In a previous note, we looked at China’s endeavors in key digital technologies with the strategic aim to gain more digital independence from the West. An important prerequisite is having access to integrated circuits as they are an important enabling factor for any type of computational activity. Hence, China has set itself the target to become largely self-sufficient in 2025. Even though this ambition seems feasible in the light of its enormous national market and governmental financial support, we should also consider some of the considerable challenges that China faces towards digital independence.
Even though China has been able to become largely self-reliant in technology domains like online platforms (e.g. social media, e-commerce) and artificial intelligence, it still largely depends on the U.S. when it comes to its supply of semiconductors, a key enabler for any computational endeavor. China is worldwide the biggest importer of integrated circuits (ICs), annually importing $200 billion in computer chips (60% of the worldwide supply) of which only 16% is manufactured domestically. This vulnerable position is lately exemplified by the U.S. sanctions against ZTE, where the existence of a large Chinese company is suddenly at the mercy of U.S. policy. However, kick-starting a domestic chip industry is not a mere matter of brute financial force, but needs to consider a few important organizational, political and technological hurdles. Firstly, we already see that China faces resistance in bootstrapping themselves into the semiconductor industry by facing huge resistance in acquisitions. In total China has placed $34 billion in bids for U.S. semiconductor companies alone since 2015. However, China was only able to close $4.4 billion in semiconductor acquisition deals globally. Consequently, China is now looking for ways in growing it from the ground up domestically. However, Khong, Zang & Ramu (2015) foresee in their research four organizational hurdles that need to be overcome: “firstly, there is a need to build an industrial ecosystem around the semiconductor firms. Secondly, the universities should play a greater role to supply the human capital to continuously develop home grown firms and innovation capabilities. Thirdly, efforts must be taken to push domestic firms to integrate globally so as to appropriate systemic synergies (Rasiah, 2010). Fourthly, efforts must be taken to strengthen network cohesion between institutions, meso-organizations and the firms”. In addition, setting development targets by the government does not guarantee high quality industrial output. In the research of Kroll (2016) there are multiple accounts of low quality patents that merely serve the purpose to meet government targets. More generally, in a highly competitive market, intellectual property rights could also be less respected. One example, is the Hanxin microchip which was presumably designed by prof. Chen, but later turned out to just be a derivative from a Motorola chip, which in turn could erode trust in Chinese ICs altogether. From a political perspective, China could also face increased resistance as a consequence of a seemingly successful state-planned IC strategy. In addition to blocking acquisitions, it is not unthinkable that the U.S. government will also put its weight behind its own domestic industry. We can see that Intel is already lobbying for government support in order to fight the impending IC battle. As a result China’s ambition to bypass the U.S. could become a moving target, potentially escalating into an arms race with IC price deflation and an IC market crash to boot. Another political attack vector could be found in information warfare, in which growing suspicion towards Chinese tech being used for government intel will also extend to Chinese semiconductors. For instance, last month ZTE and Huawei devices have been removed from military base stores due to security risks. Next to these organizational and politcal hurdles, there are also technological drivers. One characteristic is that the integrated circuit industry is a heterogeneous space, where there are many types of computer chips ranging from random access memory (RAM), central processing units (CPU) to graphics processing units (GPU) to more specialized chips (e.g. ASICs, FPGAs, quantum computers). Here we can see that China has already gained a considerable position in NOR memory and is betting on DRAM and 3D NAND. Hence, the heterogeneity allows for developing an edge in one of these domains, however it is much more difficult to gain independence in all the different types of ICs. Lastly, the success of ICs is not only a matter of producing the fastest and cheapest chip, but also about being compatible with a larger ecosystem of hardware and software. One example is the Fangzhou 1 embedded chip that was launched in 2001; despite its functional competitiveness, it was not well received, even by domestic manufacturers, since Intel chips had less incompatibility risks and therefor a quicker path to market. In response, the Chinese government could theoretically use mandates to force local OEMs to use a certain percentage of Chinese parts in their systems, however, this would run the risk that these producers will be put at a competitive disadvantage due to lower quality ICs.
