What are semiconductors, and why are they vital to the global economy?

This transcript has been generated using speech recognition software and may contain errors. Please check its accuracy against the audio.

Robin Pomeroy, host, Radio Davos: What’s the thing that is vital to keep the global economy flowing, is the size of a fingernail, but costs billions of dollars to produce - clue: you may well be holding at least a few in your hand right now.

Chris Miller, associate professor of international history at the Fletcher School, Tufts University, Boston: If you take your smartphone and open it up, you'll find many different chips inside: chips running the operating system, chips managing the Bluetooth, the Wifi, the audio, the camera, the connection with the cell network.

It's not only devices that you think of as having computing power. Take a new car, for example, that car will on average have 1,000 semiconductors

Robin Pomeroy: Welcome to Radio Davos, the podcast from the World Economic Forum that looks at the biggest challenges and how we might solve them.

This week: they’re everywhere in the things we use, and are at the heart of the artificial intelligence revolution, so just what are semiconductors and why are they so vital to the global economy?

Chris Miller: Today, trade in chips is a fundamental pillar of the international trading system. It's actually so important that China spends as much money each year importing chips as it spends importing oil.

Robin Pomeroy: This historian has written a book that charts the rise and rise of chips and explains why the vast factories they are made in are so expensive and complex.

Chris Miller: A new fab will cost $20 or $25 billion dollars, the most expensive factories in human history.

Robin Pomeroy: Subscribe to Radio Davos wherever you get your podcasts, or visit wef.ch/podcasts.

I’m Robin Pomeroy at the World Economic Forum, and with this explainer on the vital importance of chips…

Chris Miller: We’re surrounded by thousands and thousands of chips, most of which we are only vaguely aware of because we never see them, they’re buried deep inside of electronic devices.

Robin Pomeroy: This is Radio Davos.

Welcome back to Radio Davos, we’re weekly again now, after the summer break, and we start the new term with a look at something small but hugely important, both in terms of our everyday lives, but also in the grand scheme of global trade, economic and geopolitics.

Semiconductors make the world go round, and the most cutting-edge versions are necessary to propel the advances in artificial intelligence that everyone’s been reading about - and listening about - if you are a subscriber to Radio Davos - please do check out our series on artificial intelligence from the last couple of months.

Yet, the global semiconductor industry is, let’s say, complicated. Its early formation and current dynamics are brilliantly detailed in a book called “Chip War,” written by this week's guest Chris Miller, he's an associate professor of international history at the Fletcher School at Tufts University in Boston in the USA.

Chris Miller spoke with us about his extremely-well-timed book, what makes chips so vital and difficult to manufacture, and why meddling with supply chains for geopolitical reasons could have unintended consequences.

Chris Miller spoke to my colleague John Letzing - here’s John, starting with the basics:

John Letzing: For somebody who maybe only has a vague notion of what a semiconductor is, which is probably a lot of people, can you tell us a little bit about what is a chip and what does it do.

Chris Miller: A chip is a tiny piece, of silicon in most cases, often the size of your fingernail and on it are carved millions or nowadays often billions of tiny transistors, which are little electric switches that flip on and off.

And when they turn a circuit they produce a one, when they turn off they produce a zero. And all of the ones and zeros undergirding all digital computing, in your phone or PC or data centres, is just millions and billions of these tiny circuits flipping on and off, and so the process of chipmaking is just the process of carving many of these microscopic circuits into tiny pieces of silicon.

John Letzing: You know, I think the average person also might have a conception of maybe a single chip in their phone, for example, whereas that's not. Can you tell me a little bit about the abundance of these things?

Chris Miller: Today, almost anything with an on-off switch, except for the most simple lights, have at least one and often dozens or some cases hundreds of semiconductors inside.

So certainly if you take your smartphone and open it up, you'll find many different chips inside: chips running the operating system, chips managing the Bluetooth, the Wifi , the audio, the camera, the connection with the cell network, specialised chips to optimise the speed at which your phone processes video. There are dozens of devices inside of a typical phone.

It's not only devices that you think of as having computing power because today almost every type of device is being connected to the internet and given some sort of basic computing. So your refrigerator, your microwave, your coffee maker, they all have chips in them too, sometimes quite simple chips, but often increasingly sophisticated chips.

If you take a new car, for example, that car will on average have 1,000 semiconductors inside, some of them doing things like making the windows move up when they press the button. But others are much more complex, managing any sort of autonomous driving systems you have, communicating with the cellphone network.

And so we're surrounded by thousands and thousands of chips, most of which we are only vaguely aware of because we never see them. They're buried deep inside of electronic devices.

John Letzing: You make this really nice comparison in your book about what some countries, maybe a lot of countries, spend on something like oil relative to chips. And I guess in terms of a basic building block for a growing economy, where do chips rank?

Chris Miller: Well, semiconductors are one of the most widely traded goods internationally, and that's especially true and not only about chips themselves, but about the devices they make possible: phones, PCs, servers. And so today, trade in chips is a fundamental pillar of international trading system. It's actually so important that China spends as much money each year importing chips as it spends importing oil, which is sort of an extraordinary fact.

We're used to thinking about oil being a strategic commodity, but if you're sitting in Beijing, you're just as worried about the price of semiconductors, your access to them, as you are worried about your ability to import oil.

As we put more and more chips in more and more devices, it's become increasingly critical for companies to access the chips they need for manufacturing. In the past couple of years, when there was a semiconductor shortage globally illustrated just how dependent the world's manufacturers had become on access to semiconductors.

And so the auto industry is the best example of an industry that across the world faced a deficit of chips that they needed during the pandemic. In the US, Japan and Europe, auto firms had to let cars sit unfinished because they couldn't acquire often a handful of semiconductors that they needed for critical features. And the global auto industry was estimated to have lost several hundred billion dollars in lost sales because of shortages of semiconductors.

And that illustrates just how important chips have become. I think until recently, auto CEOs didn't really think of their companies as major consumers of semiconductors. They thought their critical components were the engine, the bodies of cars, or even the software systems. But actually that the chips today are one of the most critical components for cars.

If you're looking to the future, actually, cars are another case study of how this is continuing to intensify in its importance. A new car will have of a thousand chips inside of it, as I mentioned, but an electric vehicle will have several thousand dollars' worth of semiconductor components inside of them because you need chips not only for the usual features like the windows and the driving, but also to manage electric vehicle powertrains.

So as we project forward, we're seeing more and more chips put in more and more types of devices, which means that company executives in industries that seem far from the tech sector are having to spend more and more time thinking about how to get access to the chips that they need.

John Letzing: It's fascinating, despite this growing reliance, you describe a global industry, global supply chains consisting of, I think you refer to them as, choke points. I mean, is it fair to say that maybe this industry isn't quite as globalised as some people might imagine?

Chris Miller: Well, it's an interesting paradox.

On the one hand, the industry's consumers are global. Buyers of chips for smartphones or autos are in almost every country. And the industry has only been able to scale the way it has because it has access to this global market. So almost every major chip maker that tries, and in most cases succeeds, to sell to an international market.

But the production process is actually shockingly concentrated for many types of advanced chipmaking equipment or software or manufacturing processes, there's just a couple of companies in the world capable of undertaking the type of production that you need and for many critical components actually just one company that has the know how to do cutting edge production.

So you've got a global market, but you've got highly concentrated production, and that does create the potential choke points that you're referencing.

The other factor that people often don't fully realise is the diversity of types of chips, and that's great because it means we've got many different types of chips for many different applications, but it also means that you can't very easily swap in one chip for another.

And so when you think about China importing as much chips as it does in oil. Oil is basically something you can easily swap in and out one barrel for another. There's a bit of variation between the types of crude different countries produce, but not much. But there's huge variation in chips, and so often it's the case that just a single facility can produce the type of chip that you need. It would take months and huge sums of money to bring up production in a different facility. So if one plant gets knocked off line it cause huge issues or industries that are downstream.

John Letzing: A big part of the appeal of this book, I think, for people, is this fully fleshed out history that you provide with this cast of characters. And one striking thing I think is the way these early players like Fairchild Semiconductor and Texas Instruments, they relied on these things like the Apollo program and nuclear weaponry to grow. Is it fair to say that geostrategic rivalry, in that instance Cold War rivalry, fuelled this industry from from the beginning?

Chris Miller: I think that is fair to say. And here, too, it's a complex story. Because, on the one hand, there's no doubt that the the U.S. government's willingness to spend vast sums of money trying to buy advanced chips for military equipment was a key driver of early growth. Throughout the 1960s, which was the first decade when you could purchase chips, most of them went to either defence or to space uses like NASA's.

But it's also the case that the industry only scaled because it quickly found out how to take technologies that were developed for defence applications and turn them to a vast civilian market.

And so today around 98% of chips that are produced go to civilian use cases rather than defence use cases. And so the companies and the countries that have succeeded have done so by taking these technologies that are often developed with government help and turning them into mass market products.

And ultimately the mass market is where the money is. And it's the civilian market that has let companies build the capacity to scale their production and therefore drive down costs in a way that have made chips ubiquitous.

John Letzing: Another thing you describe in terms of this sort of Cold War rivalry is the way that the US was able to get ahead, not least because of the way it was able to spread out R&D funding among all of these different allied countries in East Asia and elsewhere, whereas the Soviet Union was sort of doing it all in-house or trying to. Does that speak to the present era in a way, and this notion that we can just re-shore all of this stuff?

Chris Miller: I think one of the things you see from the very earliest days in the chip industry is that no country was doing it alone.

So you go back to the earliest days of Fairchild Semiconductor, one of the earliest start-ups the chip industry, just a couple of years after they were founded, they opened their first assembly facility in Hong Kong. So they were a trans-Pacific company from almost day one.

And I think that's important because the volumes of investment that are needed to make cutting-edge facilities have always been large and there has been a need from day one to spread out that investment amongst a large number of customers, but also to do so in the places that are the most efficient, to drive down costs and make production and R&D investment economically viable.

And so what you find is that the chip industry was one of the first industries really to globalise its value chains, such that by the 1960s you already had companies that were doing assembly in one country, R&D in a different country, manufacturing in a third country. And that's been critical to the industry's success.

And so as you allude to, now there's pressure from governments to rejig supply chains in a way that makes different governments feel like their supply chains are more resilient to geopolitical conflict. But that's got to be done in a way that takes into account, first off, how complex the supply chains have become, and indeed they're extraordinarily complex, but also the logic behind, because there are huge efficiencies that have been gained from this internationalisation of production.

And so if we try to rejig supply chains in the wrong way, you could end up imposing massive costs that aren't just bad for companies, but also slow down the rate of technological progress. So that's a really difficult and complex balance that has got to be struck.

John Letzing: So for the average person, they might think what could be so hard - you build a factory and then you start making these things. So maybe can you tell me a little bit about what's involved and what is the sort of cost involved?

Chris Miller: Well, just the dollar value alone, I think, speaks to the difficulty. A new fab will cost $20-$25 billion, so they are the most expensive factories in human history.

And the scale is is really hard to comprehend unless you've had a chance to walk around one. I'll give you one anecdote. TSMC, the world's biggest chip maker, Taiwanese firm, is building a new facility in Arizona which I had a chance to visit recently, and they have on site the seventh biggest crane that exists in the world to move around huge metal beams as they build these vast facilities.

So you've got scale at a really extraordinary way. But then that massive scale has to be used to produce microscopic scale. And so the transistors on the chips that will be produced inside of this facility, will each one be roughly the size of a coronavirus, produced by the trillions and trillions and trillions.

And so the combination of mass scale plus absolutely microscopic scale is something that you wouldn't see in any other industry. And that's why there's so much precision involved and there's so few companies that have learned to undertake this type of manufacturing in an economically viable way.

John Letzing: You mentioned TSMC. Can you tell me a bit about how it was that so many of the world's chips came to be produced there, and what the genius is behind their their business model?

Chris Miller: Well, TSMC was founded in 1987 by a businessman named Maurice Chang, who actually spent his early career building the chip industry himself while working at Texas Instruments, outside of Dallas.

He was present at the creation of the chip industry, and he had an idea that he was mulling over over the course of the late seventies, early eighties, that the cost of chip making was rising year after year, the complexity was growing, and he realised that companies increasingly wouldn't want to have to undertake manufacturing themselves, just getting too complex, too costly. And there were economies of scale that meant it was more sensible to have a smaller number of firms spending more time focusing on manufacturing.

And he envisioned sort of doing what Gutenberg did for books, in the semiconductor industry. Gutenberg, didn't write any books, he only printed them. And Maurice Chang envisioned a company that wouldn't design any chips, would only manufacture them.

At the time it was a really radical idea because there were no companies that only designed chips. Everyone both designed and manufactured in-house.

But he projected these trends of growing costs, of growing complexity into the future, and he set up TSMC around the idea of doing zero chip design only manufacturing for outside customers. And the Taiwanese government was very supportive, gave him his initial funding.

And the past 35 years have shown that this business model has been an extraordinary success. And so he has attracted the world's biggest tech companies, from Apple to Google to Nvidia, to produce their chips with his company. And it's ridden this trend of economies of scale further and further and further. And today, it's the world's largest chipmaker and it's also the world's most advanced producer of cutting edge processor chips.

John Letzing: Give us a sense, percentage-wise, of all the chips being made now, what's coming out of TSMC, in terms of chips in general, and these advanced chips in particular.

Chris Miller: When it comes to the most advanced chips, the types of chips that are running the processor inside of your smartphone or at the chipset that Nvidia uses to train AI algorithms, around 90% are produced by TSMC in just a couple of facilities in Taiwan.

So there's just extraordinary concentration. No one else in the world can do it at that level of precision or with that vast manufacturing capacity.

John Letzing: Yes, and they obviously are making a lot of the chips that are powering artificial intelligence. In terms of AI in particular, can you give me a sense of competition within that specific realm and how that's impacted the chip industry?

Chris Miller: It really has and the key trend here is that the amount of data used to train a cutting-edge AI system has been doubling once every 6 to 9 months, according to the best research on the subject. And that means that processing power, the ability to make sense of this data, to use it to train your AI system, has been a real limiting factor.

And today Nvidia has around a 90% market share in the chips that are used to train AI systems. And right now there's such a shortage of the chips that Nvidia makes, called GPUs [Graphics Processing Units], that there's actually venture capital firms that are not investing dollars in AI startups, they're just giving them the chips, because there's such a scramble under way to acquire access to these chips that they're more valuable than the dollars in the startup space.

And it's important because the the types of computing that you undertake to train an AI system is different than computing in your phone or in your PC. So Nvidia has cornered the market for this type of computing. It's also developed the software applications that go on top. And so, with only a couple of exceptions, almost all of the firms that are raising large sums of money to train big AI systems are using Nvidia's chips for that purpose.

John Letzing: There are so many interesting serendipitous turns of event in your book. Is Nvidia's realisation at a certain point that these chips that were being used largely for video games had amazing applications in AI? Does that qualify?

Chris Miller: It really does and it surprised the company just as much as it surprised outside observers.

It was video games and then in more recent years cryptocurrency mining that were the two drivers of Nvidia's growth.

And in some ways there's a sort of counterintuitive conclusion that actually the key impact of cryptocurrency on the world is probably not going to be in the creation of Bitcoin or Ethereum, but is going to be in driving Nvidia's growth over the last decade as the company was investing billions of dollars a year trying to not focus on crypto mining but actually build the chips and the software applications needed for AI.

But certainly it was a wild journey for Jensen Huang, the CEO, in the several decades since he founded it. He would readily admit that they had no idea that AI would be their primary business several decades down the road, because, as you say, they were focussed on graphics and graphics in computer games and video games when they started.

John Letzing: Can you give us the high level view of what it is about an advanced chip that you might need for AI that works in a way that's different from other chips?

Chris Miller: The key difference between a chip called a CPU, which is the main chip in your phone or your PC, and a chip called a GPU, which is the type of chip that you use for AI training, is that CPUs do many different types of things one after the other, sequentially, whereas GPUs are good for parallel processing, they can do multiple parallel things at the same time.

And so when you're training AI systems, that parallel processing provides huge increases in speed. And so if you want to train AI, you can do it on a CPU, but the speed is much slower, the energy intensivity is much higher. And so the cost of AI training just shoots up exponentially.

And that matters because companies are already spending vast sums on AI training. Sam Altman of Open AI has publicly predicted that he might spend up to $100 billion developing, OpenAI's systems and that will largely be spent on paying for CPUs. So that the cost of compute is really enormous. And so there's huge gains to be had from finding the most efficient chip for a given type of AI workload.

John Letzing: Silicon Valley has this global reputation for relentlessly churning out new competitors and disrupting markets and refreshing competition. Is it ironic in any way that pretty much all of these tech companies coming out of there or anywhere have to rely on a small handful of companies in the chip industry?

Chris Miller: It is something of a paradox because on the one hand there's no doubt you've got extraordinary disruptive innovation start-ups like Nvidia emerging. And Nvidia, started by three chip designers who met in a Denny's in a rough neighbourhood of San Jose in their earliest days. So a real classic startup story - not a garage in Palo Alto, but close.

And then you've got next to them firms that have almost monopolised their market and have been in their market position for a decade or longer.

I think there are two reasons why you've got these dynamics.

One is the capital investment needed in chipmaking is so vast that it's really hard to start up new firms in the manufacturing business. In chip design it's easier. But in chip manufacturing, good luck raising $20 billion from your family and friends to get your chip manufacturing startup off the ground.

So that's one. Second is the amount of complexity involved in getting close to the cutting edge is just far, far larger than in most other industries. The material science knowledge you need in the manufacturing, the chemistry, the physics. There's so many different disciplines that have to be integrated. And unless you're able to get all of them at the same time, at the cutting edge, you're not anywhere close. Because being second best in this industry is a guarantee that you won't have the customer base, you'll be far from the cutting edge. And so there's huge returns from being the best, which means that it's very, very difficult to be a startup because you've got huge barriers to entry that you face in your early years.

John Letzing: And given the dynamics that you just described, I think it's fair to say, and you say in the book, that no country really assumes that it can just do it alone. You describe a definite desire to create a sort of distinct non-U.S. supply chain. And I guess I'm wondering what what do you think about the prospects for that? Is that something that you see potentially realistically happening?

Chris Miller: Well, I think right now the dynamic is that you've got a clear move by the US and Japan, joined by the Netherlands, to limit the transfer of advanced chipmaking tools to China. And China obviously is not going to accept that without trying to find ways around it.

And so Chinese firms, the Chinese government, are investing quite heavily trying to build up alternatives. But the challenge is that in many of these market segments where you've got just one or two firms capable of producing ultra precise manufacturing equipment it's just very, very difficult to replicate.

So if you look at some of the tools that are used to make chips, and these are the types of tools that have been restricted in recent months, some of these tools have hundreds of thousands of components, including the flattest mirrors humans have ever made, the most powerful lasers ever deployed in commercial devices.

They're just extraordinarily complex, far more complex than spaceships that can go to the moon or almost any type of medical device. And so the idea that new firms are going to enter the market to learn how to produce at this extraordinary level of precision is something that's not being done.

The trend we have right now in the chip industry is an effort to replicate the capabilities of these firms. The question is how long will it take for competitors, almost certainly from China, to learn these capabilities. And I think the answer is it's going to take a fair amount of time, just given the complexity involved.

John Letzing: And it's not just geopolitical risk that threatens these choke points. I think one really interesting aspect that you raise in the book is the threat of natural disaster. There's this corny James Bond movie from the eighties where the villain wants to destroy Silicon Valley with an earthquake. But in fact, as you point out in the book, that's maybe not such a farfetched threat. Can you tell me a little bit about what's what's involved there, what's the risk is?

Chris Miller: I think when you zoom out and realise that we built the chip industry in Silicon Valley, in Taiwan, in Japan, a number of highly seismically active zones, in hindsight, it doesn't look all that wise.

Even production of the lithography tools, one of the types of tools you need to make chip-making equipment, is done in the Netherlands, is only, I was recently reminded, 70 feet above sea level where their factory is.

So it's easy to imagine ways that natural disasters could cause huge disruptions, especially given the concentration that we discussed. Now, the chip industry spends huge sums of money trying to insure itself against these capabilities. So all these facilities have redundant power and all sorts of systems designed to make sure that they've got as much insurance as you can realistically have.

But in the case of a big earthquake located in the wrong region, it's not difficult to imagine the ways natural disasters can cause not just billions, but in the worst case, a trillion dollars of damage.

John Letzing: For a lot of people, maybe the only time they became aware of semiconductors was during the pandemic with the shortages there. But you raise a really interesting argument in the book, which is that you're saying that it really wasn't as much to do with supply chain hiccups as it was to do with demand growth. So can you maybe tell me a little bit what was behind that?

Chris Miller: One fact that surprises a lot of people is that semiconductor production globally increased every single year of the pandemic. So we produce far more chips after the pandemic than we did before it.

And so there were some disruptions when factories were closed due to COVID restrictions. But actually the chip industry did, I think, a really extraordinary job in making sure that those disruptions were as limited as possible.

And in some ways it makes sense because inside of a chipmaking facility, it's not just COVID that induces companies to take cleanliness very seriously. You know, a single speck of dust can cause millions of dollars of damage. And so they were already highly environmentally controlled factories that we're talking about.

The key disruption of the pandemic was a surge of demand that was completely unpredicted, demand for new PCs as people started working from home, more data centre infrastructure as companies worked to upgrade their communications capabilities.

And at the same time, many other companies that bought chips predicted early in the pandemic that their sales would collapse. So car companies looked at the pandemic, said this is going to be economic disaster, cut their production plans and therefore called up their component suppliers and said we need fewer components.

And it turned out that that prediction was wrong. And actually, thanks to stimulus from governments, consumption kept up in ways that surprised everyone.

And so when car companies in particular called up their semiconductor component suppliers and said, actually, we're going to need the chips that we cancelled, in many cases they found themselves in the back of the line because that capacity had already been promised to people who were making PCs or electronic devices.

And so that supply chain issue, in terms of who was demanding and who was supplying, created a mismatch that caused hundreds of billions of dollars in lost sales that I mentioned earlier and it illustrated that you don't even need disruptions in supply to cause huge problems for users of chips, all you need is unexpected surges in demand.

And so I think it helped sensitise people to, 'well, wait a minute, if increasing supply coupled with an even more rapid increase in demand could cause such mass disruptions, what would happen if supply didn't increase but it decreased due to some sort of disaster? How would my business respond then?' And the answer is that most companies didn't and really still don't have good plans in place for how they would react in case the world's supply of chips decreased dramatically for a reason of natural disaster or geopolitical conflict.

John Letzing: Another compelling point that you make near the end of the book is that the demand for computing power isn't going anywhere. Probably it's not going to ever diminish. But as you say, we could run out of supply.

Chris Miller: Well, the chip industry has been unique because of Moore's Law, which says that the computing power per chip doubles roughly every two years. And that's been true since the 1960s, when Gordon Moore, who was one of the founders of Intel, first coined the phrase. And that rate of growth is unprecedented in any other segment of the economy in all of human history. Full stop. Doubling every two years. I calculated what it would look like if aeroplanes flew twice as fast every two years and we'd be well over the speed of light by now.

But the chip industry has delivered that and it's happened in a way that most people take completely for granted. Most of us are only vaguely aware that Moore's Law exists and have never really thought about the extraordinary advances it has made possible.

If you look at the cost of computing, it's collapsed over the course of our lifetimes, to say nothing of the course of the entire chip industry. The first semiconductors that were available for sale in the 1960s had four transistors on them. Today, if you go to the store and buy a new smartphone, just the primary chip will have at least 10 billion transistors on it. So from 4 to 10 billion has been the rate of progress. And the price of a chip hasn't changed much. So the cost per transistor has declined by roughly a billion-fold.

And Moore's Law is actually not a law, it's just a prediction. And that's the problem. At some point it will become impossible to shrink transistors further, which is the main strategy we've used to put more of them on chips.

We're not there yet, but I think in a decade's time there will be a lot of questions asked about can we find new pathways to shrink transistors that will be as economically viable as our prior pathways were. And if not, the cost of computing could begin to increase meaningfully because we've relied on Moore's Law thus far to deliver it, and Moore's Law is not guaranteed to continue.

Robin Pomeroy: Chris Miller’s book is called “Chip War”.

He was speaking to John Letzing, Digital Editor at the World Economic Forum's Strategic Intelligence platform, a curated mix of expert insights and contextual intelligence on topics ranging from artificial intelligence to water - you can find out more at intelligence.weforum.org.

Please subscribe to Radio Davos wherever you get your podcasts and please, if you enjoyed what you heard today, leave us a rating or review. And you can find our back catalogue - including that recent mini-series on generative artificial intelligence - on your podcast app, or at our website wef.ch/podcasts, which also has transcripts of our programmes.

And join the conversation about podcasts on the World Economic Forum Podcast club -- look for that on Facebook.

This episode of Radio Davos was written and presented by me, Robin Pomeroy, with John Letzing. Editing was by Jere Johansson. Studio production was by Gareth Nolan.

We will be back next week, but for now thanks to you for listening and goodbye.