What Happens to the Chip Supply if China Invades Taiwan?
One of TSMC’s factories in Central Taiwan Science Park in Taichung (Picture Credit: Briáxis F. Mendes (孟必思))
Taiwan matters to the world for a variety of reasons, some geopolitical, others ideological, but in recent years, one reason has become increasingly prominent: Taiwan manufactures most of the world’s semiconductor microchips. The country houses two foundries that, combined, produce 63–65% of the world’s microchips. The larger of these, Taiwan Semiconductor Manufacturing Company (TSMC), commands as much as 58% of the market while its nearest rival, Korea’s Samsung Electronics, supplies only 17%. At the denser (and thus more powerful and efficient) end of the microchip spectrum, TSMC dominates the trade, with a market share of over 90%.
Chips are integral to just about everything electronic, from household appliances like smart kettles and fridges, to computers, phones, and cars, to planes, tanks, and missiles. But it wasn’t until early 2020 that the world felt just how dependent it is on Taiwan for these chips, as a global microchip shortage hit, affecting some 169 industries, from soft drinks to steel, furniture to construction, and every type of manufacturing imaginable. The automotive industry has been hit particularly badly. With the 2020 chip shortage (that has extended into 2021, and will likely continue into 2022), the global auto industry is estimated to have lost $210 billion in revenue. 575,000 industry jobs in North America have been impacted (with workers losing work hours, pay, or even their employment altogether), factories were shut down, high-end features were slashed to decrease the number of chips needed, and used car prices jumped by 23% in Australia and nearly 25% in the US in response to the dearth of new cars. To make matters worse, electric cars use more than double the number of semiconductors that conventional cars do, undermining countries’ attempts to transition to electric vehicles. US President Joe Biden has called chips “extremely critical,” with disruptions capable of putting “American lives and livelihoods at risk,” and a member of South Korea’s ruling Democratic Party, Representative Yun Ho-jung, has called semiconductors tantamount to “the rice of the economy of the 21st century and the heart of South Korea’s economy.”
The fact that most semiconductor manufacturing takes place on Taiwan, an island off China’s east coast, which Beijing views as a renegade province and seems increasingly likely to try to invade and seize, makes the supply of these chips even more precarious. Even if China tried, but failed to take Taiwan, the war for the island would itself disrupt chip supply, and thus, the global economy. If China annexes Taiwan, it would obviously also appropriate its semiconductor foundries, suddenly giving it control of more than two thirds of the global supply of microchips, allowing it to choke off access to them to any country it likes. How might China use this power? Would it wield it with self-restraint, or would it weaponize it, doling out chips as rewards to its friends and withholding them to punish countries that displease it, like how it’s sanctioned Australia for calling for an independent probe into the origins of COVID?
If China annexes Taiwan, it would suddenly gain control of more than two thirds of the global supply of microchips, allowing it to choke off access to them to any country it likes.
If the supply of microchips was halted, this would reverberate across every chip-dependent industry (which is most of them) as new products get delayed or cancelled, and older products become more expensive to compensate for the shortfall. The world’s technological progress would slow or grind to a halt; no new Play Station 6 or Xbox; no iPhones 14, 15, and 16; no new cars, especially if they’re electric; and no new computing clusters or graphics cards. Companies and consumers will be forced to revert to earlier technologies that are less dependent on cutting-edge chips to satisfy their demands, and if they break, no replacements will be forthcoming.
Some countries are trying to hedge against this risk. The EU is investing $160 billion in tech with the goal of producing 20% of all chips in the world by 2030, the US is investing $52 billion to establish home-grown foundries, and TSMC has agreed to build a $12 billion plant in Arizona (though this plant would be capable of producing only 3% of TSMC’s output in Taiwan, making it more tokenistic than truly helpful).
But becoming self-sufficient in semiconductor manufacturing is not as simple as getting a blueprint of a TSMC foundry and recreating it elsewhere. Building even an entry-level foundry that produces 50,000 wafers a month costs about $15 billion (note that TSMC makes 2.7 million wafers a month). Furthermore, foundry workers are highly-skilled, highly-educated engineers who are in short supply, and whose experience matters as much as their knowledge. It takes between 400-1,400 steps over approximately 12 weeks to make a silicon wafer (the heart of a microchip) which is 100 layers thick, some as thin as a single atom. To complicate matters further, the raw materials must be incredibly pure, allowing for impurities equivalent to only 1 grain of sand in 16 Olympic swimming pools and, during production, a single speck of dust could destroy the whole wafer, costing millions of dollars. Taiwan has spent decades building up the capacity, expertise, and supply chains to manufacture state-of-the-art chips at this scale, and other countries will have a hard time recreating this feat in just a few years. In the words of Daniel Nenni, co-author of Fabless: The Transformation of the Semiconductor Industry: “it’s just impossible for any company or country to catch up to this huge ecosystem that’s moving forward like a freight train.” For the present, at least, the world is hugely dependent on Taiwan’s foundries for semiconductors, which makes the question of who controls them – and how they use this power – of paramount importance.