right now, or figuratively in terms of the key trends that are reshaping the world, all of them depend fundamentally on semiconductors.

(DESCRIPTION)
The book cover of the book "Chip War." The illustration on the cover is an American flag with the blue field of stars replaced by a semiconductor chip. Text, Financial Times Business Book of the Year, An Economist Best Book of the Year, New York Times Bestseller. Chip (also "integrated circuit" or "semiconductor"): a small piece of semiconducting material, usually silicon, with millions or billions of microscopic transistors carved into it.

(SPEECH)
AMANDA BOHN: I wanted to just ask you a little bit more about your book. You wrote Chip War. And I loved what The New York Times called it-- a nonfiction thriller. You don't see that every day. And you do such a nice job of taking a really complex topic in a really sophisticated industry and you make it really understandable.

And there's 500 folks out there that are listening live right now that registered for this webinar. And they're going to get a free copy. But Chris, there's some folks that haven't read that book. So maybe give us a quick summary for our listeners on the thesis of your book Chip War.

(SPEECH)
CHRIS MILLER: Well, the book argues that semiconductors are the most complex devices that humans have ever manufactured. If you go to the Apple Store nearby and buy a new iPhone, for example-- this is true of basically any smartphone-- you'll find that inside of it are chips with tens of billions of transistors, tiny little switches that turn circuits on and off. And each of these transistors fits on a chip roughly the size of your fingernail, making them the smallest devices that humans have ever produced at scale. Each transistor inside of your phone, for example, or in your computer, is roughly the size of a coronavirus. And yet we produce them by the billions and billions.

And so, bringing together the manufacturing and design capabilities that are needed to undertake this type of ultra precision manufacturing involves super complex supply chains, stretching from Europe to the United States to Japan to Taiwan. And so, there's really no industry that is as globalized as the chip industry. There's not a single country that can produce advanced chips on its own, nor is there a single company that can do it on its own, as well, because they all rely on suppliers and customers that have their own unique and ultra complex capabilities.

And so on the one hand, the chip industry is a story of extraordinary collaboration across the supply chain between different companies and countries. But this also does create vulnerabilities because there are often just a handful of companies and, in some cases, just one company that have unique capabilities that nobody else can replicate. And so when you look at the centrality of chips in the world today, it's a story of technological progress, of driving down the cost of chip-making. But it's also a story of complex supply chains that increasingly are seen as not just a strength but also, at times, a vulnerability.

(SPEECH)
AMANDA BOHN: Chris, that's a little mind-boggling, that there's this huge global dependence on something that impacts global trade and the balance of military power, and it's smaller than my fingernail-- my goodness. Let's talk about supply chain for a second. We've all seen so many papers and news pieces about the impact the various world events have had on the supply chain in the semiconductor space. So, could you lay out for us, maybe, a high-level picture of the global supply chain and help us understand the current situation?

(DESCRIPTION)
Current Supply Chain. A hand holds up a smartphone in an aisle of a warehouse. An overlay graphic shows a world map in the center with different icons radiating around it, including a magnifying glass, a Wi-Fi signal, an airplane, gears, a cloud, a chat bubble, a truck, a tablet, and a dollar sign.

(SPEECH)
CHRIS MILLER: Well, to start, if you want to make a chip, you first have to design it, lay out where each of the 10 or 20 billion transistors is supposed to go. And just given the numbers involved, this is an extraordinarily complex task. And so, there are a handful of firms that produce specialized design software that help companies automate their process of laying out components on a chip.

And then the design itself is also highly specialized because different chips do different things. In your phone, for example, there's one chip that manages the operating system, a different chip that remembers data, other chips that manage the Bluetooth or the Wi-Fi or the cell connection. There's a chip that knows whether your phone is held this way or held this way. And

(DESCRIPTION)
He holds an imaginary phone as if showing it to us. First, he holds it vertically, then he flips it horizontally.

(SPEECH)
all of these chips have to be designed to accomplish that specific task. And so different companies have different expertise in designing for these different types of use cases. And that's just the design. Then you actually have to get to the manufacturing of how do you fabricate lots of virus-sized transistors by the billions and billions. And that itself is just as complex.

(DESCRIPTION)
Text, Market Share Concentration. "These tools are produced primarily by three different companies."

(SPEECH)
There's a small number of companies that produce the tools that are capable of moving materials at almost the atomic level that's necessary to produce this precision, companies that can make tools that can deposit thin films of materials just a couple of atoms thick or hatch tiny canyons into silicon just a couple of atoms wide. These tools are among the most complex and expensive tools ever made. The most expensive cost $150 million a piece, so this is extraordinarily complex equipment.

But then you actually have to do the manufacturing. And that's a different set of companies that undertakes manufacturing processes. And when you're talking about advanced memory chips or advanced processor chips, there's just a handful of companies that can do either set of processes.

Finally, the last step is, once you've got your chip, you have to assemble it and package it into a device, into a smartphone or into a computer. And there's not a single country or a single company that can do the entire supply chain. Everyone has to collaborate with partners in different segments of the industry.

(SPEECH)
AMANDA BOHN: Thank you, Chris. I think this is a perfect time to bring Richard into the conversation and get his perspective on this, especially given his expertise at Cadence. But Richard, before you do comment on that, we'd love to hear a little bit about Cadence Design Systems and the role that Cadence plays. And then we'd love to hear a little bit about your view on the global supply chain situation.

(DESCRIPTION)
Richard replaces Amanda in the upper right corner. He is seated in a sunroom with a mountain view outside the windows.
Text, Cadence at a Glance, Computational technology for designing today's electronic systems. Leader: Computational software for Intelligent System Design, Trademark. $977 million: Q2 2023 Revenue. Culture: Innovation: created by engineers for engineers. Industry: Software and programming. About 10,800 employees worldwide. 26 global development centers. Nasdaq Ticker: CDNS. S&P 500 and Nasdaq 100 indices. Source: Cadence Earnings Press Release, July 24, 2023.

(SPEECH)
RICHARD GU: Thank you, Amanda, for having me. So, Cadence is a computational software company. And we help our customers design chips and electronic systems. As you all know, all aspects of our lives are being increasingly electrified and digitized riding on the backbone of semiconductors. So, chips are ubiquitous these days and are powering everything around us, as you mentioned, Amanda.

So it is, by far, one of the greatest human inventions. And it's extremely complicated to design and to produce. Just to give you an idea, the most sophisticated silicon has around 100 billion transistors on a 1 inch by 1 inch die. And it's mind-boggling to fathom let alone to be designed manually for that kind of scale.

They can only be designed using highly sophisticated software. And our software is essential and foundational to help automate and optimize the design process, thereby providing tremendous power, performance, [INAUDIBLE] benefits, and productivity improvement for our customers, which then gets amplified and accrued to the entire society as these chips get embedded into every aspect of our lives.

I think it's fair to say that almost any chip in the world uses some form of Cadence software and IP. We participate in all end markets, all verticals, and all geographies. And that's what Cadence is all about in a nutshell.

Amanda, back to your question on the semi ecosystem and to add to what Chris just talked about, upstream in the ecosystem, you also have companies such as ARM, that provides architectural design IP, or ISA, instruction set architecture, for computer processors. So, they license the processor IP to other companies. In order to make a chip successfully, it is critical to have close collaboration amongst the architecture IP companies, the EDA companies like Cadence, the semi systems companies, equipment, and the foundries. As you can tell, US and the Western world still control important parts of the semi ecosystem, except for manufacturing, while the manufacturing is more concentrated in Asia, hence the tension and the regulations Chris and [INAUDIBLE] just touched upon. Back to you, Amanda.

(SPEECH)
AMANDA BOHN: Thank you, Richard. I think Cadence is a fascinating firm and plays a pivotal role in the broader ecosystem. Thank you for those comments, Richard. Chris let's go a little bit deeper on the chip shortage and the implications this has had on the industry.

(DESCRIPTION)
Recent Challenges. A close-up of a semiconductor chip installed on a board.

(SPEECH)
CHRIS MILLER: Well, during the pandemic, there was indeed a shortage of semiconductors for a number of use cases, though I think this has often been misunderstood in the popular perception. Most people are surprised to know, for example, there are actually more semiconductors produced every year of the pandemic, more chips produced in 2020 than in 2019, more chips again produced in 2021. But the pandemic brought a lot of unexpected demand patterns, for example, a surge of purchases of new PCs as people prepared to work from home during the pandemic.

And so, supply and demand for different market segments was out of whack because it didn't accord to normal patterns. And so, what this meant was that, for certain types of semiconductors, there were extraordinary delays. The best example of this is in the auto industry, where the auto industry is estimated to have lost almost half a trillion dollars of sales, if you include lost sales in the US, Europe, Japan, and China all added together over the course of the pandemic. And this is because a typical new car can have 500 or 1,000 semiconductors inside of it.

And these chips are often different types of chips. You need to have every single one to actually have your car roll out of your factory parking lot. And just the delay in a single type of semiconductor can cause a $40,000 or $50,000 car to sit in the factory until that chip arrives, even if the chip itself only costs $1 or $2. And car companies were the best example of manufacturing segments that found themselves extraordinarily reliant on semiconductor supply chains that they often had very little visibility into.

And indeed, there are great anecdotes of auto CEOs calling up semiconductor manufacturers and asking, do you produce this type of chip I'm looking for, because they often didn't know. It was their component suppliers or the component suppliers of their component suppliers who were facing shortages. And so there's been a major push over the past couple of years after the chip shortage for manufacturing firms to better understand what their semiconductor supply chains look like and to rejig their purchasing processes to make sure they don't face the similar scale of difficulties that they often faced during the pandemic.

(SPEECH)
AMANDA BOHN: Yeah, Chris, for anybody that was trying to buy a car during the pandemic, they know this issue quite well. Richard, I'd love your view on this topic. What are your thoughts on the chip shortage and the fragility of the semiconductor supply chain?

(DESCRIPTION)
A pile of chips. Text, "Certain segments are now actually experiencing an over-supply."

(SPEECH)
RICHARD GU: Yes, certain segments, such as memory chips including the DRAM and NAND, have actually been a situation of glut or oversupply as of late. So, there are some green shoots that indicate that this down cycle appears to be bottoming out. And that's the fascinating part of the semi-industry, where the product cycle are multi-years, and those that can really portend the future trends and where the market is headed in order to better match demand with supply and innovation would be best positioned.

At present, clearly, the AI chips are in a high demand environment and is experiencing a supercycle. The strong demand for that will drive both design and production activities, as well as resource allocation for years to come. The AI infrastructure build out will benefit platform companies such as NVIDIA, Broadcom, et cetera, as maybe the first wave.

The second wave will likely be those vertical players, such as data centers and hyperscalers, who will design and come up with their own custom built AI chips, specifically tuned for their own applications. The third wave will be accompanied by new applications, new verticals, and new use cases of AI, such as digital biology. Back to you, Amanda.

(SPEECH)
AMANDA BOHN: Thanks, Richard. And as we talk about supply chain risk, geopolitical risk and the recent events just really have to come into the discussion here. We have the Russian invasion of Ukraine and the devastating Israel-Hamas war right now. But we also have China and Taiwan. And there are real disputes here. And the relations are tense. Chris, tell me a little bit about the nature of the conflict and maybe the implications this has on the semiconductor industry.

(DESCRIPTION)
The skyline of Taipei. Text, Importance of Taiwan. "Chips from Taiwan provide 37 percent of the world's new computing power each year."

(SPEECH)
CHRIS MILLER: Well, the China-Taiwan conflict actually predates the invention of the first semiconductors. But today, it's impossible to understand the China-Taiwan risk without looking at the implications for the chip industry. The origins of this dispute go back to 1949. Since then, the Chinese government has claimed control over Taiwan, whereas today, most Taiwanese leaders say they would prefer to be treated as an autonomous or even an independent country.

And so, this dispute has been going on for quite some time, with the US supporting Taiwan in trying to retain the status quo, in which Taiwan is self-governing but not exactly independent. And this matters for the chip industry today because Taiwan is an absolutely irreplaceable manufacturer of processor chips.

And when it comes to the most advanced processor chips, the types of chips in your phone, in your PC, and most importantly, perhaps, the types of chips that are undertaking AI processing in data centers, what you find is that around 90% of these semiconductors are made by just one company, which has all of its cutting-edge manufacturing right now in Taiwan. So if there were some sort of geopolitical escalation or military escalation between China and Taiwan, the implications wouldn't just be important for the political impact. It would also be important because it would risk disrupting the entire world's access to the most advanced processor chips.

And it's worth noting that every country in the world today depends on Taiwan for really irreplaceable capabilities. And although many people on this call will have never heard of the Taiwan Semiconductor Manufacturing Company and hardly anyone has actually bought a product from them, in fact, all of us, throughout our daily lives, rely on dozens, hundreds, even thousands of chips that are produced by TSMC because their chips are everywhere, in data centers, in the telecom’s infrastructure, in our smartphones. We simply can't live without them.

(DESCRIPTION)
The entire world depends on advanced fabrication from Taiwan and South Korea. A bar graph, titled, Breakdown of the global wafer fabrication capacity by region, 2019, shows the percentage of global capacity Memory, Logic in four different n.m. sizes, DAO, and total, for the US, China, Taiwan, South Korea, Japan, Europe, and other. The US, Europe, and other percentages are considerably smaller overall than China, Taiwan, and South Korea.

(SPEECH)
AMANDA BOHN: So, Chris, you touched a little bit on that there. But let's go a little deeper because we're trying to decode the future in this webinar. If the conflict over Taiwan were to escalate, what would be those implications?

(SPEECH)
CHRIS MILLER: Well, I think if the chip shortages during the pandemic any indication are, we should expect vast disruptions to manufacturing operations. If the pandemic chip shortages could cause half a trillion dollars of losses to the auto industry, despite that chip production increased every year of the pandemic, imagine a situation in which a third of the new processing power that's produced each year, which is the share that's produced on Taiwan, is no longer available. The implications would extend far beyond phones or PCs.

It would also implicate a whole set of manufacturing industries that you probably don't think of as being in the tech sector-- construction equipment, for example, agricultural equipment, life sciences and medical devices. Today, almost everything has semiconductors inside. And a huge share of those chips are produced in Taiwan. And so even for devices that don't necessarily require chips made in Taiwan, just the collapse in supply because the world's most important chip maker could be knocked offline would have implications for basically every sector of the economy.

(SPEECH)
AMANDA BOHN: What a great point. There's just such a massive world dependency on chips, but of all sophistication levels, not just those that are powering data centers or AI. Chris let's stay on this geopolitical risk category just a little bit longer and talk about the Russia-Ukraine conflict. I've also come to learn that you're an expert on Eastern Europe. So given your expertise, what are the implications of the Russia-Ukraine conflict on raw materials specifically that are used in the semiconductor manufacturing space?

(SPEECH)
CHRIS MILLER: Well, it's a really interesting and actually an unexpected question because neither Russia nor Ukraine produce hardly any chips. Most people were expecting the war to be irrelevant for the chip industry. Relevant for the rest of the world in many ways, but they weren't expecting a direct impact on semiconductors. But

(DESCRIPTION)
A circuit board is broken into many jagged pieces. Text, "Even countries with tiny roles in the chip industry are trying to use their role to their advantage."

(SPEECH)
in fact, when you dig into the materials that are used in manufacturing semiconductors, which includes not only the silicon that chips are made out of, but lots of layers of different metals, from cobalt to palladium, as well as all sorts of highly specialized gases, what you find is that many of those actually come from either Russia or Ukraine. And there was an active effort by Russia very early on in the war to weaponize the supply of neon gas. Before the war, half of the world's neon gas was being produced either in Russia or in Ukraine.

And so the lesson of this is that even countries that might not be seen as having a big stake in the chip industry or the computing industry more generally realize that this is such a critical infrastructure, such a critical set of companies and capabilities, that there's an incentive often for political leaders in certain countries to try to disrupt production in the hopes of causing shortages and higher inflation in the West. Now, that effort by Russia didn't really succeed. There were enough alternative sources of supply of neon gas, for example, to be brought online fast enough that production wasn't seriously disrupted. But I think it's a cautionary tale for future potential conflicts and a reminder that we need to think through the ways that other geopolitical crises could also implicate the chip industry.

(SPEECH)
AMANDA BOHN: Yeah, Chris, weaponizing a raw material like neon gas, that's just a frightening thing. So, let's pivot a little bit. This might be a good time to talk a little bit about policy. What are policymakers doing to prepare for these challenges? And do you think, Chris, are they going to have an impact?

(DESCRIPTION)
Policy Changes. A graphic shows a document with a dollar sign on top, a pen, and a bag with a dollar sign on it. Text, Subsidies and Incentives. A small diagram shows a dotted line circle connecting a factory at the top, people to the right, a truck at the bottom, and a chip to the left. Text, Tech Export Control. A large gear with a robot arm in front working on a chip. Text, Shifting Chip Production.

(SPEECH)
CHRIS MILLER: Well, you've seen over the past five years or so from all of the world's biggest economies, China, the US, Europe, Japan, an effort to shore up each of those country's own supply chains. In the US, for example, there's the CHIPS Act, which is devoting $52 billion to try to stimulate chip manufacturing in the United States. Today, the US produces around 10% of the world's chips but consumes around 25% of the world's chips. So, there's a big chunk of semiconductors that have to be imported.

And given the geopolitical concerns that we've already discussed, in Congress and in the White House, there's increasing concern that this is a security but also an economic vulnerability. And so, the US is trying to reduce the cost gap of manufacturing chips in East Asia versus in the US because it is more expensive in the US and, in doing so, to attract more investment for chipmaking. And we already see that there is a major increase in investment in new facilities relative to the prior trend.

But it's still a bit of an open question as to how sustained this trend of higher investment is going to be. It's not that much of a surprise that if the government offers incentives to invest, more companies will invest more. The question is what happens once the incentives are gone. And $52 billion might seem like a lot of money. It is a lot of money in many ways. But to the chip industry it's actually a small amount of money.

Just to put that amount of funding in comparison, in context, that's not all that much more than the Taiwan Semiconductor Company's capital expenditure budget this year, which would be around $35 billion. So, in that context, we shouldn't look at this as a game-changing sum of funds. It'll make a difference, but it's not transformative. But when you look at that next to what Europe's doing, which is something similar, what Japan is doing, which is also something similar, you do see a sustained effort by Western policymakers to invest more in diversifying the manufacturing footprint, precisely to insure themselves against the types of risks that we discussed in supply chain.

The second aspect is that there's a lot of new regulations and controls being put into place. And here you've seen, over the past couple of years in the US, the Biden administration restricting the transfer both of certain chips and also of chipmaking tools to China. And the goal here is to limit China's ability to access the types of processors that are used in artificial intelligence, training cutting edge AI systems, which today are basically exclusively designed by US firms and exclusively manufactured in Taiwan. And so, there's a whole second set of actions that are designed to restrict the transfer of US, Japanese, European, other types of technology, to China, with the aim of limiting China's AI capabilities. And so, both the promoting industry and also the protecting technology transfer, both of those remain a focus of policymakers in the US, and also in other Western capitals.

(SPEECH)
AMANDA BOHN: Thanks, Chris. So, it seems like we've kind of laid the groundwork around talking about the supply chain risk and the fragility around it, the geopolitical risk, and some of those really wonderful insights around policy changes and the challenges associated with them. At Travelers, we insure thousands of companies in the semiconductor industry space. And we partner with these customers and their agents and brokers to help them prepare for those challenges ahead.

And good risk management is all about preparation, mitigation, and prevention. So, let's pause here and dive deep in that. Richard and Chris, from each of your unique viewpoints, how can leaders in the semiconductor industry space and their customers prepare for these challenges? Chris, I'm going to start with you on that.

(DESCRIPTION)
Preparing for Challenges. A graphic shows a semicircle with five circles with graphics inside around its circumference. The one on the far left has a box with a circle around it punctuated by dots. Text, Understand Supply. The next circle to the right has a graph under a magnifying glass. Text, Track Changes. The next has a document under a magnifying glass. Text, Prepare Second Sources. The next circle has a gear in the center with four arms around it, one with a person in it. Text, Understand roles. The circle on the far right has a notebook in it with a list on it with three stars down the left side. Text, Contingency plans.

(SPEECH)
CHRIS MILLER: Well, I think the first challenge for companies in this space is to put a lot more thought into understanding ways the regulatory environment is changing and ways the supply chain is changing. The fact that every major economy is trying to reshape how the supply chain functions adds a lot more uncertainty and certainly more variables to track than has ever been the case. So, companies are spending more time just trying to understand the landscape than they've ever had to in the past.

The second aspect for customers of semiconductors is to think hard about their suppliers. Companies are trying harder to find second sources of supply, whether that's having two companies supply them or at least sourcing semiconductors from two different geographies. And that's not always possible. There are some cases where there really is only one supplier that can produce the type of chip you need. But thinking about diversification of supply is something that basically every manufacturing company is doing.

In addition to that, companies are also thinking harder about the amount of supply they have in inventory. One of the things you found with auto firms during the pandemic was that companies that had the most just-in-time in their supply chain had the most difficulty in responding to the chip shortage. And so, for certain critical components that are hard to find second sources of, companies are thinking about whether they need to stockpile a bit more inventory to give them resilience in case of disruption.

And finally, another point that is getting more attention is thinking about potential security risks to hardware. As every device becomes more and more reliant on many different types of semiconductors, there's more concern than ever on making sure that the security implications of sourcing from certain types of companies are understood. And I think, just like we've seen a huge growth over the last decade in focus on cybersecurity and software, connections to the internet, hardware security, I think, is going to be a growing focus over the next decade, as we understand more about the challenges of sourcing hardware from certain supply chain actors.

(SPEECH)
AMANDA BOHN: Thank you for that, Chris. Richard, what's your view on this?

(SPEECH)
RICHARD GU: I think I agree with most of the points Chris just made. I think the only thing I would add or amplify is that, really the chance, all the risk mitigation favors the prepared mind. So from my vantage point, having a contingency plan for various scenarios is imperative. Pick Cadence as an example.

We have about 26 global development centers where we can shift design R&D work around and dial up or down depending on the customer needs and the circumstances in a very agile way. I remember when the Russia-Ukraine conflict first broke out. We were able to close down the office in Russia and move the workload somewhere else rapidly. But again, software is much easier. And the physical plant and supply chain takes longer and requires, really, forethinking and planning.

(SPEECH)
AMANDA BOHN: Thank you Richard. I'll round this out. I think, Chris and Richard, you offered great suggestions.

(DESCRIPTION)
Text, Risk Implications. A row of chips are standing upright on a table. The last chip is at the edge of the table about to fall off.

(SPEECH)
I want to underscore the point around diversification. I think that's so important. And I think the industry learned a lot during the chip shortages around why diversification in the product line as well as your suppliers is important.

But equally as important is taking your company's robust quality control program and ensuring that your backup supplier can meet the level of quality that you expect from your primary supplier. Inferior or lower quality product can create risk. And you don't want to be figuring this out when you're in a pinch for a raw material or a subcomponent. So thoughtful and tested contingency plans really does reduce risk.

Also, firms in the semiconductor space are global, of course. They may design in the United States. But they might ship-- or they might manufacture overseas and then ship worldwide. Or they might design and manufacture here in the US, but the critical components are coming from overseas. So having an insurance program for your foreign exposures that's just as robust as your domestic program is really important, and then also making sure that those coverages are dovetailing. And also, having both programs foreign and domestic with the same insurance carrier really ensures that there isn't going to be a coverage dispute between those two carriers. That does reduce risk.

OK, so let's switch gears here. Richard, when I started focusing on the technology space many, many years ago, I was learning about the semiconductor industry. And I recall learning about Moore's law. And

(DESCRIPTION)
A laptop screen shows a simple line graph with a clear upward trend. Text, Moore's Law: An observation that the number of transistors on a microchip double about every two years, though the cost is halved.

(SPEECH)
I know, Richard and Chris, you guys are experts on Moore's law and really familiar with it. But for those that aren't, this was a forecast that was made in 1965 by Gordon Moore, that the number of components on a chip or an integrated circuit would double approximately every two years. It's actually one of the greatest technological predictions in the last half century. So, Richard, with that as a backdrop, can you describe the current state of semiconductor technology?

(SPEECH)
RICHARD GU: Sure. It's a great question, Amanda. So, some say Moore's law is dead. And some say Moore's law is alive and well. I think the real answer is maybe somewhere in the middle, in that Moore's law is slowing down over the past five, 10.

Years, but there are still plenty of innovation and opportunities to be had. Dennard scaling, which is really the real science behind Moore's law, observes that transistors are -- as transistors are reduced in size, their power density stays constant. And that stopped a while back. And what we have been doing over the years is really to get more space scaling as opposed to power and performance scaling.

So that's the reason that the chips are getting bigger and bigger these days. And the power and heat dissipation are increasingly becoming one of the key challenges as we scale the area. As a matter of fact, the number of transistors we're packing into Intel chip is expected to increase from 100 billion transistors right now to a trillion in a decade. And so, when Moore's law slows down, it creates tremendous opportunities for many areas, such as 3DIC, or what we call chiplet, where we can use creative packaging technologies to interpose and connect multiple heterogeneous types of chips together to gain the maximum efficiency.

Another example of the opportunity is by applying AI to our software to do more with less. So generative AI is a perfect example of leveraging reinforcement learning and parallel computing to come up with a dramatically better product, while continuing to improve productivity and shorten time to market.

(SPEECH)
AMANDA BOHN: So, Richard, the technology is becoming ubiquitous. It's at the heart of everything, all of our lives. Moore's law is slowing down, leading to an opportunity for some innovation, it sounds like. And I want you to help us decode the future. How do you see the industry evolving and maybe the new opportunities and challenges it presents?

(DESCRIPTION)
A semiconductor chip stands upright on a black background. The square in the center of the chip is glowing white and hundreds of horizontal white lines are shooting outward from the square to the right across the black background. Text, An Evolving Industry.

(SPEECH)
RICHARD GU: Yeah, I'll try. So, semiconductor industry has evolved over time and continues to advance relentlessly downloads to now like 3 and 2 nanometers now. So those innovations and disruptions really breathe new areas and industries over the years, such as PC revolution, internet, data center, the 5G and mobile, and autonomous driving these days. So, what we are seeing is that the silicon and systems are increasingly converging together.

A great example of that is an electric vehicle, where at the heart of it is the chip that drives the key decision-making. But it's also a combination of hardware and software, mechanical and electrical, all coming together. More importantly, it is processing millions of signals and data to help make the most informed decision.

So with that in mind, the players that really have the breadth of the portfolio to support the merge of the electrical and mechanical will win out. We'll also add the tectonic shift of the new AI revolution, which could be, by far, one of the most transformative innovations in human history. So what it will do is to free up, really, the human beings from mundane and repetitive tasks and focus more on the value-added work. If you look at the next decade, the number of transistors is going to go 10x.

But the workload will grow even more, maybe 30 to 40x, due to the ever-increasing complexity of the chips. So ahead of us is a cliff of labor shortage as we simply don't have enough universities and colleges to produce that many engineers. So, what it means, that we have to rely on AI and automation to do the heavy lifting. And those are the real answers to help us bridge the gap.

And innovative packaging, as we just touched upon on 3DIC, will be critical for us to eke out more efficiency as Moore's law slows down. And the great thing about it is that you could put heterogeneous dies together to not only gain scaling and performance but also reap the benefits of faster innovation velocity since you don't need to redesign everything for your next gen product. You just need to swap out a certain piece of that and redesign.

(SPEECH)
AMANDA BOHN: Thanks for that, Richard. I think it's just going to be fascinating to see how AI is bridging that gap and helping that labor shortage, or if maybe the labor shortage will improve. And it's going to be really fascinating to see how these companies continue to evolve their business model in response to a changing market. So, given this constantly evolving industry, Richard, how do you see the technology evolving and maybe some of those risks and challenges that are on the horizon that these companies need to prepare for?

(DESCRIPTION)
A man stands at the bow of an origami paper boat. He is looking out into the distance over the water with a telescope. Text, What's on the Horizon?

(SPEECH)
RICHARD GU: Great question, Amanda. I think risks are an inherent part of any business. And with every risk comes great opportunities too. I'll give you an example. Like quantum computing could be a disruptive technology down the road. But it's still quite early days. And it's faced with big barriers of success, such as the extreme low temperatures it requires to operate and the superconducting.

Also tying back to the discussion between you and Chris on the geopolitical risk, which lurk around, building some redundancy and backup plan and diversification is fairly important. Cybersecurity is a risk that we all have to deal with in the digital age. And to fend it off requires a concerted effort from the industry, both from hardware and software standpoint.

(SPEECH)
AMANDA BOHN: Richard, thank you. I think we're focusing on very similar things. And your point about cybersecurity and the risks and threats as it relates specifically to chips, I feel like we could talk for an entire hour just about that topic.

(DESCRIPTION)
Implications for Life Sciences. An abstract graphic shows brain scan images arranged in a network of cubes and circuits.

(SPEECH)
But let's go deep on one area, if I could, and get specific around the implications for the life sciences sector. This industry is a key focus at Travelers.

It's one of our largest sectors within the practice. And we're really optimistic about the potential for growth in the sector, largely because of the explosion of telehealth capabilities during COVID, the application of AI in medicine, and then the innovations that we're seeing in the pharma manufacturing space around new drug delivery systems. So, Richard, from what we've discussed today, how do you see the technology evolving and the risks it presents, specifically for the life sciences industry?

(SPEECH)
RICHARD GU: Yeah, life sciences and digital biology are really exciting new frontiers, where technologies, especially AI, can be a great game-changer. I used to work for Pfizer years ago before coming to the Silicon Valley. And I remember those days when companies were pumping billions of dollars into developing a molecule and had to wait five to 10 years before any potential drug can be put on the market. And even that is a crapshoot.

So, what's super exciting about the current status is that, when you peel back the onion of a drug discovery molecular simulation and a chip design simulation, the fundamental algorithms are surprisingly similar. So, by combining math, biology, and computer science, we believe technology has a big role to play in the future of drug discovery. And this is still early stage, but very promising.

As Cadence, we acquired a company called Open AI about a year ago, which leverages simulation technologies to accelerate drug discovery. And it holds a lot of potential for this area. The way we are thinking about digital biology and life sciences is more like a three-layered cake. In the middle is the core algorithms of the drug simulation using math and computer science and biology. And at the top, we are leveraging AI to optimize and orchestrate the entire workflow. And at the bottom is what we call computational hardware, where parallel computing infrastructure, such as GPUs, can bring tremendous value in terms help us accelerate the computation.

(SPEECH)
AMANDA BOHN: A three-layer cake analogy, I love that. Thanks for that perspective, Richard. I totally agree with you. I think life sciences is an area that's really ripe for some disruption, especially around utilizing AI and bringing those life-saving drugs to market faster.

(DESCRIPTION)
Text, Risk Implications. A square embedded chip with much smaller chips at each corner, and with gold circuits coming out from each side.

(SPEECH)
It's really going to be fascinating to see how this all unfolds, not only for the life sciences sector but overall, the impact that AI is going to have, the hopeful success of the CHIPS Act, and whether the China-Taiwan tensions ease, and several other things. The risk implications from an underwriting perspective are really challenging to pinpoint right now, which is honestly one of the reasons I love underwriting in the tech and life sciences space. It's kind of a marriage of the future and the unknown with innovation all merged together.

We don't know what semiconductor devices will do in the future or where they will be. I think Chris has one in his dishwasher. I have a device in my cat. And I know Richard has a cell phone with several of these devices. But where will these devices be in 10 years?

Are they going to be in my shoes? Are they going to be in my sandwich that I'm going to eat for lunch today? In addition, they're getting even more powerful. And we've seen this with the rapid advances in AI chips as of late. So where will these devices be, how will we be using them, and what will their power limitations be?

As underwriters, we're always trying to predict the bad day. We want to be able to predict the loss so that we can plan for it, price for it, but more importantly, help mitigate it. And this can be challenging when you don't know what the future end use of the devices will be. Despite the risk implications being really hard to anticipate, continued partnership with our customers as they move along this exciting innovation journey, it's going to be more important than ever.

(DESCRIPTION)
Text, Final Thoughts. A collection of chips laid out on a surface, with one chip standing upright on one corner.

(SPEECH)
And now, Chris, Richard, we're on to final thoughts. Richard let's start with you. What are some of your final thoughts that you want our audience to know or take away from today's discussion?

(SPEECH)
RICHARD GU: Yeah, first off, I'd like to thank you, Amanda, and the Travelers group for having me. And I also wanted to thank Chris for his fantastic insights on the industry and his partnership. So, we're very fortunate to play a role in one of the most dynamic and critical industries for the global economy. To me, the risk-reward for the semiconductor industry is compelling, as you think about how the industry impacts the socioeconomic and technological contours for years to come. So, thank you, again, Amanda and Chris.

(SPEECH)
AMANDA BOHN: Thanks, Richard. On to you, Chris, with closing thoughts.

(SPEECH)
CHRIS MILLER: Well, I would just say, Amanda, that you opened by calling attention to the fact that we're all surrounded by dozens, hundreds, even thousands of chips everywhere we go, in our phones and our dishwashers and our automobiles. And that's been made possible by this extraordinary improvement in capability that Moore's law provided, as well as extraordinary declining cost, which has been great, enabling AI and all sorts of other capabilities. But it also does increase the risk because we become more and more dependent on semiconductors as every year passes. And we need to think harder than ever before about how we mitigate those risks to chip supply, given that our lives, our society, our economy, nothing can work without a vast supply of semiconductors.

(SPEECH)
AMANDA BOHN: Thank you for that, Chris and Richard. I think our discussion is proof that this industry is complex and presents challenges that business leaders and risk managers really need to prepare for. For those that are in the semiconductor industry, don't worry, there's plenty of help along the way.

Be sure that you're partnering with an agent or broker that really specializes in the space, in addition to an insurance carrier that does that, as well, like Travelers. You want your carrier to deeply understand the industry as well as the evolution that the industry has seen to ensure they can help you partner on those risk mitigation strategies. That will go a long way in helping your firm ensure that it can avoid the pitfalls and capitalize on those opportunities.

I want to thank Chris Miller and Richard Gu for your partnership today. Thanks for helping us decode the future. And

(DESCRIPTION)
Text, Thank you.

(SPEECH)
to those listening, we hope you enjoyed today's webinar and thanks for participating.

(DESCRIPTION)
Logo, Travelers. Text, Learn more. travelers.com.

2023 The Travelers Indemnity Company. All rights reserved. Travelers and the Travelers Umbrella logo are registered trademarks of The Travelers Indemnity Company in the US and other countries.