The latest round of state investments in China’s semiconductor industry reveals an evolving approach by the Chinese government characterised by greater state involvement and strategic oversight.

This introduces notable adjustments to the model that has applied to China’s semiconductor industry since 2014, one in which state agencies have featured as key investors. The evolution we’re seeing is a response to US-led sanctions on China’s chip sector. Another factor is the disappointing performance of the state-entrepreneurship model, due to fraud and corruption.

The funding round, revealed in May, is the third phase of the National Integrated Circuit Industry Investment Fund, or Big Fund 3.0. In it, Chinese policymakers are trying to find a balance between, on one hand, bolstering high-end innovation to ramp up their de-Americanisation drive and, on the other, fostering existing capabilities in older chip technologies.

This latest push is likely to help China catch up with the United States faster. That’s the opposite to what the United States has been trying to achieve.

In terms of investment, China has upped the ante by promising a whopping 344 billion yuan (US$48 billion) for the 15 years to 2039. The amount is more than double the investment in the second phase, about $19 billion. It looks like a competitive response to the nearly US$53 billion of US government investments under Washington’s 2022 CHIPS and Science Act, and there’s no sign that the escalation will stop.

The fund is registered under the Beijing Municipal Corporation, with 19 shareholders, including the five biggest banks in China. The involvement of key banking institutions is a major change compared with the first and second phases. The investments made by Big Fund 3.0 will be managed by Sino IC Capital, a state-run company. The task of strategic coordination and delegation is allotted to the China Science and Technology Commission, a government agency.

This funding structure shows a greater degree of centralisation and state oversight, probably to avoid a repetition of past wrongdoings in management of government funds. Also, the appointment of Zhang Xin, a former senior inspector at China’s Ministry of Industry and Information Technology and a known industry specialist, as the fund’s president signals that Beijing aims to rely more on active semiconductor technocrats to ensure effective fund management.

The overhauling in China’s chip development strategy is a clear indication that Washington’s policies are giving the Chinese chip sector and China’s overall AI development a tough time. The Big Fund is expected to invest heavily in critically vulnerable areas such as equipment and raw materials, for which China depends heavily on Western suppliers. Responding to the crisis created by US-led sanctions, which ban the sale to China of high-end chip equipment, Big Fund 3.0 focuses on enhancing China’s independent capabilities. China’s strategy is to cut US restrictions by investing more in producing mature-node chips, which are in short supply.

Under Big Fund 3.0, the investors will focus on large-scale wafer manufacturing plants. Such plants will enable China to make high-bandwidth-memory (HBM) chips used in 5G-enabled devices, high-end AI systems, and so on. HBM chips provide very wide ranges of channels for data processing, are ideally suited to train large volumes of data in AI modelling and are essential to an advanced cloud computing industry. China aims to focus on producing HBM chips and domestically sourced machinery to address bottlenecks emerging in the domestic cloud computing industry due to US sanctions, and ultimately to advance China’s AI development.

The fund aims to strengthen the entire domestic chip supply chain and build a robust, independent ecosystem. The Chinese strategy is to build comprehensive chip capabilities. That will also enable investments in less sophisticated chips, which are made with older technologies but are useful in consumer devices, and compound semiconductors, such as gallium nitride, which are used in advance applications such as autonomous vehicles.

Unsophisticated chips and gallium nitride ones are both beyond the scope of US sanctions. If Chinese semiconductor firms make optimal use of state funding, China may be able to lead in production of them. If so, overcapacity and overproduction will probably follow, as is often seen elsewhere in Chinese manufacturing. That would significantly affect prices for other makers of such chips in Malaysia, Taiwan and elsewhere.

As a result of US sanctions, the scope of Chinese policies has widened across tech domains aimed at strengthening supply-chain independence and domestic market demand. A huge amount of state support will probably enhance the level of China’s indigenous capability, though complete supply-chain independence is far from possible.

While the prospects of this course correction look promising, a big question remains: whether Chinese chipmakers can climb the ladder to produce HBM chips.