Designing Out U.S. Technology

Commentary by William Alan Reinsch

Published October 15, 2024

When the Biden administration imposed new export controls on certain semiconductors and their manufacturing tools in October 2022, they raised three issues-the effect on U.S. company revenues, the effect on China's technology policy, and whether they would incentivize other countries, including China, to "design out" U.S. technology in order to avoid the controls.

Now that two years have passed, answers to the first two questions are coming into view. The revenue impact was small on chipmakers, larger on tool manufacturers, and will get bigger over time as more advanced chips fall within the scope of the rules. The controls did not change China's policy of developing its own technology in this sector, but they no doubt accelerated China's efforts.

The answer to the third question remains more of a mystery, and it is one the Scholl Chair is studying. We are researching four parts of the semiconductor supply chain: packaging, manufacturing tools, subcomponents, and design. The first report, on packaging, was published recently and is the focus of this week's column.

First, the papers look at two alternative responses to the export controls. One, designing out, means finding and adopting alternative sources for U.S. technology. The other, designing around, means developing new technologies that bypass U.S. technology. Either, if successful, will lead to a shift of semiconductor supply chains away from reliance on the United States, which means there will also be significant negative revenue consequences for U.S. firms.

Packaging used to be the last stop in the fabrication process, a relatively low-value step compared to the rest of the supply chain and often outsourced to third parties. Advanced packaging, in contrast, is becoming increasingly important as a means of developing more advanced chips and as a way to get around limitations on chip density. As chip construction approaches the 1-2 nanometer level, the path to further development will lie in stacking multiple connected chips in new architectures-chiplets-that will be enabled by advanced packaging. Equally important, chiplet architecture also enables the assembly of packages of less capable chips which, working together, can match or exceed the capabilities of the most advanced semiconductors. As the CSIS paper says,

Importantly, chiplet technology can be used to create microelectronic systems that perform like those containing leading-edge semiconductors-without any access to the advanced manufacturing technology required to make such chips.

Advanced packaging technologies provide an opportunity for China to design around U.S. export controls on advanced chips-bypassing U.S. technology controls by developing novel production capabilities. Chinese firms are already implementing such pathways. In 2022, Huawei and HiSilicon started exploring 3D chip stacking designs, an advanced technique that vertically integrates multiple wafer dies into one package as a way to navigate U.S. sanctions.

In 2022, YMTC, the leading Chinese manufacturer of NAND flash memory-a semiconductor that retains data without power, widely used in memory cards and solid-state drives-utilized advanced packaging techniques to create a cutting-edge memory chip. The packaging architecture, known as Xtacking, incorporates hybrid bonding, a hallmark of advanced packaging processes, showcasing China's dedication to leveraging packaging for innovation.

One of the most concerning instances of Chinese companies using advanced packaging to circumvent U.S. restrictions is the Jasminer X4, a cryptocurrency mining chip. This chip effectively utilizes DRAM-to-logic hybrid bonding, a packaging technique that stacks advanced logic chips on top of memory chips to enhance performance and reduce energy consumption.

So, at least in advanced packaging, it appears China is making significant efforts to design around Western technology. Subsequent papers will look at other parts of the supply chain and, I suspect, will come to similar conclusions.

This is important because the outcome of Chinese efforts has serious implications for our export control policies. The emerging consensus is that the U.S. controls have served their purpose in the short term by holding back Chinese efforts to advance in the semiconductor sector. Long-term consequences are less certain, but CSIS research thus far suggests that over the long term-which in this sector is only a few years-the costs to U.S. industry may do more harm to our security than the benefits the controls provide. That is because the United States' ability to run faster-to stay ahead of our adversaries technologically-is directly related to the economic health of its companies. The private sector is the primary innovator in critical technology sectors, and, simply put, if it is not making money, it is not investing in research and development for next-generation technologies-the ones that will keep us ahead of China.

Designing out and designing around, along with the expanding scope of U.S. controls, will have a negative impact on the domestic industry as China develops alternatives to its reliance on Western chips and tools. Maintaining leadership in advanced packaging in particular will be an essential part of maintaining overall leadership in the semiconductor sector, which means that at some point a serious review of U.S. export controls will be necessary.

William A. Reinsch holds the Scholl Chair in International Business at the Center for Strategic and International Studies in Washington, D.C.

Programs & Projects