A Bombshell For U.S.-China Tech Relations
Export control regulations have set the stage for years of intense tech competition
A few weeks back I published a 3000-word article in Noema Magazine on the long-term implications of US export controls and the future of US-China tech competition. I’ll be running some short excerpts below but I’d encourage you to read the whole thing.
I start off the piece by exploring how US policymaking evolved over forty years from explicitly supporting China’s military advancement to the new Sullivan tech doctrine trying to “maintain as large of a lead as possible” on “force multiplying technologies.”
Then I explore how the export controls will play out in the Chinese semiconductor ecosystem.
Chinese firms like SMIC may have been able to produce small runs of 7nm chips in the past year, but without American tech, doing so at scale will be impossible for perhaps a decade or more. The hundreds of billions of dollars of R&D invested over decades, not to mention the network effects and positive feedback loops for improving both design software and tooling, mean that the chokepoints the Biden administration aims to hold onto will prove durable into the medium term…
Reuters reported yesterday that “China is working on a more than 1 trillion yuan ($143 billion) support package for its semiconductor industry, three sources said, in a major step towards self sufficiency in chips and to counter U.S. moves aimed at slowing its technological advances. Beijing plans to roll out what will be one of its biggest fiscal incentive packages, allocated over five years, mainly as subsidies and tax credits to bolster semiconductor production and research activities at home, said the sources.”
Next, I get into why self-reliance is so core to contemporary CCP ideology.
For Beijing, relying on the Commerce Department’s goodwill to access critical technology is an unacceptable outcome and an anathema to current CCP ideology. The roots of the CCP’s fixation on developing homegrown technological capacity run deep, going all the way back to the early days of the USSR, when Lenin blamed Russia’s loss in WWI on technological backwardness. “It is necessary to master the highest technology or be crushed,” he declared after signing the Treaty of Brest-Litovsk that marked the end of Russia’s involvement in the war…
China’s leaders post-Mao have embraced a vision of, as Julian Gewirtz calls it, “actionable futurism — a futurism that required a response from the state and indeed shaped a wide range of important S&T policies.” In 1979, senior leaders Hu Yaobang, Hu Qiaomu and Deng Liqun said in conversation with Deng Xiaoping that “rapid development” of technology meant that “if we do not grasp the opportunity and catch up, our gap with developed countries in the economic and technological domains will grow larger” — a sentiment that Sullivan’s statement echoes. Just last week, Xi Jinping said in his Party Congress speech that “to meet China’s strategic needs, we will concentrate resources on original and pioneering scientific and technological research to achieve breakthroughs in core technologies in key fields.”
As such, the Biden administration, by explicitly attempting to contain Chinese technological development, hits a nerve for the CCP.
I conclude with what this means for the future of US-China policy.
Ultimately, the Biden administration’s actions have locked the U.S. and China on a course for years of vigorous tech competition. However, the quality of America’s execution of its strategy is very much in the air. Unfortunately, as Carnegie Mellon’s Erica Fuchs writes, “the intellectual underpinnings to inform technology strategy at the national level are limited.” State-of-the-art research on innovation cannot even settle on a definition of critical technology, much less help policymakers figure out which levers to pull. The U.S. must invest in giving policymakers a deeper understanding of both the Chinese science and technology policymaking ecosystem, as well as analytical foundations like possible technological futures and the ins and outs of critical supply chains. That will give the U.S. and its allies the best chance of maintaining their technological edge and securing the regional stability that lies at the heart of the Biden administration’s aims.
This analysis of the October EAR update is consistent with most western media reporting - that the updated sanctions: a) were a surprise to CCP leadership; and b) that it will cripple the mainland semi sector.
In fact, China has long anticipated a de-coupling of advanced chip fab technology (see the 2014 Guidelines to Promote National Integrated Circuit Industry) and, as is noted in this article, has invested heavily to minimize that risk, arguably with questionable ROI. But that doesn;t mean that China is not capable of independently developing sub-10nm fabs.
Another important factor is that China is the single largest chip market in the world - more than 50% of total demand. Once again, US foreign p[olicy has failed to consider the unintended consequences of sanctions, and that is reflected in equity prices for fabless US chip manufacturers. As with oil and other sanctions directed at Russia, there will be significant aftershocks when the dust settles on the latest EAR.
It's important to recognize that the bulk of chip demand is still with older 10+nm chips (mostly 28nm chips), where China has a compelling cost advantage that will likely be sustainable for some time to come. The recent sanctions were specifically intended to limit China's ability to develop local super-computer and AI capabilities that could be used ofr both commercial and military purposes. It's likely these sanctions will be successful in the short term.
But the real question is whether China has reached a critical mass in its ability to innovate (in the semi and other sectors). IMO, it has - if you consider its technical accomplishments in critical fields like space exploration, aeronautics, naval warfare, biotech, etc, there is a compelling argument that the latest sanctions are a day late and a dollar short.
TSMC introduced 7nm chip manufacturing in 2018 and is only now ramping up 3nm capabilities. SMIC already has limited 7nm capabilities - while China is behind the curve, it's doubtful it will take 10 years to catch up to today's state-of-the-art wafer manufacturing.
And finally, although many (most) advancements come from the private sector, if someone were to compute the number of Chinese ex-pat grad students in leading US STEM programs, it would be evident that the US has not (yet) cut-off China from basic research.
Like him or not, Chairman Xi plays the long game. And unlike the US, China's hi-tech strategy is coordinated, albeit rife with corruption. This topic should be revisited every 6 months to see if assumptions that were baked into US sanctions are durable over the long run. If they're anything like the logic that went into sanctions against Russia, we will see important unintended consequences from US actions taken in 2022.
deleted