China’s Evolving Semiconductor Strategy

China announced yesterday that it has established its third state-backed investment fund to boost its semiconductor industry, this one valued at US$47.5 billion, according to a filing with a government-run companies registry. The announcement is the latest in a decade-long drive to speed up development of its indigenous semiconductor sector and close the gap with the West. In 2014,  China established a $21 billion national investment fund for chipmaking and sought to acquire several leading western chip firms. Its bold ambitions captured the interest of global industry and investors and rattled officials in western capitals. Ten years later, China’s semiconductor industry is at a crossroads, facing an increasingly complex and even hostile geopolitical environment that includes an array of western export controls and a global chipmaking investment arms race.

Semiconductors are critical both for China’s high-tech growth and national security. They are also a key battleground in the growing geopolitical competition between China and the United States.

Semiconductors are critical both for China’s high-tech growth and national security. They are also a key battleground in the growing geopolitical competition between China and the United States. A central question is whether and how China will overcome increasingly severe and widening restrictions on high-end semiconductor technology, while also attempting to reduce its reliance on foreign supply chains. The buzzwords from China’s 14th Five-Year Plan and the 20th Party Congress that now define China’s semiconductor strategy are “science and technology self-reliance” and “secure and controllable supply chains.” But these are no easy tasks. To date, no country or company has managed to build a self-sufficient semiconductor supply chain. Will China be different?

U.S.-imposed sanctions against China have created serious challenges to the country’s progress in advanced semiconductors and the key technologies that rely on them, such as artificial intelligence (AI). Due to an inability to acquire the most cutting-edge chip production equipment, China is still two or three generations behind chipmakers in Taiwan, Korea, and the United States in producing leading-edge logic chips. Chinese AI designers are now forbidden to use the most advanced logic foundries outside of China and AI labs have also been cut off from purchasing the most advanced AI chips. Amidst an industrial revolution catalyzed by nascent frontier AI technology, including large language models, Chinese researchers are right to worry that they will struggle to keep pace. One of China’s state-owned memory chip makers, Yangtze Memory Technologies Co. (YMTC), has also been hit with damaging sanctions that have cut off their ability to serve global markets. Added to this, China has still not made necessary progress in replacing critical advanced lithography systems and other high-end chip production tools that it previously procured from Netherlands-based firm Advanced Semiconductor Materials Lithography. Despite years of sustained efforts, China still relies on its existing install base of imported machines.

China now accounts for nearly a quarter of global 300mm chip manufacturing capacity, up from 12 percent in 2014, with the balance of production shifting from predominantly multinational to domestic chipmaking operations.

Despite the fierce geopolitical headwinds, China’s semiconductor industry has made significant progress over the last decade. China now accounts for nearly a quarter of global 300mm chip manufacturing capacity, up from 12 percent in 2014, with the balance of production shifting from predominantly multinational to domestic chipmaking operations. China’s leading chip foundry, Semiconductor Manufacturing International Corporation (SMIC), has tripled revenue and doubled capacity to become the world’s third-largest foundry. Especially in “mature node”—that is, 28nm and above—chip production, China is rapidly expanding capacity and is on pace to add more than 18 new chip fabs in 2024 alone. China has also built up an impressive memory chip industry from nothing (memory chips being those involved in data storage, versus logic chips, which are involved in computation), and has now a range of competitive Chinese fabless chip designers producing everything from advanced AI chips, and microcontrollers for industrial goods, to power management chips for electric vehicles. In the upstream supply chain, including chipmaking equipment, materials, and software design tools, Chinese companies are accelerating their growth, albeit from a low base.

China’s strategy to promote semiconductor technology innovation and increase self-reliance is multi-pronged; some elements appear more likely to be successful while others face serious obstacles. First, in the near and medium term, China will benefit from a large and booming market for industrial goods that do not require the most advanced chips. This includes the “new three” industries—electric vehicles, batteries, and solar cells—but also industrial robotics, medical devices, and consumer electronics that comprise the sectors prioritized for development under the “new productive forces” strategy. China is building an immense number of legacy chip factories that could serve these markets, as evidenced in skyrocketing chip tool imports and sales. However, there are indications that China is building more supply than demand, which has stirred fears of oversupply in western capitals. Already, Chinese foundries are engaged in a price war with their domestic competitors that has spilled over to impact similar firms in Taiwan and South Korea. In response, the U.S. government has doubled tariffs on the import of semiconductors, and has many other tools it could use to further address Chinese legacy chip oversupply.

Second, in the realm of advanced technology, China faces an uphill climb in attempting to close the gap with global chip technology leaders. For advanced logic, China is working to develop manufacturing techniques that may extend the viability of their lithography tools to imprint 5nm chips (still two generations behind the state-of-the-art), but it will be challenging to produce enough chips to meet burgeoning Chinese demand for advanced AI chips. This effort is now being led by Huawei, in partnership with China’s leading chip firms and research institutes. In the realm of memory chips, Chinese chipmakers hope to take advantage of loopholes in the export controls issued by the United States in 2022, which currently do not restrict 3D in-die stacking of DRAM (dynamic random-access memory chips). For developing chips to power China’s AI ambitions, China is researching ways to leverage 3D integration of chips for both logic and memory, and is exploring the development of wafer-scale level chips. However, Chinese efforts are still nascent and far behind the established global leaders who are today leveraging all of these technologies for high-volume commercial production.

There are no shortcuts to technology leadership in the global semiconductor industry, and China is no exception. While China will continue to make progress in semiconductor innovation, it will likely find itself falling even further behind in some areas as western firms race ahead. However, in the short to medium term, China is likely to be able to make do with the mature chip technology it already has, while significantly boosting its self-sufficiency in the corresponding legacy chip supply chain. In terms of government policy, given that China has already established a national plan, fund, and advisory group for the chip sector, the priority now is policy coordination, which is being managed by the new Central Committee on Science & Technology. Central efforts behind the scenes may help to better coordinate research and development efforts between the state and private sector, and prevent an all-out race to the bottom price war in legacy chips. Over the long term, China will also likely continue to invest in semiconductor-related basic research and talent development, two often forgotten, but essential elements to a winning chip strategy.  While $47.5 billion in new funding will help, money alone won’t close the technology gap for China, and other nations not encumbered by export controls are now spending billions. For the foreseeable future China will likely be a “fast-follower,” perennially challenged in keeping pace with global leaders.

Jimmy Goodrich is an IGCC nonresident fellow. The views expressed by the author are his own and do not reflect those of any organization with which he is affiliated.

Thumbnail credit: Wikimedia Commons