Taiwan Semiconductor Manufacturing Co (TSMC) on May 15 announced that it would build an advanced semiconductor fab that would mainly deploy 4-nanometer process technology in Phoenix, Arizona.
Most Taiwanese are not aware that Phoenix has become an important chip manufacturing center in the US. Although South San Francisco in the Bay Area is known as Silicon Valley, the tech industry there today has little to do with silicon, or chip, production, except for a few old Intel fabs. Instead, it is more focused on upstream semiconductor industries, such as the Internet, biotech, green energy and artificial intelligence.
Apple, for example, designs its own mobile phones; Google designed a tensor processing unit for artificial intelligence (AI); Nvidia’s graphics processor has been widely adopted by AI, blockchain and bitcoin companies around the world in the past few years. These upstream companies own highly innovative technologies and they basically outsource all their manufacturing to companies like TSMC or United Microelectronics Corp (UMC).
TSMC, which has more than nine advanced IC fabs in Taiwan, is the world’s largest contract chipmaker with the most advanced manufacturing quality. However, in other US high-tech sectors, such as defense, space technology, medical equipment, 5G and self-driving vehicles, plants capable of supplying small batches of high-tech chips are still in demand.
TSMC’s decision to build a fab in the US is clearly based on the huge potential it sees for providing its professional services locally. Regardless of how much political pressure from the US and Taiwanese governments was behind TSMC’s decision, it is a milestone in the internationalization of Taiwan’s technology industry. Past governments have encouraged Taiwanese companies to make global deployments, but without the leading companies taking the lead, who would dare to go it alone?
TSMC aims to start detailed plans for its Phoenix facility next year and to complete construction in 2022-2023, with 4-nanometer chip production targeted to begin in early 2024.
TSMC has a fab in the Southern Taiwan Science Park that has started mass production of 5-nanometer chips. Its plan to set up a more advanced fab deploying 2 to 3-nanometer process technology in Kaohsiung’s Luzhu Science Park has already been approved by the government.
With the emergence of a large number of new technologies, demand for high-end applications is expected to focus on 4 to 7-nanometer chip technology within the next five to seven years. TSMC’s main competitors, such as Samsung and Intel, are doing their utmost to catch up, making 4 to 7-nanometer technology their focus for the next 10 years.
TSMC’s bold plan to build its second 4 to 5-nanometer fab on US soil to be closer to potentially huge new customers is a wise and timely move. Building an advanced fab in the US is based on several important strategic considerations.
First, separating production sites and capabilities is an attempt to reduce the risk of becoming involved in the intensifying US-China technology dispute.
Second, continuing to boost investment in high-end process technologies would help TSMC maintain its leading status in the industry in the long term.
Third, top US university laboratories are leading the world in the research and development of new materials, precision optics, communication technology and digital healthcare, providing geographical proximity for the TSMC US fab to develop next generation 1-nanometer technology.
Sustainable development of an enterprise requires continuous efforts by generations of talent. TSMC founder and former chief executive Morris Chang (張忠謀) achieved great things, and the two co-CEOs who took over are also doing a good job.
It has been said that TSMC is a black hole that sucks in Taiwan’s high-end talent as one of every two graduates from the master’s or doctoral programs of the nation’s four top technology universities end up working for TSMC.
However, this top talent, who are now over the age of 40 and have been trained by TSMC for more than 15 years, are not like their predecessors from the two previous generations, who studied in the US and had experience working in large global companies. The question now is how to better equip them to lead this world-leading enterprise. Transferring them to Phoenix would clearly be a feasible shortcut to giving them the training and experience they need.
From another perspective, building TSMC’s second 4-nanometer plant in Phoenix would release 3,000 to 4,000 graduates from Taiwan’s elite universities to the job market. This would boost the workforce needed by other enterprises with high potential. This group represents Taiwan’s top science and engineering talent, and they would be a great human resources asset to Taiwan and a positive force to transform and upgrade the nation’s industry and enhance its international competitiveness.
Tu Tze-chen is former head of the Industrial Technology Research Institute’s Knowledge-based Economy and Competitiveness Center.
Translated by Lin Lee-kai
Taiwan has lost Trump. Or so a former State Department official and lobbyist would have us believe. Writing for online outlet Domino Theory in an article titled “How Taiwan lost Trump,” Christian Whiton provides a litany of reasons that the William Lai (賴清德) and Donald Trump administrations have supposedly fallen out — and it’s all Lai’s fault. Although many of Whiton’s claims are misleading or ill-informed, the article is helpfully, if unintentionally, revealing of a key aspect of the MAGA worldview. Whiton complains of the ruling Democratic Progressive Party’s “inability to understand and relate to the New Right in America.” Many
US lobbyist Christian Whiton has published an update to his article, “How Taiwan Lost Trump,” discussed on the editorial page on Sunday. His new article, titled “What Taiwan Should Do” refers to the three articles published in the Taipei Times, saying that none had offered a solution to the problems he identified. That is fair. The articles pushed back on points Whiton made that were felt partisan, misdirected or uninformed; in this response, he offers solutions of his own. While many are on point and he would find no disagreement here, the nuances of the political and historical complexities in
Taiwan faces an image challenge even among its allies, as it must constantly counter falsehoods and misrepresentations spread by its more powerful neighbor, the People’s Republic of China (PRC). While Taiwan refrains from disparaging its troublesome neighbor to other countries, the PRC is working not only to forge a narrative about itself, its intentions and value to the international community, but is also spreading lies about Taiwan. Governments, parliamentary groups and civil societies worldwide are caught in this narrative tug-of-war, each responding in their own way. National governments have the power to push back against what they know to be
Taiwan is to hold a referendum on Saturday next week to decide whether the Ma-anshan Nuclear Power Plant, which was shut down in May after 40 years of service, should restart operations for as long as another 20 years. The referendum was proposed by the opposition Taiwan People’s Party (TPP) and passed in the legislature with support from the opposition Chinese Nationalist Party (KMT). Its question reads: “Do you agree that the Ma-anshan Nuclear Power Plant should continue operations upon approval by the competent authority and confirmation that there are no safety concerns?” Supporters of the proposal argue that nuclear power
RSS