Who is winning the chip wars? With Chris Miller

Episode Summary

Executive Summary: The episode examines the global chip war and argues the U.S. remains ahead but less secure than five years ago, mainly because China has poured state-backed investment into semiconductors. The discussion centers on Taiwan’s centrality, AI and military dependence on advanced chips, U.S. diversification efforts via the CHIPS Act and export controls, and China’s push for self-sufficiency.

Main Topics: Who is winning the chip war? (Priority: 5/5): Chris Miller rates the U.S. as still strong but weaker than five years ago, moving from an '8' to a '6' because China has become a more effective competitor. He stresses that the real contest is between an established international supply chain and China’s rising domestic capabilities. Taiwan’s central role and systemic risk (Priority: 5/5): The conversation explains that Taiwan produces most of the world’s most advanced chips and a large share of lower-end chips, making any disruption there a global economic shock with severe effects on consumer devices, manufacturing, and data centers. AI, military power, and strategic dependence (Priority: 5/5): The hosts explore why governments care about chips: advanced semiconductors are essential for training AI systems and building autonomous military and intelligence capabilities, so chip access translates directly into strategic advantage. China’s industrial policy and self-sufficiency drive (Priority: 4/5): China’s chip strategy combines heavy public investment funds and pressure on domestic firms to buy local chips. The goal is both greater domestic production and reduced dependence on foreign imports, though catching the cutting edge remains difficult. U.S. policy: CHIPS Act, diversification, and export controls (Priority: 5/5): The CHIPS Act is presented mainly as a diversification policy to reduce reliance on Taiwan rather than a pure technology-leadership strategy. Export controls on advanced tools aim to slow China’s progress, while tariffs and restrictions may be used to curb lower-end dependence on Chinese chips. International alignment and the durability of controls (Priority: 3/5): Despite occasional political friction, the U.S., Japan, the Netherlands, Korea, and Taiwan share security concerns about China. That alignment, plus overlapping corporate interests, makes the current export-control coalition more durable than it may appear. Supply-chain resilience vs. ‘insurance policy’ (Priority: 4/5): Miller argues chip subsidies should be seen as insurance against a catastrophic Taiwan Strait disruption rather than a conventional industrial policy success metric. The spending may look inefficient if no crisis occurs, but it is intended to prevent trillion-dollar losses in a worst case.

Key Arguments: The U.S. is still ahead in chips, but China has narrowed the gap through sustained industrial policy and large-scale state investment. Taiwan is the critical chokepoint in the global chip system; a loss of access would rapidly disrupt electronics, data centers, and general manufacturing. Advanced chips are a strategic resource because chip quality determines AI performance, which in turn affects economic, military, and intelligence power. China’s challenge is harder than simply building more fabs because the cutting edge keeps moving as firms like NVIDIA and TSMC push it forward. The CHIPS Act is mainly about shifting production geography and reducing concentration risk, not bringing every leading-edge chip fully onshore. Export controls have likely slowed China’s progress at the high end, but they are less effective for lower-end tools because those supply chains are less concentrated. The West’s chip strategy is based on maintaining a diversified, allied supply chain rather than fully nationalizing semiconductor production. China’s low-end chip self-sufficiency goal is more achievable than its high-end leadership goal because commodity manufacturing is easier to replicate than frontier innovation. A lot of what looks like irreplaceable supply-chain dependence can be mitigated by market adjustment, but Taiwan-related chip dependence is unusually inelastic and dangerous. Chip subsidies should be understood as purchasing insurance against rare but catastrophic geopolitical disruption, not as a standard ROI-driven investment program.

Data Points: U.S. chip-war position five years ago: 8/10 - Miller’s retrospective rating of the U.S. position in the chip war five years earlier. U.S. chip-war position now: 6/10 - Miller’s current rating of the U.S. position due to stronger Chinese competition. Taiwan share of most high-end processor chips: Around 90% - Taiwan’s production share of the most advanced processor chips. China’s chip imports vs. oil imports: Equal spending each year - China spends as much importing chips as it spends importing oil. 2014: First national integrated circuit investment fund established - Marks the start of China’s major national-level chip investment push. U.S. tariff on chips from China: Doubled earlier this year - The Biden administration increased tariffs on imported Chinese chips. Chip fab build time: A couple of years - Time needed to build a single chip-making fab, illustrating slow capacity replacement. Cost penalty to build in the U.S./Europe vs. Taiwan/Korea: 20%–40% - Higher production costs outside the established East Asian chip ecosystem. High-end lithography market share: 100% - ASML’s share of the high-end lithography tool market. Forecast horizon for China’s chip market share growth: Next half decade - Best available forecasts suggest China’s chip output share will continue rising for several years. Scale of Chinese national integrated circuit investment funds: Three national funds, each about the size of the U.S. CHIPS Act - Illustrates the scale of China’s central government spending on semiconductors. Potential disruption from Taiwan crisis: Trillions of dollars - Estimated cost to the world economy from a major Taiwan Strait disruption. Neon gas supply shock context: 50% reduction - Example of a mineral supply shock that did not cause major disruption because supply adjusted quickly.

Pivotal Quotes: "It was an eight before, it's in a six right now because China's been a very effective competitor." — Chris Miller: Describing the U.S. position in the chip war over the last five years. "If there were a conflict in the Taiwan Straits, that would be knocked offline, and the U.S. wouldn't be able to access the chips that it needs from Taiwan." — Chris Miller: Explaining the U.S. military and strategic vulnerability tied to Taiwan-based production. "The key question is: where do you have supply that's really inelastic and demand that's really inelastic." — Chris Miller: Summing up how to think about dangerous supply-chain dependencies.

Implications: The chip war is a long-term contest, not a one-off race. Expect more subsidies, export controls, tariffs, and allied coordination as governments treat semiconductors as strategic insurance against Taiwan-related and AI-era shocks.