The debate around the potential for conflict between the U.S. and China over Taiwan has reached disturbing levels, with recent articles suggesting the two countries could be “sleepwalking into conflict.” There have even been calls for the U.S. to consider bombing key Taiwanese chipmaking sites in the event of a Chinese invasion, an action former Trump administration national security adviser Robert O’Brien hinted may be necessary in a recent interview with The Wire China.
At a recent closed door event on Taiwan I attended at Brookings, a Washington think tank, discussions over the U.S.’s potential to intervene in a conflict ended with talk of China possibly using tactical nuclear weapons to hold off that prospect. I was struck, meanwhile, by the lack of consideration given to alternative approaches, such as coupling deterrence with assurance, as moderate observers have called for; or how to better integrate China into its regional economy via trade agreements, or other multilateral approaches that might raise the stakes of an invasion for Beijing.
Instead, many participants, though not all, channeled the deterrence-only approach of so-called “realists” such as former Trump administration officials Elbridge Colby and Matt Pottinger. This typically includes calls for doubling U.S. military spending, strengthening the defense-industrial base, and arming Taiwan to the teeth.
The deterrence-only discourse currently in vogue seems entirely inadequate to the moment — a 19th century discussion about great power projection in a 21st century world of global interconnection and complex, tightly coupled global technology supply chains.
Hence the 800-pound gorilla in the room — the inconvenient reality that the world depends heavily on Taiwan for the most advanced semiconductors driving the AI revolution, and that supply chains for most other types of semiconductors have some critical nexus with the island — was given short shrift. One participant even commented that semiconductors are important, “but not everything.”
In fact, Taiwan, semiconductors and TSMC are, if not everything, close to everything. By any measure, Taiwan’s unique position in global IT supply chains is the single most important development since the historic documents governing U.S., China and Taiwan relations were first hammered out in the 1970s and 80s.
The emergence of an industry on Taiwan whose cutoff from the global economy would instantly create massive economic ripple effects, disrupting shipping, food, energy and financial markets, remains extraordinary.
When I first went to Taiwan in 2003, the island was primarily a key producer of components used in desktop computers. Things have certainly changed since then. U.S. officials now often repeat a salient data point to illustrate the importance of global chip foundry leader Taiwan Semiconductor Corporation (TSMC), namely that it now manufactures over 90 percent of advanced node semiconductors. These are the chips that power our smartphones, laptops, desktops, tablets, and EVs, and are used around the world in data centers to train advanced AI models.
But Taiwan’s business ecosystem is about much more. Taiwanese companies such as UMC are also major players in manufacturing mature node semiconductors, while design firm Mediatek and its thousands of engineers play a key role in second tier mobile phones and many other devices.
…when Taiwan watchers dismiss the technology component in discussions about deterrence as being about “mere semiconductors”… this is at best a misguided downplaying of a key issue…
Taiwanese firms are also key players in advanced packaging, while the island is home to the R&D operations of all major computer and server firms, as well as leading semiconductor manufacturing tool makers. The island’s importance to the global tech industry was on full display when it hosted the Computex conference in early June, an event the bosses of Intel, AMD, Qualcomm, Nvidia, and dozens of others all attended.
AMD’s CEO Lisa Su (top left), Qualcomm’s CEO Cristiano Amon (top right), Nvidia’s CEO Jensen Huang (bottom left), and Intel’s CEO Pat Gelsinger (bottom right) present during the Computex conference, June 2024, in Taipei. Credit: Computex
Recently, the likes of Rhodium and Bloomberg have attempted to quantify the global economic impact of a blockade or broader conflict involving Taiwan, with numbers ranging from $5 to $15 trillion. Such figures suggest that any conflict, from a prolonged blockade to an actual kinetic exchange, would be catastrophic for the global economy.
Yet even these numbers may vastly underestimate the potential collateral damage. A prolonged conflict could result in the irreparable disruption of the business environment in Taiwan, and the partial or complete scattering of the critical mass of companies, researchers, knowledge, and chemistry that has driven the global IT sector for the last two decades. Discussions with industry insiders suggest that it would take decades and tens of trillions of dollars to recreate such an environment anywhere else in the world, even if it could be accomplished.
Other effects could be swift and devastating. To cite just one example, the current market cap of the three most valuable companies globally — Microsoft, Apple, and Nvidia — is close to $10 trillion. All are highly dependent on continued daily access to Taiwan, as are at least half of the companies in the S&P 500 index. The potential for a devastating global equity market crash is extremely high.
So when Taiwan watchers dismiss the technology component in discussions about deterrence as being about “mere semiconductors,” or state that “TSMC is just one company,” this is at best a misguided downplaying of a key issue — and at worst, a woefully inadequate acknowledgement of how critical the technology portion of the complex issues across the Taiwan Strait is.
When I wrote a report about the Geopolitics of Semiconductors in 2020, the few people aware of TSMC’s critical role in global supply chains were industry regulars. Four years later that has changed dramatically. But the greater awareness of Taiwan’s importance has come alongside ever more discussion around the potential for conflict between the United States and China over the island.
Some, including myself, have become concerned that U.S. policy could undermine Taiwan’s “Silicon Shield” and increase the potential for conflict — because as U.S. export controls made it harder for Chinese manufacturers to obtain Taiwanese chips, they would aim more for self-reliance and less dependence on Taiwan for advanced semiconductors.
By contrast, other commentators have begun speculating on the potential for the U.S. to consider disabling and even bombing TSMC foundry facilities in the event of a Chinese military move against Taiwan. In April, historian Niall Ferguson claimed that the U.S. government would act to destroy the TSMC foundries in such an event. Trump administration officials such as former national security advisor Robert O’Brien have also speculated on this issue.
Such statements display a lack of knowledge of semiconductor industry supply chains, and also anger officials in Taiwan who say they would do their best to prevent anyone from damaging TSMC foundries in a conflict scenario. Proposing the bombing of TSMC and other sites borders on the irresponsible and demonstrates a lack of awareness of economic realities.
Back at the D.C. think tank conference, much emphasis was placed on maintaining the status quo, while ensuring that Beijing would not conclude that the potential for peaceful unification has not been foreclosed. The Biden administration has stuck to the line that the U.S. supports the One China Policy — which acknowledges that Taiwan is part of China while not accepting Beijing’s view of Chinese sovereignty — while stressing U.S. support for a peaceful resolution of Taiwan’s status.
Meantime, one glaring omission from the current discourse concerns the U.S.’s extraterritorial export controls that have progressively cut Chinese firms operating in sensitive sectors such as AI off from using TSMC to manufacture their designs.
Inside TSMC’s 12 inch wafer fab. Credit: TSMC
What is the long-term impact of these controls? How would they be implemented in the event of a change in Taiwan’s status? U.S. policy has to date not considered these questions adequately. A rethink is critical given that in the event of peaceful unification, issues around sovereignty and export controls involving Taiwan and TSMC would be significant. The U.S. and like minded allies, for example, could not quickly and plausibly impose rules that cut off leading Western firms from using TSMC foundry services: that would have the same effect as a conflict-induced forced cut off.
…we appear stuck in an endless cycle of crisis and response, with military exercises bringing the increased potential for an accident to incite a wider conflict…
These are the types of issues that serious analysts of Taiwan should be thinking about: how to begin building new structures and understandings that help avoid future conflict, while considering what would happen in the event of a peaceful solution of the Taiwan issue — something the U.S. government, officially at least, supports.
Finally, what is clearly missing from the discussion is consideration of a new framework around U.S.-China-Taiwan relations, which acknowledges the current situation, what has changed since 1982, and how to think about rebuilding trust and real peace and stability across the Taiwan Strait. Without this, we appear stuck in an endless cycle of crisis and response, with military exercises bringing the increased potential for an accident to incite a wider conflict, accusations of ‘salami tactics’ being used on both sides, and each party continuing to stick to their preferred interpretation of the status quo.
This is not a recipe for stability, and will only lead to further attempts to de-risk global IT supply chains — attempts that will remain impossible to fully realize, given that the concentration of advanced node production at TSMC will persist for the next decade and beyond.
One approach hinted at by Taiwan’s new President Lai Ching-te is to consider its advanced semiconductor industry capacity as a global asset, which needs to be protected at all costs. Still, the real issue remains how to avoid conflict in the first place. This will require significant new, out-of-the-consensus thinking — starting with an acknowledgement that yes, semiconductors matter, and matter profoundly, to Taiwan, the U.S., and to the global economy.
Paul Triolo is Senior Vice President for China and Technology Policy Lead at Albright Stonebridge Group. DGA Albright Stonebridge Group works with a variety of companies in many sectors, including healthcare, fashion, agriculture, and technology. The firm advises companies across the technology stack, including working with semiconductor firms, consumer electronics companies, social media companies, firms part of the Al supply chain, and firms in key verticals deploying Al cross their business operations. Mr. Triolo has been writing regularly on technology policy related issues, including export controls, since 2016.