Instant Genius - The birth, evolution and future of microchips

Episode Date: July 12, 2026

It’s a good bet that if most of us were to glance around our living rooms or offices, we’d quickly spot all manner of electronic devices that are powered by microchips. Since their introduction in the 1950s, chips have become one of the fundamental building blocks of the various digital technologies we rely on in our daily lives. What impact has this had on the development of technology, economics and global politics over the last several decades, and what roles will chips continue to play in our global future? In today’s episode, we’re joined by Prof Rakesh Kumar, John Bardeen Faculty Scholar in the Electrical and Computer Engineering Department at the University of Illinois, to talk about his latest book, The Chip Age – How Chips Shaped Our Past and Will Define Our Future. He tells us about the key role the space race played in the rapid development of microchips, explains how the production of chips involves collaboration between countries in almost every part of the world and talks us through some of the various economic, political and environmental issues that may influence the future of this vital technology. Learn more about your ad choices. Visit podcastchoices.com/adchoices

Hello and welcome to Instant Genius, a bite-sized masterclass in podcast form. Every Monday and Friday, you'll hear world-leading scientists and experts talking about the most fascinating ideas in science and technology today. I'm Jason Goodyear, commissioning editor at BBC Science Focus. It's a good bet that if most of us were to glance around our living rooms or offices, we'd quickly spot all manner of electronic devices that are powered by microchips. Since their introduction in the 1950s, chips have become one of the most fundamental building blocks of the various digital technologies we all rely on in our daily lives.

But what impact has this innovation had on the development of technology, economics and global politics over the last several decades, and what roles will chips continue to play in our global future? In today's episode, we're joined by Professor Rakesh Kumar, John Bardeen Faculty Scholar in the electrical and computer engineering department at the University of Illinois to talk about his latest book, The Chip Age, how chips shaped our past and will define our future. He tells us about the key role the space race played in the rapid development of microchips, explains how the production of chips involves collaboration between countries in almost every part of the world, and talks us through

some of the various economic, political and environmental issues that may influence the future of this vital technology. So welcome to the podcast. Thanks so much for joining us. Thank you for having me. So today we're talking about your latest book, The Chip Age, How Chips Shaped Our Past and We'll Define Our Future. So everyone listening is going to have heard of Michael Chips before. But I bet most people don't actually know exactly what they are. So I think that's the best place to start. So we're talking about chips. What exactly do we mean? So when we are talking about chips, we are essentially talking about these really fingernail or

posted size devices that sit at the heart of most of your electronic devices. So let's say you have a television, you have a computer, you have a missile, there are these devices that sit at the heart of these systems that really powered the whole computing, the whole data processing, the whole storage that's needed in order to enable these systems. So that's what we are talking about here. Yeah, so let's have a look into that a bit further then. What are they actually made from? Broadly speaking, a chip is made out of silicon. So, you know, the good old silicon that you find in sand? So, in fact, the way you start making chips is you start with sand. And then through a set of a bunch of processes, you're converting them into these things that really are powering our civilizations. That's quite incredible.

So let's go back to the start of the story then. You know, when were they first invented? So chips have a fascinating story. You know, once people learned how to build transistors, it became very clear to them that if you could pack these transistors onto one substrate, one device, it can do pretty incredible things. So it's not some kind of certain ability. You know, a lot of people realize that they need to build something like a chip where they can pack a lot of transistors on the same substrate.

In fact, there was a British gentleman who conceived of the idea of a chip, and he worked on different prototypes. In fact, he even created mock-ups of how, chip would look like. In reality, the first chip was built by this gentleman named Jack Kilby at Texas Instruments in the US. Interestingly, even the chip that he built wasn't quite mass manufacturable. So it's something that he built something where you could put a bunch of transistors on a substrate and you could get it to do something. But the way he built it, it wasn't something that you could stamp over and over and then sell it, build it at high volumes or

or sell it at low cost. So there was this other person, Robert Noyes, who finally figured out how to build it in a way that you can manufacture these at high volumes and sell them at a low enough price that they could have an impact. Yeah, so we've mentioned transistors there a couple of times. So just so we all get up to speed without doing this sort of electronics 101, you know, what's a transistor? What does it do within a circuit? So think of a transistor as essentially a switch. So think of it as some. something that essentially switches between zero and one. And today, all of computing, even when you write software, essentially all of this gets

translated into a set of control signals for these little switches, the transistors, that convert your intent, the programmer's intent, into something that essentially drives different kinds of circuits using these zeros and ones. So transistors are these devices that are the heart of these chips. And then one of the innovations of the chip was that we can put a lot of these switches, these one and zero devices, into a small space. That's right. In fact, it's one of the most incredible thing about the chip industry.

So today, on a single chip, you can have billions to hundreds of billions to, in some cases, trillions of these switches, which is quite amazing. So imagine trillions of these devices on a single chip. So we're talking about the development of chips, so from the idea to the way that they're now, as you mentioned earlier, they're literally, you only need to look around the room and that you're sitting in and they're everywhere. So when did they really start picking off, you know, what were some of the key events that when, you know, like you say, the chip age, when the chip age began.

So while that potential was cleared, you also needed a lot of investment into this new technology in order to really make an impact. Thankfully, that was the time when the U.S. was working on its Minuteman missile program. Similarly, the U.S. was also working on this Apollo program to go to the moon, and both of these programs would benefit greatly from the chips. Government realized that immediately. So, U.S. government was actually the first large customer of the chip technology. And this was brilliant for this technology, because since the technology was in its infancy, it needed

a lot of initial investment and it got that investment from the government. So it starts coming back up to modern times. So from these early days, do we know how many chips are currently produced, say, in a given year, in one year? Globally, more than one trillion chips are shipped every year. In fact, the annual sales of chips would exceed a trillion in a couple of years. So having said that, that's an awful lot of manufacturing going on there. So how to make these things and at such volume as well. So chips are made in these factories that we often call fabs, fabrication plants or foundries. So let's step back and think about the process right from the beginning.

So essentially you start with a certain kind of quartz, which is essentially sand. Then you melt this quartz and then now you have this molten material sitting in a cylinder. Then what you do is you grow a giant crystal out of this molten sand effectively. So now you have this thing called, you know, single crystal silicon. Once you have this giant cylinder of a single crystal silicon, you essentially then create a bunch of disks, thin disks out of these cylinders. So essentially you're slicing the cylinder over and over and over to create a bunch of thin disks. You call these thin disks, vapors. Then what you do is you polish these wafers so that they are the smoothest disks you'd have ever seen.

So once you now have these highly smooth discs, the wafers, then you start injecting it with different kinds of chemicals and materials in order to achieve certain kinds of properties, electrical properties into these vapors. Once you've done that, then there's this really fascinating process where you have some artwork or what we call chip design. that's essentially in a mask. Think of a mask is another disk that has a set of patterns sort of cut out on it. What you do is you place the mask on top of this wafer and then you shine this light, ultraviolet light through this mask. So essentially the light goes through this mask onto the wafer and then etches this artwork or the pattern on top of this wafer.

That's how you essentially place a chip design or put a chip design on top of a wave. wafer. So you do this over and over using multiple masks. So now at the end of this process, you have a wafer that has chip design imprinted on it and, in fact, multiple copies of it. Okay. So once you have that, then you essentially dice these chips. So essentially you cut these chips out using some kind of saw, for example. So you now have a bunch of these chips that you've cut out of this wafer. What you do then is you package these chips. So chips by themselves are very fragile. So you package these ships into sort of some kind of covering, if you will, for protection and for integrity, et cetera.

And then these packaged chips, you have to test for correctness. So you do different kinds of testing to see whether the chips work the way they were intended. And once you have these tested package ships, then you can put them in the different kinds of applications that you have in mind. Yeah. So sort of sticking with that, of course, we all know the huge impact that. chip technology has had on all aspects of our life. But a lot of people don't think about, you know, as we just talked about there, how are they made? But where the materials come from and the impacts that that can have, which you talk about a lot in the book. So where do we get

traditionally and currently, where do we get the materials to make all of these chips? Yeah, so that's a great question. So chips are these fascinating devices that require almost every element in the periodic table. So imagine all these different kinds of materials that you need in order to finally deliver a chip. So these materials are sourced from all over. In fact, chips are one of the most collaborative exercises that humanity goes through. So let's step back again. So to build chips, you need different kinds of input materials, silicon, different kinds of chemicals that you're using to inject the desirable properties into the wafer. You need different kinds of manufacturing tools.

You need different kinds of design tools. You need different kinds of testing tools. All of these get done in different parts of the world. So a lot of manufacturing happens in Asia. A lot of design happens in the U.S. A lot of equipment gets built in Europe. A lot of chemicals that you need to polish the wafers to deliver the right properties onto the wafer gets done in Japan.

So this is really an exercise that the entire world participates in. So you talk about the various impacts that sourcing these raw materials can have, as you mentioned, there globally. So what are some of the key sort of stories and scenarios and perhaps even concerns there? So when it comes to sourcing different kinds of input materials or tools or equipment for the chip building process, there are several issues that come up. So let's start with materials.

So one, a lot of materials get produced in regions that are, frankly, conflict zones. So the availability of those materials is strongly dependent on these different geopolitical events that are going on. Similarly, some materials are sourced from regions of the world that don't have very good human right records. So there are reports of child labor being used in producing some of the materials that ultimately make their way into the chips.

Similarly, there are these different kinds of choke points. So there are parts of the chip supply chain that are produced by one or a couple of countries. So these countries essentially can use that as a geopolitical leverage and really hold the world hostage when it comes to the production of chips. So there are all these different kinds of concerns that exist. around chip production. How about things like impacts on the natural world or the environment? Because obviously we're getting these materials from somewhere. That's right.

So chip production itself can have significant impact in terms of the need for water, the need for energy, and that has different kinds of implications. So consider the US scenario. In the US, a lot of chip production takes place in areas that often see droughts that really lack water. So when you're drawing a lot of water to produce chips, that really has an impact on the water availability of your local community. Similarly, a lot of chip production involves different kinds of emissions.

So you're emitting different kinds of gases into the environment. You're sometimes polluting the water that you're using. So you have environmental impact, impact on the quality of air, impact on the quality of water. Furthermore, chip production requires. requires a lot of electricity. In fact, so much so that in countries like Taiwan, where chip production is a major economic driver, it has real impact on the quality of power that they see.

In fact, Taiwan has seen multiple blackouts that can be directly attributed to chip production. So there's clear impact on electricity, on the quality of air, the quality of water. These all have to be kept in mind. So it's not only the production of chips, it's also the usage of chips. Some of the AI data centers today use chips that overall can consume gigawatts of power. That's a lot of power.

So when the chips are consuming so much of power, that again has impact on your electricity bills, that again has impact on the reliability of the electricity grid, the quality of power that you get, et cetera. Because chips need a lot of power, they need a lot of water to cool these chips. And we go back to the same questions. Well, is this water getting drawn out of the portable water source? Is this affecting the volume and the quality of water in the community? What is the power source that powering all these ships?

Are these ships getting powered by coal plants? Are these ships getting powered by natural gas? Or are these ships getting powered by renewable energy? So even when it comes to the usage, ships are directly or indirectly responsible for a lot of water usage, a lot of electricity usage, they have real footprint on the environment. Yeah, so you sort of touch on there talking about these large data centers and the use of chips and the sort of growth of AI.

And often people talk about things called GPUs. So could you just tell us what they are and how they fit into this picture, please? So there are different kinds of chips. Different kinds of chips are good for different kinds of applications. AI applications particularly benefit from this specific kind of chip called GPU or graphics processing unit. Interestingly, GPUs were invented, frankly, to let you play video games for the computer graphics.

And at some point, it became very clear that the same kind of chips are very good for AI applications. There was a sort of a self-reinforcing cycle that got created between AI and the GPUs. As AI applications grew, the GPU industry grew, and as the GPU industry grew, and as the GPU grew, a lot of AI applications got built on top of it and AI itself grew. So we are in this interesting cycle where GPUs are critical for evolution of AI, and the evolution of AI means that we need more and more GPUs. Yeah, so you mentioned there, like the huge amounts of power that they use and the amount of calling that they need when they're operating. So, you know,

you think, well, that's a huge problem. So are there any ways that people are working on we can perhaps provide a solution to these problems. So maybe make chips that operate at lower temperatures, for example. So this is a big problem. And indeed, there's a lot of work being done on how do you reduce the amount of energy that GPUs consume. So there are all kinds of techniques people are looking at. So the techniques all the way at the algorithm level.

So people are trying to design better AI algorithms that can deliver the same performance at much less power. People are trying to build better chips, more energy-efficient chips, where they're using better circuits, better cooling technologies, all of that, so that the energy footprint is reduced. However, I must warn that increasing the efficiency has not always delivered lower overall energy. It's called Jevon's paradox. Humans, we are sort of greedy. If you make something more efficient for me, there's a much higher likelihood that I'm going to use more of it.

And that's what's been happening. In fact, when you look at the energy efficiency of chips, it has really gone up over the years. Today's chips are orders of magnitude more efficient than chips that existed only a couple of decades back. But we are now finding new ways to use these chips, new applications to power using their chips. So we are just using more and more chips,

and that's increasing the overall energy footprint. So while it's commendable that there's work being done on reducing the energy of chips, that is this other factor that's almost psychological, that's almost tied to how we think as humans, that's about Jevon's Paradox, where we want to use more of something that's more efficient. And I don't know what we do about that. You just reminded me of something that I was going to ask then. So we, you know, as we've been talking about, chips have evolved quite a lot over the, you know, the last several decades or whatever. But whenever this topic comes up, people often bring up something.

something called Moore's Law. So what exactly is that? And, you know, where are we on that curve now? So Gordon Moore was one of the founders of Intel. And he made this observation that, and this is 1965, this is long back. He made this observation that the number of transistors on a chip was essentially doubling every 18 to 24 months. And once he made this observation, people took notice.

And it sort of became like a goal where the entire industry worked towards making it true. So it's not really a law. It was an observation that was made, and then the industry decided to take it as a goal, and then it became sort of a self-reinforcing thing, where the entire industry would come together and would try to deliver miniaturization of these switches or the transistors in such a way that you can double these every 18 to 24 months, and that's the cadence we were on over and over. Now, whether Moore's Law still continues is a matter of great debate.

A lot of people believe that Moore's Law has slowed down, but it depends on how you look at it. If the question is, are we packing transistors per unit area, according to Morsella today, the answer is no? On the other hand, if the question is, is the number of transistors on a chip still following Moore's law? The answer is yes. Now, the reason why both are true is because increasingly we are now being more and more creative in terms of how we are putting together chips,

how we are putting together the transistors on chips. For example, today, we are going in the Z dimension. So think of it as building a city. So, you know, initially, let's say there's a lot of land. They're not a lot of people. You bring people in, and then people build houses, and then people grow outwards. And at some point, the city is filled with people.

Then how do you, you know, accommodate more people? Well, one way in which you accumulate more people is by building skyscrapers. So you go in the Z dimension, the third dimension. And that's what we have started doing with chips. So we have started putting transistors in the third dimension. So transistors are now layered on top of each other, and that's how the total number of transistors on a chip is still falling more slow,

even though the density at which we are packing these transistors is no longer falling more slow. There's another trick that we are playing now. There's this very interesting phenomenon around chips, where if you make a chip too large, too large, the likelihood of it not working shoots up. So think of it as a bunch of dust particles in the factory where you're building the chips. The bigger you make the chip, the higher the likelihood that the dust particle is going to hit your chip and kill it. So you don't want to make chips bigger than threshold.

So essentially, the way chips historically worked is initially the size of chips kept increasing, and at some point you stopped increasing the size of the chips. So once you stop increasing size of chips, the only way in which you pack more and more features by increasing the density at which you're packing the transistors, by essentially making the transistor smaller. But at some point, it became harder and harder for us to make these transistors smaller and pack them in this sort of small area of the area of the chip. One new trick that we are playing now is packaging multiple of these chips together in a way

that these chips together look like one large chip. So we are taking these chips that we didn't want to grow the size of because of these concerns about reliability, but then we are taking multiple of these chips together, packaging them together, and now they look like much bigger chips.

So that's another way in which we are effectively increasing the number of transistors a single chip can have because a single chip now consists of a bunch of these little chips that we sometimes call chiplets. So going back to the very start, we talked out the explosion of conventional chips and the various economic and political forces that drove that. Do you think, you know, we need something like that with these quantum chips to really get them over the line? So one of the reasons why you see a lot of geopolitical tensions or geopolitical maneuvers around

chips, why you see a lot of investments around chips, is because it's not only that chips power today's applications, but chips are going to underlie some of the most strategic technologies of future. So consider AI. As you know, a lot of people believe that the country or the region or the set of values that wins the AI race is the one that's going to dominate the world. So in order to win the AI race, it's critical to have access to the most advanced ships that these AI applications run on top of. So one of the reasons why countries want access to these advanced ships or want sovereignty over these advanced ships is so that they can compete in this AI race so that they can get the geopolitical and economic spoils of AI going forward.

It's not only AI that's the strategic technology that chips are critical for. Under the strategic technology that chips are critical for is quantum computing. Quantum computing, again, is one of those technologies where the ceiling is very high. You can potentially enable applications and systems and get military. and economic advantages that would be very difficult to beat. Chips underlie the development of quantum technologies. So again, chips are critical to a technology like that. Biotechnology.

Again, there's enormous potential around biotechnology, both in terms of economic benefits as well as military benefits. Again, chips are critical to the development of biotechnology. Renewable energy, being able to lead and win the renewable energy, also requires access to chips, mastery over chips. So that's another reason why chips have become the spawn in the geopolitical game. So we have to recognize that chips are not only important to power today's applications. They are also critical to the strategic technologies for the future.

And that is the reason why we see a lot of attention around chips, especially when it comes to geopolitical games. Thank you for listening to this episode of Instant Genius, brought to you from the team behind BBC Science Focus. That was Professor Rakesh Kumar. To discover more about the topics we've just discussed, check out this book, The Chip Age, how chips shaped our past and will define our future. If you liked what you just heard, then please do consider subscribing to Instant Genius on your preferred podcast platform. If you'd like to see our guests and hosts in person, then why not check out our YouTube channel?

at ScienceFocus. The current issue of BBC Science Focus magazine is out now. Pick up a copy wherever you buy your favourite magazines or download us on your app store of choice. You can also find us on Apple News or online at sciencefocus.com.