The looming global battle over semiconductor supply

What are semiconductors?

As the name implies, semiconductors are materials whose physical properties fall between those that can conduct electricity (such as copper and aluminium) and insulators that can’t (glass and ceramics). Silicon, germanium or carbon are all natural semiconductors. In a process called “doping”, small amounts of “dopant” impurities can be added to pure semiconductor crystals of compounds such as gallium arsenide or cadmium selenide, which causes big changes in the conductivity of the material. It is this property that makes them useful in the building of electronic circuits. When two types of semiconductor are put together, you can build a diode, a crucial electronic component that acts as a “turnstile”, conducting electrical current in only one direction. You can then build a transistor, created by using three layers of semiconductors, and then a silicon chip, which is a small piece of silicon that can hold thousands of transistors. These are the foundations of the microprocessors that make modern computing possible. 

Semiconductors are important then?

They are essential to modern societies, in so far as we rely on electronic devices. Over the past half century, the process of building chips has (famously) been constantly refined and improved, so that today tens of millions of transistors can be cheaply formed onto a single chip, and most people carry phenomenal computing power around on their phones. Indeed, the word “semiconductor” (the essential component) is now often used interchangeably with “chip” (the end product). And about a trillion chips are made each year, or 128 for every person on the planet . Already, a typical electric car, for example, has about 3,000 different chips in it, and demand is set to soar further as (for example) artificial intelligence (AI) and data-crunching develop, and more industrial machines are connected and fitted with sensors. The $450bn global market for microchips is notoriously cyclical, with companies struggling to meet demand during upswings, and slowdowns sending share prices crashing. But even so, the current shortage has got Western businesses and governments spooked – for both economic and geopolitical reasons.

What’s going on?

Chipmakers are booming as a result of the pandemic. The world, stuck at home, has snapped up videogames, laptops and televisions. Share prices have soared: the world’s listed chip-makers are now worth a collective $4trn, four times their value four years ago. But the chipmakers have struggled to meet demand, and in recent weeks a dozen carmakers, including Volkswagen, Honda, General Motors and Fiat Chrysler, have all had to close down production lines because they haven’t been able to source the necessary chips. The shortage is expected to last six months, and as many as half a million vehicles could ultimately be affected, say analysts. The supply crunch has raised questions about the changing auto industry and how it can protect its interests as it transitions to electric vehicles. But it has also shone a light on the remarkable consolidation in the semiconductor industry, which involves only a tiny number of big players – mostly in Taiwan and South Korea.

How is the industry changing?

The semiconductor sector is at a historic inflexion point, where the Moore’s Law (which holds that the cost of computer power will fall by half every 18 months to two years) is breaking down, says The Economist. Each generation of chips is becoming more challenging technically to make than the last, and the cost of building factories far greater – leading to impressive “effervescence” on the design side, but potentially worrying overconsolidation in manufacturing. Nvidia, for example, which designs chips for gaming and AI, is now the most valuable US chip firm, worth more than $320bn (it is bidding for the UK chip designer Arm). Apple has begun using its own chip in its Mac computers; Amazon is also developing chips for its data centres.

And the manufacturing side?

By contrast, when it comes to actually making chips, the number of manufacturers at the industry’s cutting-edge has fallen from over 25 two decades ago to just three. Of those, America’s Intel is in trouble, and is reportedly set to outsource its 3nm (nanometre) chips to the global leader, Taiwan Semiconductor Manufacturing Company (TSMC). The third big player is South Korea’s Samsung, meaning that the world is currently highly reliant on two big firms – one of them based on an island claimed by China, and another right next door to an unstable dictatorship. “In the 20th century the world’s biggest economic choke-point involved oil being shipped through the Strait of Hormuz”, says The Economist. “Soon it will be silicon etched in a few technology parks in South Korea and Taiwan.”

Why does this matter?

In an age where chips are becoming foundational to economic progress, and where technology is already being enlisted in the emerging great power rivalry between the US and China, such dependence on one or two firms – and one global region – looks unsustainable. The current shortage, and the spotlight it has shone on Taiwan’s dominance, will add urgency to efforts – in Beijing and Washington, but also in Japan and Europe – to increase self-reliance in chip-making. It also has the potential to exacerbate brewing tensions over Taiwan’s status. Creating cutting-edge chip-making capacity in Europe will be the work of many years, says Joerg Wuttke, president of the EU Chamber of Commerce in China. In the meantime, not just capacity problems but the geopolitics of the chip industry – the possibility of short-notice government interventions, export controls, and associated political risk – mean that more shortages can be expected. “So better get prepared.”