THE incredibly complex, high-stakes business of making semiconductors has always been a battle of corporate giants. Now it’s also a race among governments. These critical bits of technology – also known as integrated circuits or, more commonly, just chips – may be the tiniest yet most exacting products ever manufactured.
And because they’re so difficult and costly to produce, there’s a worldwide reliance on just a handful of companies, a dependence that was brought into stark relief by shortages during the pandemic.
Access to chips has also become a geopolitical weapon, with the US ratcheting up curbs on exports to China to contain the rise of an economic rival, and spending tens of billions of dollars to repatriate some chip production from Asia.
1. Why the war over chips?
There is silicon at the heart of every artificial intelligence breakthrough, range upgrade for electric vehicles or guidance system for hypersonic missiles. Most of the world’s leading chip technology originates in the US. China is the biggest market for the electronic components and has a growing desire to make more of the chips it uses itself. That has made the industry a focal point for Washington as it tries to limit the rise of its Asian rival and address what it says are national security concerns.
For its part, Beijing has poured billions into efforts to build its own chip industry and lessen its reliance on imports that are increasingly subject to restrictions by the US. At the same time, the US and Europe are setting aside huge sums of government money to bring back physical production of chips, reducing what they say is a dangerous reliance on a few facilities in East Asia.
2. Why are chips so critical?
They are what is needed to process and understand the mountains of data that have come to rival oil as the lifeblood of the economy. Made from materials deposited on disks of silicon, chips can perform a variety of functions. Memory chips, which store data, are relatively simple and are traded like commodities. Logic chips, which run programs and act as the brains of a device, are more complex and expensive.
GET BT IN YOUR INBOX DAILY
Start and end each day with the latest news stories and analyses delivered straight to your inbox.
Access to components such as Nvidia’s H100 AI accelerator has become linked to both national security and the fortunes of giant companies such as Alphabet’s Google and Microsoft as they race to build out giant data centers and steal the lead in what’s seen as the future of computing. But even every-day devices are increasingly reliant on chips.
Every press of a button in a car full of gadgetry requires simple chips to translate that touch into electronic signals. And all battery-powered devices need chips to convert and regulate the flow of electricity.
3. Who controls supply?
Chipmaking has become an increasingly precarious and exclusive business. New plants have a price tag of more than US$20 billion, take years to build and need to be run flat-out for 24 hours a day to turn a profit. The scale required has reduced the number of companies with leading-edge technology to just three – Taiwan Semiconductor Manufacturing (TSMC), South Korea’s Samsung Electronics and Intel of the US.
TSMC and Samsung act as so-called foundries, providing outsourced manufacturing for companies around the world. The world’s biggest tech firms are dependent on access to the best manufacturing, most of which is located in Taiwan. Intel used to focus on making chips for its own use, but is also now trying to compete with TSMC and Samsung for contract manufacturing business.
Lower down the food chain, there is a huge industry that makes so-called analog chips. Companies such as Texas Instruments and STMicroelectronics are leading makers of the components that do things like adjust power inside smartphones, control temperatures and turn sound into electrical pulses.
This is the area that China, blocked from access to many of the machines needed to make more cutting edge parts, is targeting, investing heavily to boost production and grab market share.
4. How is the global chip battle playing out?
Despite a huge Chinese spending spree, the country’s chipmakers still depend on US technology, and their access to overseas chip production technology is shrinking.
- The US imposed tighter export controls in 2023 on some chips and chipmaking equipment to stop China from developing capabilities that Washington regards as potential military threats, such as supercomputers and artificial intelligence. In October, the rules were tightened further and agreements reached with Japan and the Netherlands to follow suit were due to come into effect in 2024. Then, in early 2024, the US pressed allies to further tighten restrictions on China’s access to semiconductor technology, aiming at plugging holes in export controls. The Biden administration is trying to bring Germany and South Korea into a China chip curb agreement that already includes Japan and the Netherlands, since all four countries are home to key firms in the semiconductor supply chain.
- Some of China’s technology leaders, including Huawei Technologies, have been placed on the so-called entity list. American chip technology suppliers have to get US government approval to sell to these blacklisted companies, a step meant to limit China’s ability to develop high-end chips and build cutting-edge AI applications. But Huawei in 2023 announced the Mate 60 Pro phone powered by a new Kirin 9000s chip. That processor was made by Shanghai-based Semiconductor Manufacturing International Corp. with so-called 7-nanometer technology, more advanced than US rules allow.
- US politicians have decided they need to do more than just hold back China. The 2022 Chips and Science Act set aside US$39 billion for direct grants, as well as loans and loan guarantees worth US$75 billion, to revitalize American chipmaking after decades of production shifting to Asia. Officials had unveiled six preliminary awards as of mid-April: three to firms that produce older-generation semiconductors, plus multibillion-dollar packages for Intel, TSMC and South Korea’s Samsung Electronics.
- Europe has joined the race to reduce the concentration of production in East Asia. European Union nations agreed in November on a 43 billion euros (S$62.5 billion) plan to jump-start their semiconductor output. The goal is to double production in the bloc to 20 per cent of the global market by 2030.
- Japan has earmarked about 4 trillion yen (S$35.2 billion) in government funds for revitalising its semiconductor sector and hopes that spending in the field, including private sector support, can reach 10 trillion yen. Among the plan’s goals is the tripling of domestically produced chip sales by 2030.
5. How does Taiwan fit into all this?
The island democracy emerged as the dominant player in outsourced chipmaking partly because of a government decision in the 1970s to promote the electronics industry. TSMC almost single-handedly created the “foundry” business model – building chips designed by others. It is an approach that chip users embraced as the cost of new plants skyrocketed.
Big customers like Apple gave TSMC the massive volume to build industry-leading expertise, and now the world relies on it. The company overtook Intel in terms of revenue in 2022. Matching its scale and skills would take years and cost a fortune.
Politics have made the race about more than money, though, with the US signaling it will continue efforts to restrict China’s access to American-designed chips made in Taiwan’s foundries. China has long claimed the island, just 100 miles off its coast, as its own territory and threatened to invade to prevent its formal independence. BLOOMBERG