Semiconductors explained in 16 mins | Chris Miller

0:00 When I started my research on semiconductors,  I thought that because chips were everywhere, chips were easy to make, and because nuclear bombs  were only controlled by a handful of governments, they were hard to make. But what I realized  is it's actually the exact opposite. If you take nuclear weapons, that technology has  barely improved since the 1960s. But chips are everywhere because they're cheap and they're  tiny, and making things very inexpensive and very small is extraordinarily difficult, which is why  there's just a couple companies in the world that can do it. And it's done so not just for a couple  years, it's done so now for over half a century.

0:32 And that's why when you compare progress in the  computing industry to progress anywhere else, there's really no comparison. I'm Chris  Miller, a professor at The Fletcher School and author of, "Chip War: The Fight for  the World's Most Critical Technology." - [Announcer] Semiconductors 101 - Well, a chip is a piece of silicon, often the  size of your fingernail. And in it is carved thousands, or millions, in some cases billions  of tiny devices called transistors, which flip circuits on or off, on and off. And when they're  on, they produce a one. When they're off, they produce a zero. And all of the ones and zeros  undergirding computing, undergirding data storage, all of your Instagram likes, all of your text  messages, these are all just long strings of ones and zeros, which are created on the chip  by these circuits flipping on and off. There are a couple different categories of chips. Some  chips process data, other chips remember data, and a third category turns real world signals,  like audio or pictures into ones and zeros so that they can then be processed or remembered. And  so when we look at the world, we see pictures. But when a phone, for example, uses its camera to look  at the world, it takes in lots of rays of light, and then has to learn how to convert those into  ones and zeros that can be stored. And so there's very specific sensors for pictures, for sound,  for radio waves that use semiconductors to convert these real world signals into strings in ones of  zeros that can then be re-represented as pictures later on, for example, when you pull a photo up  on your phone. All of this is done by different types of semiconductors. So, generally, chips have  a foundation of silicon, but there are dozens of other materials that are layered on top to make  the transistors at such tiny scale. So a typical advanced chip could have several dozen materials.  The foundation is silicon, but there are many other chemicals involved in the process. Yeah,  it's true that sand is from silicon and so are chips, but the similarities basically end there.  The silicon that's used in manufacturing chips is among the most purified elements that we have. And  the reason is that when you're manufacturing chips with tiny transistors, you need to place almost  every atom perfectly to make those chips work.

2:41 Which means that if your silicon, or any  of the other materials that you're using, has even a single atomic impurity, it can cause  defects in the way your chip functions. And so the production of the silicon wafers that are  used in the chip manufacturing process requires extraordinary levels of purity. There's really  just four companies in the world today that are capable of producing silicon wafers at the right  level of purity at the scale that's required for contemporary manufacturing. The good news is  that there's silicon everywhere. It's one of the most widely-distributed elements in the Earth's  crust. The hard part is really the refining and the purification of silicon to make sure there  aren't any impurities that could disrupt the manufacturing process. So on top of your silicon,  you could have boron, gallium, gallium arsenide, lots of different chemicals that are used, and  every chip maker has its own proprietary process.

3:29 So we don't really know, inside of a typical  chip, what materials are used, because chip makers usually keep it pretty secretive. That's their  special sauce that lets them manufacture chips with the right level of capability. Now we're not  gonna run out of silicon, nor will we run out of the other materials that are generally used in  chipmaking. There are some concerns that certain materials are predominantly refined and processed  in a single country. So for some of the materials like gallium and germanium, China produces around  90% of those materials. So there's geopolitical issues that could interrupt supply, but it's not  gonna be that we're running out of the capability to produce them. I visited a bunch of chipmaking  facilities over the course of the research. The interesting thing though is that, when you  go inside one of these massive facilities, called fabs, what you find is that there are  huge machines and not much else. Because the manufacturing process has to be extraordinarily  automated because humans are way too imprecise for manufacturing at nanometer scale. And so inside of  a chipmaking facility, there are very few humans, and lots of big machines that, from the outside,  are impressive in their size, but you can't see what's actually happening because it's happening  at microscopic level. So there are a handful of companies that play a big role in the making  of the machines that make chips, a couple in the United States, one in the Netherlands, and one  other large one in Japan. Five companies play the dominant role in the manufacture of the machines  that make chips. And in some ways, it's actually harder to make the machines that make chips than  it is to make the chips themselves. Because these tools are among the most precise tools that have  ever been deployed. Just to give you one example, ASML, a company based in the Netherlands, produces  machines that are used in the manufacture of almost every high-end chip today. And these  machines are capable of manipulating materials at basically the atomic level to produce chips  with billions and billions of transistors like those that are inside of your phone or that are  used for training AI systems. So there's a pretty small number of companies that make chips.  And when you look at specific types of chips, you find that there's even more concentration.  The biggest chip maker in the world is the Taiwan Semiconductor Manufacturing Company.  When it comes to advanced processor chips, like the chips in your phone, or the chips in  your computer, TSMC makes around 90% of them. So they've got an extraordinary market share, and are  probably the most important semiconductor company, and arguably the most important company, in  the world, because the chips that they produce, we rely on for basically everything. There's  been a lot of consolidation in the chip industry over the past couple of decades, and it's been  driven by economics and by technology. Today, a single cutting edge chipmaking facility can cost  $20 billion, one of the most expensive factories in all of human history. And so there's just a  couple of companies that can afford to put up that sum of money on a regular basis to build more  and more cutting edge facilities. And to make that work financially, you've gotta produce a ton of  chips. And so there are huge benefits that accrue to the largest firms. The more chips you produce,  the more your cost structure makes sense, and the better your technology gets, because you learn  from every chip you manufacture, you gather data from it, and you tweak your manufacturing process  to make sure you've got fewer and fewer impurities at every step. And so TSMC is both the world's  largest chip maker, but it's also the world's most advanced, precisely because it gathers more  data than anyone else. Because chipmaking requires ultra-purified materials and hugely complex  equipment, there's not a single company that can do it on its own. Everyone requires a set of  partnerships with supply chain providers to give them the materials, and the intellectual property,  and the software and the tools that they need to produce advanced chips. And so if you take for  example, the primary processor inside of your smartphone, it was probably made in Taiwan, but it  was made in Taiwan using chipmaking tools from the Netherlands, and from the United States, and from  Japan. It was produced using chemicals from Japan, and then often assembled and packaged in Malaysia  before ending up inside of your smartphone. And that's typical. A typical chip requires components  and materials sourced from dozens of different companies because the process is simply too  hard for any one company to do on its own. So, a nanometer is a billionth of a meter, and chips  today are measured in nanometers. If you look at the chip inside of your phone, for example,  and try to measure the size of the transistors, of which there will be billions on your  smartphone chip, each one of these will be measured in a handful of nanometers. And so  that makes them only slightly larger than atoms, smaller than any sort of living thing, far smaller  than a bacteria, smaller than a mitochondria, half the size, for the most cutting edge  transistors, of a coronavirus. There's basically nothing we manufacture at such tiny  scale as we do with semiconductors. Every year, we make more transistors than we've made all other  goods combined in all of human history. And in fact, nothing else really comes close. A typical  smartphone chip could have 10 billion transistors just in the main processor chip. A big data center  run by Google or Amazon Web Services would have more transistors than you could plausibly count.  We know that we make more transistors than there are cells in the human body, for example. We  don't even know how many we make in aggregate, because there are just so many. Moore's Law  predicts that the number of transistors per chip, and as a result, the computing power per chip  will double every couple of years. And that's been empirically true since the 1960s, which means  that the capabilities of chips have gotten vastly better, and continue to get much, much better at  a faster rate than anything else. So I like to think, for example, of airplanes to illustrate  the difference. If airplanes doubled in speed every two years from the 1960s up to the present,  we'd be flying faster, literally, than the speed of light. But chips have done that. Chips have  increased in that capability because the scale of the transistors has shrunk to the level that  today we're manufacturing them smaller than even viruses. And that has enabled the explosion of  computing power, both in terms of the computing capabilities in high-powered data centers or in  your phone, but also the application of computing to all sorts of devices. 'Cause today, there's  computing everywhere. It's in your dishwasher, it's in your refrigerator, it's in your coffee  maker, it's in your car. And it's possible to put computing everywhere because today it's so cheap,  we can produce it almost for free. And that has enabled the application of chips to all sorts of  different devices. To understand the change and the rate of innovation, in the 1950s, you could  hold a single transistor in your hand. Today, you can hold 10 billion transistors in your hand  in a chip that's the size of your fingernail. And that's not an expensive chip, that's a chip that  often will just cost $50 or so. So the rate of shrinking transistors, as well as the rate of  decline in their cost, has been unparalleled in any other segment of the economy. So before  transistors, computers used vacuum tubes, which are sort of light bulb like-devices that  would turn on and off, on and off to produce the ones and zeros. And they were cutting edge for  their time, but they had huge inefficiencies. They wasted a lot of heat, for example, they worked  pretty slowly. And they also, because they created light, attracted moths, and so computers had to be  regularly debugged in the early days of computing, which meant removing moths from the lights that  they were attracted to. You can see why it was hard to scale that up into a 10 billion unit  system. You know, I think the transistor is the key reason why we've been able to scale down.  There's really nothing else, if you look all across the economy, that has shrunk in size and  shrunk in cost at that level. And it's done so not just for a couple years, it's done so now for over  half a century. And that's why when you compare progress in the computing industry to progress  anywhere else, there's really no comparison. Well, Moore's Law is not a law of nature, it's not a law  of physics. We wish it were, because then we could rely on it to keep delivering advances far into  the future. But it's really a law of economics. It says that, if you're able to find a way to shrink,  shrink your transistors smaller, then you will be able to find a larger market as well. And that  has incentivized huge investments in shrinking, in improving manufacturing processes, and making  chemicals more purified to enable it, which has sustained this rate of advance. And if ever it  turns out that the economics are on Moore's Law break down, the technology will immediately break  down as well. Thankfully, the good news is that, right now, we're seeing a new wave of excitement  about ways you can deploy computing, which has led to a surge of new investment into AI, but also  a surge of new investment into semiconductors, because it's now clear that if we can shrink even  further, we'll enable a whole new era of advances in artificial intelligence that rely on even more  computing than we've been able to muster thus far.

12:02 You can define Moore's Law in a bunch  of different ways. Is it based on the 2D size of the transistor, or the 3D size of the  transistor? Is it based on the processing speed that comes out of it? And I think there's a lot  of people in the industry that are trying to sell a certain chip with given characteristics  that have an incentive to say Moore's Law, based on the other characteristics, has come to  a halt. If you look at the rate of increase of machine learning semiconductors, for example,  chips that are optimized for AI capabilities, they've been doubling in their capabilities every  two years for the past decade or so. In other words, exactly what Gordon Moore predicted when  he set out Moore's Law in 1965. And so my view is that when you zoom out and look at the rate of  technological progress, there's really no slowdown that's happening. When I started my research on  semiconductors, I thought that because chips were everywhere, chips were easy to make, and because  nuclear bombs were only controlled by a handful of governments, they were hard to make. But what  I realized is it's actually the exact opposite.

12:59 If you take nuclear weapons, that technology has  barely improved since the 1960s. It's so easy to make nuclear bombs, even the North Koreans can  do it. But chips are everywhere because they're cheap and they're tiny, and making things very  inexpensive and very small is extraordinarily difficult, which is why there's just a couple  companies in the world that can do it at the cutting edge. And the reason is that it's brutally  expensive, and it requires manufacturing processes that get better, and better, and better every  single year. And so if you're trying to catch up to the cutting edge in the chip industry, you're  not trying to catch up to a static cutting edge, you're trying to catch up to a cutting edge that  is racing forward at the rate of Moore's Law, doubling every two years. And so it's a race  between companies, but it's the fastest race humans have ever undertaken, which is why it's  extraordinarily difficult to reach the cutting edge. A couple years ago, it became harder to  shrink transistors in two-dimensional format.

13:50 For a long time, chips were made, they were just  described as planar chips, chips in a plane, in which all the transistors were on the same  level. Now we've started making transistors that have three dimensions, because we're  learning to stack them on top of each other to package more of them together in a way that  produces more computing power. And so one of the key trends over the next couple of years is  going to be more 3D construction of groups of transistors, which will enable more of them to  be crammed into a small amount of space. So the machines that make chips are extraordinarily  precise in their manufacturing. For example, there are tools that can lay down thin films  of material that are just a couple of atoms thick with basically perfect uniformity. And to  pattern the transistors on a piece of silicon, you use a tool called a lithography tool. And today  there's one company, ASML, of the Netherlands, which makes most of the world's lithography  tools. And for the most advanced chips, these tools can cost $350 million a piece for a  single tool. And they cost so much because they require some of the most precise components  ever used, like a mirror that's the flattest mirror humans have ever made, a laser that's the  most powerful laser ever deployed in a commercial device, and a ball of tin that falls through  a vacuum that is struck twice by that laser, explodes into a plasma measuring 40 times  the temperature of the surface of the Sun, and this plasma emits light at just the right  wavelength, 13.5 nanometers, to be bounced off the mirrors in exactly the right geometry  and land on your chip to carve the transistors into the silicon. It's the most complex and  expensive machine that humans have ever made, and it's required to make all of the most advanced  chips. Today, there are just three companies capable of producing cutting edge processor  chips, the types of chips that go in phones, or computers, or are used for AI. And it used  to be a larger number of companies that could produce at the cutting edge, but it's shrunk  into three, and might in the future shrink only to two for two reasons. First, the expense  is extraordinary. $20 billion per facility is a level of spending that many governments can't  afford, to say nothing of companies. But second, the scale required to manufacture efficiently  is vast. And that means that the benefits accrue to the largest firm. And in this case,  that's TSMC, the Taiwanese firm that's at the center of the chip industry. That's why they  manufacture on 90% of the most advanced chips, because they're cheaper, and they're better than  their competitors when it comes to manufacturing.