A Compound Problem: Dialing in U.S. Semiconductor Strategy

Semiconductors are center stage in today’s U.S.-China geopolitical competition. And both sides are firing. The CHIPS Act, which just turned one year old, stands as the marque measure of U.S. initiative in the fight.

At the same time – less obviously but also more significantly – Beijing is continuing to invest in domestic capacity in order to increase both competitiveness and control. And those domestic development efforts are being led by none other than Huawei, perhaps the Chinese State champion most explicitly targeted by U.S. high-tech regulatory action over the past few years. Of late, for example, Chinese press has celebrated Huawei’s progress in establishing production capacity for semiconductors in Wuhan’s Optics Valley: “In the storm of sanctions around the world, it is still able to hold its ground and fight unremittingly for technological innovation. Its latest feat - the establishment of a new chip factory in China - marks the opening of a new chapter in Huawei's self-developed chips.”

Clearly, U.S. strategy needs better to appreciate Beijing’s competitive positioning, industrial ecosystem, and policy – including how tradeoffs are made in Chinese industrial policy and investment between legacy and emerging, upstream and downstream. Take, for example, the question of China’s recent export restriction moves, and how the PRC might build on them. Beijing assesses that its dominance in semiconductor raw materials like gallium and germanium can strike fear into global leaders (governments and companies) in elemental, silicon-based chip manufacturing. Beijing also assesses that that dominance positions China to leapfrog in compound and optical semiconductors, like those Huawei is churning out in Wuhan. So even as Beijing strikes tit for tat with export restrictions, it also invests in the newer, more emerging market of compound and optical semiconductors.

But understanding China’s gambit is only the first step. U.S. strategy needs also to update and pursue a competitive strategy for both defending against Chinese co-option of existing markets and seizing new markets. The CHIPS Act and its encouragement of new fab proposals – like those that have been floated by the larger, silicon-based (or elemental) semiconductor companies like Intel, TSMC, and others – is a great initial mark of progress. But CHIPS act implementation needs also to recognize competitive opportunity provided by equally critical projects for national security focused on compound and optical semiconductors.  These compounds such as indium phosphide or gallium arsenide use light and electricity through the creation of Photonic Integrated Circuits (PICs). PICs form the heart of critical infrastructure such as subsea cables for telecommunications and data centers that are the backbone of AI infrastructure. These projects don’t require billions of dollars like new elemental fabs, but they do require substantial policy support and investment to compete against China’s highly subsidized industry in the immediate.

While Washington twiddles its thumbs and figures out potential plans for implementing CHIPS program activities, China is forging ahead. That forge is being led by known bad actors, like Huawei, and all along the value chain of elemental and compound semiconductors. And it’s already bearing fruit: Huawei’s recent investment in Yunnan Xinyao Semiconductors Materials Co., Ltd. is producing 150,000 wafers per year for over 90 clients.  American production capability for the same is severely limited. The competitive implications are made clear by the fact that Huawei’s bid is clearly framed by Chinese industry analysts as focused on supply chain security: “Huawei’s investment in Yunnan Xinyao Semiconductor Materials Co., Ltd. is to safeguard its supply chain security in semiconductors for optical communication.” Without a U.S. competitive response, subsea cables and AI infrastructures that depend on these inputs will be in a worse spot than the networking gear that Huawei previously came to dominate globally.

Congress recently highlighted the importance of this material on a bipartisan, bicameral basis during debate on the National Defense Authorization Act.  Senator Padilla and CHIPS act author Senator Cornyn and Congressmen Jimmy Panetta, Rob Wittman, and Susan Wild introduced an amendment requesting a study on the use of indium phosphide PICs in national security supply chains. In particular, the study is meant to examine PICs for use in secure communications, quantum information, AI, LIDAR, and directed energy weapons. Those applications make clear that this battle is an important one for both commercial and security competitions against China.  

Congress can also increase pressure on China’s core semiconductor nodes with moves like expanding Section 889 and 5949 enforcement and authority to a broader swathe of Chinese champions, such as Huawei and its subsidiaries, including HiSilicon, and its investments, including the stake Huawei holds in Yunnan Xinyao Semiconductors Materials Co., Ltd. More strategic targeting of export restriction designations and outbound investment screening are just as necessary for defending against China’s extant positioning and future direction.                                   

The CHIPS act and Congress must continue to compete all along the semiconductor value chain while finding opportunities for outsize impact in the supply chain for semiconductors, including in compound materials. Without quick and strategic action, the semiconductor race will be over before it begins.  

Emily de La Bruyère and Nathan Picarsic are Senior Fellows at the Foundation for Defense of Democracies and co-founders of Horizon Advisory.