Paul van Gerven
6 May 2016

After LEDs and solar panels, China is now setting its sights on semiconductors. International chip makers that want access to the Chinese market would do well to set up local operations.

China’s semiconductor production has grown massively in recent years. Where in 2000 the country only manufactured 2.2 billion dollars’ worth of chips and related devices, that number was 60 billion last year. That’s a little less than one fifth of the global market, and twice as much as Europe manufactures.

Yet it’s not nearly enough for the Chinese government. It ruffles the powers that be in Beijing that of the 145 billion dollars in semiconductors that China uses to produce electronics, at most one third is manufactured at home. The imbalance has been growing for years (see figure 1). What’s more, China manufactures primarily the simple stuff; the country’s role in advanced chips is insignificant.

After making major inroads into LED and solar panel production, China is now setting its sights on semiconductor manufacturing. In 2014 the State Council presented its national guidelines for semiconductor industry development, which stipulate that by 2025 China should manufacture 70 per cent of the chips that it uses (see figure 2). According to consulting firm McKinsey, that can only happen in the foundry sector, for example, if the entire world builds all its new fab capacity in China for the coming decade – so ambitious are the country’s plans.

Grafiek BC03 2016 Chinese circuits
Figure 1: China’s semiconductor manufacturing can’t keep up with domestic demand. Source: PWC

Scraping by

In fits and starts, the Chinese government has actually been trying to stimulate the domestic semiconductor industry for decades. Most of these efforts had little effect, because too little money was reserved for them and because investment decisions were left to the bureaucrats. This time, China’s taking a different approach. At 100 to 150 billion dollars, the war chest is well filled for the coming fifteen years, and the decision on how to spend that money will be entrusted to people who are on the front lines. Companies that have been identified as national champions will take the lead.

Yet it’s no small feat that China has set itself. In the past Taiwan and South Korea, and before them Japan, were able to fight their way into the semiconductor industry, but those were different times. Growth has slowed, and only the largest company or two in each sector is able to turn a profit. The rest are just scraping by.

Grafiek BC03 2016 Chinese targets
Figure 2: If China is to meet its own IC objectives, it will have to absorb all the foundry capacity to be built in the coming ten years. Source: McKinsey

In addition, China is dealing with a technological deficit so great that only Intel spends more on R&D. The country thinks in terms of cost competitiveness, but will have to teach itself the role of innovator. Drawing on foreign expertise is proving difficult: Micron, SK Hynix and Fairchild are among the companies that have turned down offers on all or part of their business. Few companies are eager to share their secrets with a country that has a poor reputation for protecting intellectual property.

Many countries also have legislation that simply forbids the transfer of strategic technology. The only country where China has managed to get a few toes in the door is Taiwan, but that country’s recently elected new government wants to increase its distance from the mainland.


But semiconductors are a strategic priority for China; profit can come later. The country is not afraid to mould the game to suit itself through draconian measures. A few years ago, for example, it provoked a war of attrition in the photovoltaic industry in order to penetrate the solar panel manufacturing market.

This time, however, it’s more likely that China will follow the same strategy it used in the rare earth industry. With a near-monopoly in production, the country limited the export of these technologically vital materials several years ago. Foreign companies that were unable to find alternative sources found themselves forced to start something up in China. (Hastily initiated mining operations outside China seem to have contained the damage wrought by its ploy: the quota was lifted last year, in part thanks to pressure from the World Trade Organization.)

In semiconductors, China can use access to the Chinese market as a lever. Companies must accept Chinese investments or start operations on Chinese soil to be able to continue supplying their products – a story all too familiar to Qualcomm: with the antitrust authorities nipping at its heels, the company has recently begun working with Chinese organizations.

TSMC is also admitting defeat. The company recently announced it would build its first fab on the mainland. The foundry has always served the entire world from Taiwan, but must suddenly cater to ‘local needs’. China’s need for a native semiconductor industry, that is.