人間魚詩生活誌 / 人間魚詩社
  • 首頁
  • 金像獎詩人百萬賞
    • 徵競逐辦法
    • 心中理想的徵獎
  • 主題特輯
    • 人物報導
    • 特別企劃 >
      • 打破矽盾問到底 Kolas X《晶片戰爭》作者Chris Miller >
        • Kolas與Chris Miller的下課時間
        • 「矽盾」救台灣? 晶片戰爭作者Chris Miller:遠遠不能保證
        • 守住王牌
      • KOLAS ANd FRIENDS -Breaking Through the 'Silicon Shield' >
        • Breaking Through the 'Silicon Shield': Kolas Yotaka Talks with Chris Miller
        • Can the 'Silicon Shield' Save Taiwan? ‘Chip War’Author Chris Miller:“Far From Guaranteed”
        • Hold on to the Trump Card
      • 詩電影珍藏特輯
      • 與民主同行——美國總統大選後的台灣與烏克蘭
      • 走在甘迺迪街 ――Volodymyr與Kolas的對話
      • 如果美國的國家利益 與我們的不同?
      • Walking towards Democracy: Taiwan and Ukraine after the U.S. Presidential Election
      • A Stroll Along Kennedy Street: A Conversation Between Volodymyr and Kolas
      • What If America's National Interest Is Different From Ours?
    • 攝影詩
    • 談詩論詩 >
      • 鄭慧如
      • 孟樊
      • 張日郡
      • 傅詩予
      • 溫任平
      • 吳長耀
    • 為何是/不是XX圖鑑 >
      • 寫真後話—— 專訪攝影藝術家 瀧本幹也
  • 詩生活誌
    • 2025 Mar. Vol.20
    • 2024 Dec. Vol.19
    • 2024 特刊
    • 2024 Sep. Vol.18
    • 2024 Jun. Vol.17
    • 2024 Mar. Vol.16
    • 2023 Oct. Vol.15
    • 夏季號 2023 Jul. Vol.14
    • 春季號 2023 Apr. Vol.13
    • 冬季號 2023 Jan. Vol.12
    • 秋季號 2022 Oct. Vol.11
    • 夏季號 2022 Jun. Vol.10
    • 春季號 2022 Apr. Vol.09
    • 冬季號 2022 Jan. Vol.08
    • 秋季號 2021 Oct. Vol.07
    • 夏季號 2021 Jul. Vol.06
    • 過往季刊 >
      • 冬季號 2022 Jun.Vol.08
      • 秋季號 2021 Oct.Vol.07
      • 秋冬號 2020 Dec.Vol.05
      • 春夏號 2020 Jun.Vol.04
      • 秋冬號 2019 Dec.Vol.03
      • 夏季號 2019 Aug.Vol.02
      • 創刊號 2019 Apr.Vol.01
      • 368 人間唸詩趣
  • 月電子詩報
    • 第064期
    • 第063期
    • 第062期
    • 第061期
    • 第060期
    • 第059期
    • 第058期
    • 第057期
    • 第056期
    • 第055期
    • 第054期
    • 第053期
    • 第052期
    • 第051期
    • 第050期
    • 第049期
    • 第048期
    • 第047期
    • 第046期
    • 第045期
    • 第044期
    • 第043期
    • 第042期
    • 第041期
    • 第038、039、040期
    • 第035、036、037期
    • 第032、033、034期
    • 第029、030、031期
    • 第027、028期
    • 第025、026期
    • 第023、024期
    • 第021、022期
    • 第020期
    • 第019期
    • 第018期
    • 第017期
    • 第016期
    • 第015期
    • 第013、014期
    • 第012期
    • 第011期
    • 第010期
    • 第009期
    • 第008期
    • 第007期
    • 第006期
    • 第004-005期
    • 第003期
    • 第002期
    • 第001期
  • 年度金像獎詩人
    • 最新公告
    • 歷屆辦法
    • 第二屆頒獎典禮
    • 第三屆頒獎典禮
    • 第四屆頒獎典禮
  • 插畫詩
  • 實驗性新詩型
    • 什麼是實驗性新詩型六行詩?
    • 徵稿公告
    • 第一波徵稿入選詩作
    • 第二波徵稿入選詩作
    • 第三波徵稿入選詩作
    • 第四波徵稿入選詩作
  • 詩電影
    • 為什麼我主張拍詩電影
    • 詩電影《穿過日影的翅膀》
    • 詩電影《刮傷廚房》
    • 詩電影《大見解》
    • 詩電影《本月》
    • 詩電影《權術》
    • 詩電影首映會 映後座談
  • 專欄
    • 我們賴以生存的虛擬世界
    • 肌肉的記憶
    • 不要自卑
    • Kolas Yotaka
    • 蕭義玲
    • 蕭伶伃
    • 蕭伶妤
    • 楊姍樺
    • 黃文成
    • 愛/人間/KOLAS 強大的女孩
    • 愛/人間/KOLAS 做人類還有何意義?
    • 為地方點上創意的亮點—— 卜唯平與「禾翌創意」及與台灣各地農會的故
    • 那日,與女神鑾轎一同疾行
    • 那些回不去的地方 蕭義玲
    • 黃文成
    • KOLAS YOTAKA
    • 蕭伶伃
    • 王婉育
    • 蕭伶伃
    • 傅雅雯
    • 蕭伶伃
    • 楊姍樺
    • 田原
    • 郭瀅瀅
    • 林彧
    • 許丁江
  • 詩人專區
  • 商品
  • 關於
  • 首頁
  • 金像獎詩人百萬賞
    • 徵競逐辦法
    • 心中理想的徵獎
  • 主題特輯
    • 人物報導
    • 特別企劃 >
      • 打破矽盾問到底 Kolas X《晶片戰爭》作者Chris Miller >
        • Kolas與Chris Miller的下課時間
        • 「矽盾」救台灣? 晶片戰爭作者Chris Miller:遠遠不能保證
        • 守住王牌
      • KOLAS ANd FRIENDS -Breaking Through the 'Silicon Shield' >
        • Breaking Through the 'Silicon Shield': Kolas Yotaka Talks with Chris Miller
        • Can the 'Silicon Shield' Save Taiwan? ‘Chip War’Author Chris Miller:“Far From Guaranteed”
        • Hold on to the Trump Card
      • 詩電影珍藏特輯
      • 與民主同行——美國總統大選後的台灣與烏克蘭
      • 走在甘迺迪街 ――Volodymyr與Kolas的對話
      • 如果美國的國家利益 與我們的不同?
      • Walking towards Democracy: Taiwan and Ukraine after the U.S. Presidential Election
      • A Stroll Along Kennedy Street: A Conversation Between Volodymyr and Kolas
      • What If America's National Interest Is Different From Ours?
    • 攝影詩
    • 談詩論詩 >
      • 鄭慧如
      • 孟樊
      • 張日郡
      • 傅詩予
      • 溫任平
      • 吳長耀
    • 為何是/不是XX圖鑑 >
      • 寫真後話—— 專訪攝影藝術家 瀧本幹也
  • 詩生活誌
    • 2025 Mar. Vol.20
    • 2024 Dec. Vol.19
    • 2024 特刊
    • 2024 Sep. Vol.18
    • 2024 Jun. Vol.17
    • 2024 Mar. Vol.16
    • 2023 Oct. Vol.15
    • 夏季號 2023 Jul. Vol.14
    • 春季號 2023 Apr. Vol.13
    • 冬季號 2023 Jan. Vol.12
    • 秋季號 2022 Oct. Vol.11
    • 夏季號 2022 Jun. Vol.10
    • 春季號 2022 Apr. Vol.09
    • 冬季號 2022 Jan. Vol.08
    • 秋季號 2021 Oct. Vol.07
    • 夏季號 2021 Jul. Vol.06
    • 過往季刊 >
      • 冬季號 2022 Jun.Vol.08
      • 秋季號 2021 Oct.Vol.07
      • 秋冬號 2020 Dec.Vol.05
      • 春夏號 2020 Jun.Vol.04
      • 秋冬號 2019 Dec.Vol.03
      • 夏季號 2019 Aug.Vol.02
      • 創刊號 2019 Apr.Vol.01
      • 368 人間唸詩趣
  • 月電子詩報
    • 第064期
    • 第063期
    • 第062期
    • 第061期
    • 第060期
    • 第059期
    • 第058期
    • 第057期
    • 第056期
    • 第055期
    • 第054期
    • 第053期
    • 第052期
    • 第051期
    • 第050期
    • 第049期
    • 第048期
    • 第047期
    • 第046期
    • 第045期
    • 第044期
    • 第043期
    • 第042期
    • 第041期
    • 第038、039、040期
    • 第035、036、037期
    • 第032、033、034期
    • 第029、030、031期
    • 第027、028期
    • 第025、026期
    • 第023、024期
    • 第021、022期
    • 第020期
    • 第019期
    • 第018期
    • 第017期
    • 第016期
    • 第015期
    • 第013、014期
    • 第012期
    • 第011期
    • 第010期
    • 第009期
    • 第008期
    • 第007期
    • 第006期
    • 第004-005期
    • 第003期
    • 第002期
    • 第001期
  • 年度金像獎詩人
    • 最新公告
    • 歷屆辦法
    • 第二屆頒獎典禮
    • 第三屆頒獎典禮
    • 第四屆頒獎典禮
  • 插畫詩
  • 實驗性新詩型
    • 什麼是實驗性新詩型六行詩?
    • 徵稿公告
    • 第一波徵稿入選詩作
    • 第二波徵稿入選詩作
    • 第三波徵稿入選詩作
    • 第四波徵稿入選詩作
  • 詩電影
    • 為什麼我主張拍詩電影
    • 詩電影《穿過日影的翅膀》
    • 詩電影《刮傷廚房》
    • 詩電影《大見解》
    • 詩電影《本月》
    • 詩電影《權術》
    • 詩電影首映會 映後座談
  • 專欄
    • 我們賴以生存的虛擬世界
    • 肌肉的記憶
    • 不要自卑
    • Kolas Yotaka
    • 蕭義玲
    • 蕭伶伃
    • 蕭伶妤
    • 楊姍樺
    • 黃文成
    • 愛/人間/KOLAS 強大的女孩
    • 愛/人間/KOLAS 做人類還有何意義?
    • 為地方點上創意的亮點—— 卜唯平與「禾翌創意」及與台灣各地農會的故
    • 那日,與女神鑾轎一同疾行
    • 那些回不去的地方 蕭義玲
    • 黃文成
    • KOLAS YOTAKA
    • 蕭伶伃
    • 王婉育
    • 蕭伶伃
    • 傅雅雯
    • 蕭伶伃
    • 楊姍樺
    • 田原
    • 郭瀅瀅
    • 林彧
    • 許丁江
  • 詩人專區
  • 商品
  • 關於
圖片

Can the 'Silicon Shield' Save Taiwan? ‘Chip War’Author Chris Miller:
​“Far From Guaranteed”


By | Kolas Yotaka

When I asked Chris Miller what he thought of claims that Taiwan, because of TSMC, holds a "Silicon Shield" that prevents the Chinese Communist Party from daring to attack, he responded without hesitation, “This has been far from guaranteed.” Chris believes that China's nationalism and aggressive ambitions are preventing them from being rational in their economic decisions. Taiwan must always prepare for the worst and strengthen its national defense.

When Trump was inaugurated for his second term, the ceremony had to be moved indoors due to an unusual cold snap. Following the inauguration, when Trump officially took office, an unprecedented global political upheaval was triggered. Domestically, Trump issued an executive order to create a temporary organization called the "Department of Government Efficiency” (DOGE), appointing tech magnate Elon Musk as its head, with the goal of slashing $1 trillion (approximately 34 trillion New Taiwan Dollars, roughly Taiwan's entire national government budget for a decade) from government agency budgets, and firing over 2 million U.S. federal employees. Among those targeted for termination were those deemed by the Trump administration to be disloyal, pro-Democrat, or hired on the basis of being minorities, including African-Americans, Latinos, and transgender individuals. Virtually the entire staff of USAID, key personnel at the Department of Defense and the State Department, including sensitive agencies like the Atomic Energy Commission and the CIA, were also terminated. Many of the targeted organizations were established after World War II during the Cold War to counter the communist Soviet Union. Now, as the U.S. seemingly weakens itself, China and Russia likely look on with hidden delight.

Under Trump's leadership, the U.S. began distancing itself from its allies, loudly proclaiming "America First" and "Put American taxpayers first," aiming to gain domestic support by legitimizing these moves. However, the call for "America First" has immediate consequences for Ukraine and Taiwan. In Ukraine’s case, the Trump administration leans on "America First” to argue that the problem should be left to Europe to handle, thereby accelerating the U.S. withdrawal of military and financial support for Ukraine in its conflict with Russia. As for Taiwan, "America First" is leading the Trump administration to speed up the relocation of the semiconductor industry to the U.S., demanding that advanced chips needed by the U.S. be produced domestically rather than in Taiwan. If not, a 100% tariff will be imposed on all imported chips to prevent Taiwan from "stealing our chips.”

In 2001, an experienced Australian journalist, Craig Addison, published a book titled "Silicon Shield: Taiwan’s Protection Against Chinese Attack." The book discusses how Taiwan's crucial semiconductor industry significantly impacts global economic development, suggesting that China would not dare attack Taiwan. Even if China had the motivation to invade, other countries would be willing to protect Taiwan, the chip producer, to safeguard their own economies. This is roughly the argument presented. Upon its release, whether seen as a valid assertion or an excess of praise, the book was largely harmless. It made people happy, boosted the collective self-esteem and confidence of Taiwanese citizens, and there was no real downside. However, this theory is merely one perspective on cross-Strait relations.

Does the "Silicon Shield" really exist? In reality, China's motivation for invading Taiwan has not diminished significantly, and other countries remain ambiguous about whether they would deploy troops to defend Taiwan. For instance, at the end of February 2025, the Chinese Communist Party carried out unannounced "live-fire training exercises" approximately 40 nautical miles off the coasts of Kaohsiung and Pingtung. They claimed it was an internal matter, stating, "This is not a diplomatic issue," and therefore did not notify Taiwan, breaking a long-standing practice of mutual notification before military exercises. Meanwhile, Trump has been asked multiple times by the media if his administration would defend Taiwan, but he has never given a straightforward answer, often choosing to focus instead on his "good relationship" with Xi Jinping. Then, in March 2025, Trump announced with great satisfaction that TSMC would invest another $100 billion in building facilities in the United States, effortlessly taking the "Silicon Shield" away from Taiwan. Trump stated that all these arrangements were made for America's "national security.”

But what about Taiwan's “national security”?


China is not being rational. The 'Silicon Shield' concept is naive.

In the final section of Chris Miller's breakout book "Chip War," titled "Taiwan's Dilemma," he mentions that when the chairman of TSMC at the time, Mark Liu, was asked whether he was concerned that the security situation in the Taiwan Strait would affect TSMC's economic outlook, he responded, "Regarding China's invasion of Taiwan, I can tell you that everyone hopes for peace in the Taiwan Strait." Chris Miller candidly stated in the book that "his (Liu's) ability to assess the risk of war remains to be seen." The book also references an article by Taiwan’s President Tsai Ing-wen in Foreign Affairs, where she argues that Taiwan's chip industry serves as a "Silicon Shield," allowing Taiwan to avoid attacks. However, Chris Miller's comments on this assertion are not far from his remarks about former TSMC chairman Mark Liu, stating that President Tsai's viewpoint is "a very optimistic perspective." He argues that the current crisis in the Taiwan Strait is more dangerous than it was in the 1950s, to the point of posing a risk of nuclear conflict, and hence, a mere "Silicon Shield" cannot effectively deter the threat.

When I asked Chris Miller if some people claim that because Taiwan has TSMC, it holds a "Silicon Shield" that deters the Chinese Communist Party from attacking Taiwan, he responded without hesitation, “This has been far from guaranteed.”

“The “silicon shield” hypothesis assumes that China – even in a crisis – will act economically rationally. Yet this has been far from guaranteed. If China were solely pursuing economic growth, it wouldn’t have foreign policy crises with many of its neighbors. You can’t explain the pressure China places on the Philippines in the South China Sea by looking at GDP growth or tech stocks—it’s about nationalism and aggression. So I don’t think we should assume economic rationality. We must prepare for worse scenarios.”

Optimists believe that no one wants to destroy Taiwan's complete semiconductor supply chain, but they fail to recognize that while everyone wishes to preserve it, many also wish to own it. Both the United States and China seek to expand their control over the semiconductor industry's supply chain, and we are currently seeing this scenario unfold. Is the concentration of chip production in any specific country an unalloyed good? Chris, who has visited Hsinchu Science Park, is in awe of Taiwan's complete semiconductor ecosystem. He has visited countries around the world that produce semiconductors and is amazed by the extraordinary scale of Taiwan's semiconductor supply chain. He believes that Taiwan's semiconductor industry is unique in the world, and throughout his research career, he has never seen such a remarkable and comprehensive semiconductor ecosystem. No other country’s semiconductor industry can match the scale of Taiwan’s. But for Chris this also raises concerns.

“Yes, I remain concerned about overconcentration.”

In Chris's view, whether for economic, business, or national security reasons, putting all your resources into a single entity is always a risky move.

“Taiwan’s and TSMC’s success has produced extraordinary concentration in the market. I don’t blame TSMC for this – I admire their success. But when it comes, for example, to producing high end GPUs, almost all are manufactured in Taiwan. Saudi Arabia produces less than 15% of the world’s oil. I think it is reasonable to simultaneously admire TSMC’s success and wish for a more geographically diversified supply chain.”

Chris politely points out that, while Saudi Arabia is referred to as a major oil-producing nation, it only accounts for 15% of the world's oil supply. Currently, 90% of high-end semiconductors are concentrated in Taiwan, including smartphone processors, AI graphics processing units (GPUs), and personal computer chips, all of which are manufactured there. Unfortunately, Taiwan continues to face the risk of conflict, which is understandably a concern for the outside world. It's important to remember that Chris is an expert in military history.

“I think investors have to be realistic and need diversified supply chains. If Google ran all of Gmail out of a single datacenter, users would ask legitimate questions about whether it is resilient to disasters. The chip industry requires similar geographic diversification.”

In 2023, when Chris came to Taiwan to share the stage with Morris Chang during the promotion of the Traditional Chinese version of his book "Chip War," he heard Morris Chang, who has witnessed multiple changes in government leadership across the U.S., China, and Taiwan, declare that "globalization is dead." For decades, the semiconductor industry has indeed witnessed significant global transformations, with dramatic geopolitical shifts and governments intervening in the distribution and profits of the semiconductor supply chain. This led Morris Chang to lament that "globalization is dead" and "free trade is dead," as government intervention, regulations, and tariff policies have constrained the semiconductor industry’s freedom to grow. Morris Chang's words still resonate, yet Chris continues to disagree with this perspective. He believes that globalization is not dead; it is merely evolving.

“If globalization were dead we wouldn’t see some countries experience historic foreign investment flows. We wouldn’t see Taiwan’s trade with the U.S. reach record highs. We wouldn’t see record foreign investment inflows to Malaysia. What’s changed is that the China-West economic relationship is now longer deepening and, in some spheres, it is starting to fragment. But I think it is wrong to equate ‘globalization’ with ‘China.’”


The United States cannot yet put up its own 'Silicon Shield’.

In early March 2025, U.S. President Trump proudly announced that TSMC would invest further in the United States. Unlike the Biden administration, which offered subsidies to encourage TSMC to come to America, Trump believes that his tariff threats have worked, resulting in TSMC investing $100 billion in the U.S. on its own. This move not only creates tens of thousands of job opportunities in the U.S. but also helps diversify TSMC's market share, with estimates suggesting that U.S. based production could become more than 40% of the global chip market. However, while countries other than Taiwan welcome the diversification and decentralization of the chip supply chain, can the U.S. really handle this market?

Morris Chang openly admitted that TSMC’s pas experience in establishing factories in the U.S. has not been smooth. Their first attempt in 1995 faced "cost issues," "talent issues," and "cultural issues." Jumping ahead to the 21st century, when TSMC went to Arizona and again attempted to establish a plant, the same problems persisted. Chris pointed out that although the U.S. is eager for TSMC to move operations there from Taiwan, the lack of diverse talent will continue to be a reason why the U.S. struggles to catch up with Taiwan.

“I think labor costs are certainly part of the question, but they’re declining in importance over time. Taiwan’s competitiveness, for example, isn’t primarily about labor costs—it is about specialization and economies of scale. I think the Trump Administration is going to continue efforts to diversify supply chains, but the U.S. will face tough tradeoffs. Diversification is very rarely cost free, and we need to be strategic about which supply chains to diversify and which supply chains are already sufficiently resilient.”

According to Chris, the ability to achieve "mass production" in the chip industry is essential for cutting labor costs, but the United States in the Trump 2.0 era still struggles to catch up to the production scale of other countries.

“Economies of scale still matter. I think the U.S in a better position today compared to five years ago, but countries like Taiwan and Korea have much more scale—and thus lower cost structures.”

It has long been the U.S.’ goal to limit China's influence. Not only did Vice President JD Vance emphasize at the "Paris AI Action Summit" in February that the Trump administration would ensure the strongest AI systems are manufactured in the U.S., using chips that are designed and produced in America, but it also seems that Trump wants to create a version of the "Silicon Shield" for the United States. 

In early March, when Trump announced that TSMC would increase its investment in U.S. production, he bluntly stated that Taiwan indeed monopolizes the market, and that the U.S. must break this monopoly in order to protect American interests:

“As you know, Taiwan pretty much has a monopoly on that market.  And I think “pretty much” is not a term that’s even appropriate.  They do have a monopoly.  And this is a tremendous move by the most powerful company in the world. 
     
It’s a matter of economic security.  It’s also a matter of national security for us.  And, at the same time, Mr. Wei (CEO of TSMC) will be able to diversify and have his tremendous presence in another place and a very safe place (the U.S.).”


After Trump returned to office, the appointments of his economic officials in the Cabinet seemed almost entirely aimed at China. Whether it's Landon Heid, the Assistant Secretary for export controls at the U.S. Department of Commerce, or Jeffrey Kessler, the Director of the Bureau of Industry and Security, both are "China hawks" viewed as likely to take strong actions against China. Recently, the Trump administration announced comprehensive export controls on AI chips to China, intending to fully ban American AI chips from entering the Chinese market. In particular, during a Senate confirmation hearing on February 27, Kessler expressed concerns about TSMC's chip supplies flowing to Huawei, emphasizing that the U.S. must enforce strict regulations. Yet, does the combination of seizing assets and imposing blockades effectively secure continued American semiconductor technology superiority over China?

“I think the right way to phrase it is: the combined partnership of the U.S, Taiwan, Japan, the Netherlands and South Korea will stay ahead of China. I think they will—but the ability to capitalize on expertise from all these partners is critical.”

It is clear that Chris believes the "Silicon Shield" cannot be upheld by a single manufacturer or a single country; it requires a collective effort from a group of people. Going it alone makes it difficult to achieve significant progress.

Watch out for the potential risks posed by 'Made in China 2025.’

In 2015, Chinese Premier Li Keqiang proposed a 30-year transformation plan for the manufacturing sector aimed at positioning China as a "manufacturing powerhouse" to break the stereotype of China producing "low-cost" products through "low labor" and "low technology." This led to the introduction of the "Made in China 2025" initiative, which encompasses a three-step plan: the first phase from 2015 to 2025, followed by the second and third phases from 2025 to 2035 and 2035 to 2045 (with 2045 marking the 100th anniversary of the founding of the People's Republic of China). The goal is to make China a dominant global technology superpower. The authoritarian Communist Party of China has a 30-year plan to strengthen the nation and its military, and is committing substantial budgets to both hardware and software development, claiming it aims to compete with democratic blocs. This is quite concerning.

In early February 2025, the Chinese AI company DeepSeek launched a chatbot, causing a stir in global stock markets. It was originally thought that the U.S. led the world in AI technology, with OpenAI taking the lead; however, it was unexpectedly surpassed by a relatively unknown Chinese company, DeepSeek. The language learning model from DeepSeek was released at a lower cost, with the company emphasizing that it is entirely made in China. The GPUs used are sourced from China, and during operations, they consume less power and rely on fewer high-end chips, claiming to achieve a level almost comparable to that of OpenAI, thus breaking the pathway for "computational power" and offering a more cost-effective option for the global AI market. Once this news broke, global stock markets plummeted, and Trump stated that the emergence of DeepSeek was indeed a wake-up call for the United States.

But Trump remains confident, believing that DeepSeek does not pose a threat to national security. He states, "This is an emerging technology. In fact, I think we will benefit from it." His instinctive business mindset perceives cheaper AI as a positive development. "I think this is a very good development," Trump remarked. He quickly set aside the national security concerns. However, Chris cautioned that the emergence of DeepSeek should prompt both Taiwan and the U.S. to be vigilant, especially regarding the national policy of "Made in China 2025.”

“Yes, we should worry. China’s industrial policies aren’t perfect, but China has caught up technologically in almost every manufacturing sector. Chips are one of the rare exceptions. Semicondcutor manufacturing is, of course, extraordinarily difficult to replicate. However, the combination of China’s government spending and the size of the Chinese market mean that we must take seriously their capabilities.”  

With the emergence of DeepSeek in China, which claims that its chips and GPUs are domestically produced, it seems that "Made in China 2025" is starting to yield results. U.S. President Trump has also proposed large technology initiatives, such as the StarGate project, demonstrating his determination to convince the world that the U.S. will maintain its leading position in the field of artificial intelligence. Both the U.S. and China need to showcase their domestic production and self-sufficiency capabilities. The competition between the U.S. and China raises the question of whether it will lead to a split in the existing chip supply chain into two systems: one for China and one for non-China. Chris believes that, in the short term, despite China's unilateral ambitions to grow stronger, it will still be unable to catch up.

“I disagree with DeepSeek. The analysts I most trust believe the company had access to up to 50,000 GPUs from Nvidia. China continues to import large volumes of Nvidia H20 chips. DeepSeek’s CEO has repeatedly publicly said his number one problem is access to high end chips. China’s still not self sufficient at the cutting edge. The main reason is the US and Dutch restrictions on selling chipmaking equipment like lithography tools. Without those restrictions, SMIC would have made much more progress in catching up.”


The Future of the 'Silicon Shield': The Defense Industry

Drones controlled by artificial intelligence, lethal autonomous weapon systems, and new types of nuclear arms have become standard features of the "new Cold War." Chris, a historian specializing in military history, has noted that most significant technological advancements throughout history have been driven by warfare. He has observed that in the conflicts in Ukraine and between Israel and Palestine over the past three years, the development of drones has become the latest evolution in national defense.

These recent wars have combined some very old trends, like the World War I-style trench warfare in the Russia-Ukraine conflict, with some new 21st century dynamics. Many analysts have noted the role of UAVs in both reconnaissance and strike operations in both conflicts. I’ve also been impressed by the central role of tech-enabled intelligence in enabling Ukrainian and Israeli operations. 

According to a New York Times article from March titled "A Thousand Snipers in the Sky—Ukraine's New War," which provides an in-depth report on the situation in Ukraine, multiple Ukrainian commanders were interviewed, and most indicated that a majority of kill missions are now carried out by drones. In three years, 70% of the casualties in the Russia-Ukraine conflict have been caused by drones. Remote-controlled killing has turned drones into nightmares for both sides, and this dramatic shift is changing the nature of future warfare. If the U.S. withdraws military aid to Ukraine, it will force both sides to adapt their warfare strategies, leading to an increased, rather than decreased, demand for relatively cost-effective drones. These harsh precedents have already altered the nature of warfare and should raise significant alert for Taiwan.

This being the case, our country must pay greater attention to the semiconductor industry, as national defense resilience depends on a complete semiconductor supply chain.

“Semiconductors are central to these new technologies in multiple ways. First are the chips that go in systems. A high-end drone might have one or two sophisticated chips and a handful of less complicated sensors and other semiconductors. Next to the chips in a specific system are the chips in a network that enable the communication and command and control which make it possible to use that system. These communication networks are full of high-end chips. Finally, there are chips in the datacenters that play an increasingly large role in developing and deploying these systems. We need to think about resilience in defense semiconductor supply chains as not only being about devices, but about the entire network that enables them.”

Chris reminds us that the drone industry is a field that Taiwan should focus on and invest in more.

“Understanding the evolution of drone supply chains provides both a lens for how electronic supply chains are changing in general—and how defense supply chains are evolving in specific. Senior military leaders in all the key countries in the Indo-Pacific realize that access to large quantities of drones will be an important determinant of any future conflict in the region. They will therefore be important for deterring conflict, too. China has historically dominated these supply chains, and it is critical that we find ways to diversify production in more reliable partners.”

As a historian, Chris believes that analyzing the past and present of the semiconductor industry is crucial for predicting its future, especially in areas that are inherently difficult to anticipate, such as politics or technology. Reflecting on history can help us understand future possibilities. Although we may not clearly see the road ahead at present, by analyzing history, we can maintain a sense of awareness that allows us to continue moving forward through the fog.

​
全文刊於《人間魚詩生活誌》第二十期
成為台灣人間魚詩社文創協會 贊助會員

台灣人間魚詩社文創協會為依法設立、非以營利為目的之社會團體。以推廣現代詩、文學及其它藝術創作,推動社會創意產業文化為宗旨。

​贊助用途:
1、協會運作
2、人間魚詩生活誌及其它書籍出版
3、詩電影拍攝
​
贊助帳號:
第一銀行(007)松貿分行 168-10-002842 社團法人台灣人間魚詩社文創協會

姐妹詩社:秋水詩社| 竊竊詩語|新詩路 |掌門詩學社
補助單位:
圖片
Copyright © Peoplefish Poetry & Life, all rights reserved.​