Pages

6 November 2022

Assessing the New Semiconductor Export Controls

Matthew Reynolds

On Friday, October 7, the U.S. Commerce Department announced a sweeping list of new export controls targeting China’s semiconductor and supercomputing industries in two separate rules. In a four-page summary, the Commerce Department’s Bureau of Industry and Security (BIS) described the rules as intended to restrict China’s “ability to both purchase and manufacture certain high-end chips used in military applications.” The new controls impose unprecedented licensing requirements on the sale of advanced chips, software, and—perhaps most critically—a wide range of semiconductor manufacturing equipment (SME) used to produce advanced chips. The controls are remarkably comprehensive in scope. They require exporters of SME, providers of the parts needed to produce or maintain SME, and even certain U.S. persons who wish to support the production or development of semiconductors in China to obtain permission from the U.S. government before doing so.

Q1: What’s behind the Biden administration’s new export controls?

A1: On September 16, U.S. national security advisor Jake Sullivan, in remarks at the Special Competitive Studies Project (SCSP) Global Emerging Technologies Summit, announced a major shift in the strategy underpinning U.S. export controls against China. Sullivan stated that the United States would no longer focus primarily on using export controls to maintain “‘relative’ advantages over competitors in certain key technologies.” Instead, he declared that the United States “must maintain as large of a lead as possible.” The previous paradigm of export controls had allowed adversaries to move up the critical technology frontier at a deliberate lag of a few generations behind the United States. The U.S. government now appears intent on blocking China’s ability to advance in certain semiconductors that have specific critical applications. This strategic shift is taking place against the backdrop of China devoting immense resources to advance its semiconductor sector, with some success, and implementing its military-civil fusion doctrine, making it increasingly difficult to distinguish civilian end-users from military end-users.

Q2: What are the supply chain chokepoints that the Biden administration aims to exploit?

A2: The most recent rounds of controls take aim at China’s weaknesses in the software used to design chips, often referred to as electronic design automation (EDA), and sophisticated SME. Although the U.S. share of chip manufacturing has declined over the last three decades, U.S. firms continue to hold dominant positions in both EDA and critical areas of SME. Without access to U.S.-owned EDA and SME, Chinese chip manufacturers are not able to manufacture advanced chips domestically. The controls on SME and EDA, taken together with controls placed on the most advanced chips used to power supercomputers and AI algorithms, constitute a two-pronged attack on China’s semiconductor industry—focused at one end on keeping the most advanced chips out of China, and at the other on preventing China from developing the capability to manufacture leading-edge chips.

Q3: Do the new controls represent a significant shift in U.S. export control strategy?

A3: The new controls mark a significant departure from the paradigm that has guided U.S. export controls for the last two and a half decades. While the Trump administration’s success in pressuring the Dutch government to block the export to China of extreme ultraviolet (EUV) lithography machines—a piece of equipment made by a single Dutch company, ASML, needed to manufacture the most advanced chips—set a de facto limit on China’s semiconductor manufacturing capabilities, there was little evidence of a systematic strategy to prevent China from advancing its semiconductor manufacturing capabilities beyond a certain technological threshold. By contrast, the Biden administration’s controls explicitly establish a point beyond which they seek to stop China’s advancement, and importantly, in a departure from previous measures, actively seek to degrade China’s current chipmaking capabilities, not just forestall their advance.

Q4: How has Beijing responded to the new controls?

A4: So far, there has been little to no response from Beijing. This may in part be because the controls were announced in the run-up to the 20th Party Congress, China’s most important political event. The party report delivered by Xi Jinping on October 16 did, however, make several references to the need for self-reliance in science and technology. While Beijing denounced the new controls as a form of “technological bullying,” the Chinese government has few policy options in the near term. The party is already lending significant financial and political support to its domestic semiconductor industry, and the recent crackdown on corruption in the semiconductor industry, which has seen many of the sector’s leading executives imprisoned, likely further complicates Beijing’s ability to respond. While Beijing could retaliate by harassing foreign firms or, as some have suggested, cutting off exports of critical minerals, China has little to gain from further escalation with the United States, especially given its current troubled economic situation. If anything, the recent controls will only confirm Beijing’s belief, dating at least back to the Trump administration’s controls on Huawei and ZTE, that the United States is seeking to contain China and block its rise, likely causing Beijing to double down on its own drive for self-reliance in chips production.

Q5: What risks are entailed by a more aggressive export control strategy?

A5: The Biden administration faces difficult trade-offs in deciding where to draw the line with controls. The October controls target equipment and technology needed to manufacture chips at and below the 14–18 nanometer (nm) process node (as a rule of thumb, the smaller the node, the more advanced the chip). While China currently accounts for 15 percent of all chips made at or above the 28 nm processing node, by one estimate, Beijing’s investment drive is expected to boost this number to 40 percent by 2025. Many consumer electronics use these legacy chips rather than those at the leading edge. The challenge for U.S. policymakers, then, will be to avoid disrupting the global supply of legacy chips, while also preventing China from acquiring advanced chips and the ability to manufacture them. Cutting China off at too large a process node, or spooking the foreign firms that China relies on for SME, could undermine China’s ability to manufacture legacy chips, exacerbating supply chain issues at a time when inflation is at its highest level in decades.

Beyond the economic repercussions, excessive controls also carry strategic risks. The controls may rankle U.S. allies and partners like Japan, South Korea, and Taiwan who themselves are suppliers of SME and producers of advanced semiconductors, potentially exacerbating these diplomatic tensions is the unilateral nature of controls. Loss of revenue could also hurt the ability of U.S. and allied firms to innovate. For the major U.S. SME firms—LAM Research, Applied Materials, and KLA—China is the largest market by revenue. Every year, firms like these reinvest billions of profits from legacy sales to stay competitive at the leading edge. While the previous export control paradigm allowed for this business model, cutting off China at too large of a node could undermine it, and paradoxically hurt U.S. innovation. At the same time, doing so could spur innovation by China’s domestic semiconductor industry. Despite Beijing’s efforts to promote its national SME firms, China’s chipmakers have long shown a clear preference for top-tier equipment from the most advanced foreign firms. But as U.S. regulations increasingly threaten to upend those supplier relationships, Chinese semiconductor manufacturers will have little choice but to rely on domestic suppliers. Stricter U.S. controls may hobble China’s competitiveness in the short to medium term, but, in the long run, more revenues flowing to China’s SME base from domestic semiconductor manufacturers could drive innovation.

No comments:

Post a Comment