China stumbles on the long march to close the technology gap4 min read
Nina Xiang is the founder of FutureLogic, a media platform bridging the Asian and global innovation economy. She is the author of “US-China Tech War: What Chinese Tech History Reveals About Future Tech Rivalry.”
Chinese president Xi Jinping has been worried about China’s reliance on foreign technology for a while, repeatedly airing his concerns ever since he became president.
Perhaps most memorably, Xi said in 2016 that the Chinese internet sector’s reliance on foreign core components was like “building a house on someone else’s foundation” that would not withstand wind and rain no matter how beautiful.
Later that year, China’s State Council issued a national policy that aimed to fundamentally reverse the country’s heavy reliance on foreign technology by 2025. But these words and policies did not translate into earnest actions until after 2018, when the lethality of U.S. sanctions was laid bare, causing something of a Sputnik moment in China.
Yet three years into Xi’s campaign to replace foreign tech with homemade versions, it is becoming clear that this effort will take much longer than Beijing originally planned. Nor can its success be taken for granted yet.
One core part of Beijing’s strategy is what is being called Xin Chuang, or building up a comprehensive self-reliant domestic information technology industry that can produce chips, operating systems and applications. Imagine a Chinese tech sector without Intel, Qualcomm, Microsoft or Android: that’s what Beijing sees as a secured tech sector free from the fear of U.S. sanctions.
This strategy is being carried out in three steps. First, China is nurturing a self-reliant market in the relatively closed government and Party-related sphere valued at dozens of billions of dollars.
Then, domestic replacements will expand to key state-owned sectors, including telecommunications, rail, electricity, health care, aerospace and energy. This market could be four to five times bigger. The last step is to cover the consumer market, which includes consumer phones, that is worth hundreds of billions of dollars.
So far, when it comes to steps one and two, domestic replacement efforts have achieved some progress. But there has been very little progress when it comes to the third step.
Phytium Technology, the absolute leader among China’s self-developed processor chip companies tailored to government entities, delivered 1.5 million chips in 2020 and expects to ship over 2 million chips this year.
It does not make sense to compare it to Intel’s 2021 shipments, which are expected 1.14 trillion. Still, the company has taken the first step in a long march, even compared to driver-assistance chip specialist and Intel subsidiary Mobileye, which delivered 19.3 million chips in 2020.
Another challenge for China’s tech industry is policy uncertainty. For example, both Phytium and Huawei Technologies’ chip design unit HiSilicon’s processors are based on chip architecture licenses they had obtained from British chip design company ARM.
Even if Beijing can block ARM’s sale to Nvidia to prevent ARM from becoming an American company, it remains uncertain whether chips based on ARM designs can be 100% secure for China.
While ARM had stopped working with Huawei after U.S. bans in 2019, the company is embroiled in a distracting dispute with ARM China over control of its China business. All of this puts into doubt the ability of Huawei and other Chinese companies to continue using ARM architecture in a highly uncertain position.
Nor is there consensus among China’s IT industry whether the country should bet on the ARM structure, which works well for future applications like internet-of-things and has a mature ecosystem, or focus on using a completely self-developed chip structure.
Moreover, Beijing’s requirements for “domestic-made” products are unclear and disparate. After proposing that a “domestic product” be defined as a product having over 50% of its total costs coming from domestic sources, Chinese agencies have yet to issue a final rule. As a result, each government organization uses a different definition.
Such policy and regulatory uncertainties cause confusion and fragmentation. From chip designers to operating systems to software developers, they must navigate different localization rules in each city. That increases costs, in addition to making interoperability across the supply chain much harder.
Because interoperability and adaptation are progressing slowly, it is common for users in government and Party-related entities to install two sets of operating systems and software: one domestic system to meet localization requirements and another system such as Microsoft’s Windows to ensure usability.
Domestic replacement among state-owned enterprises is also just beginning and faces similar challenges of fragmentation. Bank of China adopted an operating system from local supplier UnionTech, while China Construction Bank’s credit card system adopted another operating system called Kirin.
It is good to have competition initially, but the consolidation of operating systems is critical. This could take longer as various players compete for dominance.
It is unlikely that China will be able to reverse its heavy foreign reliance by 2025, or hit the industry forecast of having half of the country’s $104 billion computing market be Xin Chuang systems in 2023. Judging by the progress made thus far, it remains to be seen if China can achieve its goal of total self-reliance in the next decade, or even further in the future.
American companies are getting hurt too. Qualcomm’s shipments in China shrank 48.1% year-on-year in 2020 as Chinese phone makers sought to diversify their supply network. U.S. companies in the telecommunications, medical devices and marine equipment sectors may also find themselves soon shut out of the Chinese market as Beijing retaliates.
Both Washington and Beijing should try to avoid this lose-lose situation by returning to the negotiation table ready to compromise.