The ambitious technological goals of China semiconductors set for later in this decade turn out to be aspirational rather than achievable...
The ambitious technological goals of China semiconductors set for later in this decade turn out to be aspirational rather than achievable.
That’s true of its ambitions to lead the world in AI technology by 2030. Evidence is mounting it is also true of Beijing’s “Made in China 2025” blueprint for semiconductor production.
With a technological Cold War looming, it gets harder by the day to forecast manufacturing growth of China semiconductors. Tightened U.S. export controls designed to block access to advanced manufacturing equipment are likely to set back top producer HiSilicon and its parent company, Huawei, several scaling generations.
Given those realities, IC Insights projects Chinese-based chip manufacturing will rise to $43 billion by 2024, accounting for a mere 8.5 percent of the global IC market expected to reach upwards of $507 billion.
More sloganeering than strategy
Lending credence to the notion that Made in China is more sloganeering than strategy, the market tracker forecasts that at least half of chip manufacturing in China will continue to originate from fabs jointly operated with Intel, Samsung, SK Hynix and Taiwan Semiconductor Manufacturing Co.
Will U.S. tariffs and stiffer export controls light a fire under Chinese domestic producers? Perhaps, but meeting its technology goals will require Beijing to put substance behind its deep pockets and rhetorical flourishes. What’s needed is patient investment to build the R&D infrastructure that underpins innovation, whether its DRAMs or AI algorithms.
China’s goal of AI superiority also looks shaky. Technology scholar Dieter Ernst concluded in a recent study that China’s AI industry is immature, fragmented and focused primarily on AI apps that will yield a quick return on investment. For all the resources Beijing has thrown at AI development, China has yet to develop the ecosystem necessary for real technology innovation.
The same is true for semiconductors and other critical electronics technologies.
Other observers confirm Ernst’s skepticism about China’s AI prowess.
Beijing “set a very ambitious goal for AI industry development in China between now and 2030,” Xiaomeng Lu, international public policy manager for technology consultant Access Partnership, told a recent forum on the global AI race.
The grand plan includes three components: technology advancement, industry expansion and establishing an AI policy roadmap. While the first two goals remain “aspirational,” Lu added that China may be positioned to establish AI technical standards over the next decade.
Beyond that, not so much.
While China has achieved success in areas like supercomputing, previous five-year plans to develop indigenous technologies such as computer operating systems have fallen short of the mark. Based on IC Insight’s forecast, it appears the same is true for semiconductors.
China watchers conclude that achieving technological independence from the west will require more than piles of cash and successive five-year plans. “The policy makers have to work with the private sector players to make these things happen,” Lu concludes.