Advanced packaging is the frontier of semiconductor technology, and TSMC is in the lead – The China Project – The China Project

It’s becoming difficult to make microchips smaller, so the encasing around them is where advanced tech is going. TSMC’s advanced packaging technology has emerged as a game-changer, with repercussions extending far beyond circuitry for U.S.-Taiwan relations.
Saved.
Something went wrong. Error code 1.
This feature is only available for paying subscribers. You can continue reading without saving your preference, or you can subscribe now.
Subscribe Now
Already a subscriber? Log in here.
This feature is for paid subscribers only.
Copy this link
Copied!
Something went wrong.
Taiwan Semiconductor Manufacturing Company (TSMC) manufactures the world’s most advanced microchips. The company is not resting on its laurels: In a September 6 talk at the Semicon Taiwan industry conference, TSMC Chairman Mark Liu (劉德音 Liú Déyīn) confirmed the company’s aggressive expansion of its Chip-on-Wafer-on-Substrate (or “CoWoS”) manufacturing capacity, a move set to redefine the strategic landscape of the industry.
“Advanced packaging” is something you are going to be hearing about a lot in the near future. It’s the architecture that links, encases, and safeguards minuscule silicon chips, thus creating the ideal stable operation systems needed for smarter artificial intelligence (AI) devices and other high-end uses. (It’s worth noting that packaging was historically considered a less prestigious, blue-collar task, which led to its migration from the U.S. to Southeast Asia, China, and Taiwan.)
Moore’s Law — formulated in 1965 by Gordon Moore, a co-founder of Intel, states that the number of transistors on a microchip doubles about every two years. But the law isn’t working anymore, and it’s getting more and more difficult to shrink microchips. That’s where advanced packaging comes in: The encasing around the microchips is becoming key to high performance. As the features on semiconductors have decreased to impossibly small dimensions, chip designers have turned to these advanced packaging techniques to improve processor performance without having to further shrink chips’ transistors.
CoWoS, a specific type of advanced packaging, was developed by TSMC to manufacture computing chips like Nvidia’s top-of-the-line A100 and H100, powering a host of AI applications like ChatGPT. Notably, as of today, CoWoS manufacturing does not exist in the U.S.
CoWoS technology sidesteps Moore’s law by allowing multiple chiplets to be packed into one package, dramatically increasing performance and efficiency and reducing energy consumption, all while solving the technical issues that plague current multi-chip arrangements.
Currently, TSMC can’t produce enough CoWoS chips to meet demand, Liu said at Semicon, but the shortage should be alleviated in one and a half years thanks to plans for expanded production.
Liu’s comments highlighted the critical role of semiconductor technology in the age of AI, where tech mastery directly correlates to geopolitical sway. Liu described the journey of semiconductor technology as walking through a tunnel for the past 50 years, focusing on the singular task of shrinking transistors. “Now we are reaching the exit of this tunnel,” he said, emphasizing that while the technology becomes harder to develop, “many more possibilities lie ahead in the future.”
TSMC’s commitment to staying ahead of the competition when it comes to cutting-edge manufacturing comes as the company makes major investments in chip foundries in the U.S. At a time when tech plays a critical role in U.S.-China competition, the success or failure of TSMC’s investments — in advanced packaging and in the U.S. — will take on major geopolitical significance.
Rapid innovation is the norm in the semiconductor industry. But TSMC’s forward-thinking strategies stand out. The company initiated its advanced packaging programs like CoWoS back in 2009, led by Shang-Yi Chiang (蔣尚義 Jiǎng Shàngyì), in anticipation of the technological barriers that could hinder the growth of AI applications, but a long time before it was obvious that packaging would become key to the future of semiconductors.
As a result, TSMC’s CoWoS technology plays a critical role in the manufacturing of AMD’s Instinct MI300 advanced chiplets, and Nvidia’s decision for its H100 manufacturing keeps them uniquely dependent on TSMC, as it is the only firm capable of achieving high-volume manufacturing of such designs at a reasonable cost. Therefore, Nvidia and AMD together account for a significant portion of TSMC’s CoWoS production capacity.
But they’re not the only ones; CoWoS is especially popular for chips that need intense computational power, like those used for AI. Other industry heavyweights like Microsoft’s Athena, Amazon’s Trainium, Google’s TPU, and Intel’s NNP-T also demand CoWoS technology.
Continued investments in chip manufacturing will not only make AI applications more feasible but also exponentially amplify their potential impact across various sectors. Consider the healthcare industry, where AI-driven diagnostics and personalized treatment plans could become more precise and affordable, or national security, where real-time data analytics could offer actionable insights during critical missions. As monolithic chip development faces increasing difficulties, CoWoS represents a pivot to advanced packaging, which industry leaders like AMD’s Lisa Su (蘇姿丰 Sū Zīfēng) have already championed.
This innovation represents more than just a reconfiguration in semiconductor architecture; it introduces a pivotal element that could stir U.S.-Taiwan relations. The evolving situation underscores the United States’ intensified efforts to attract Taiwan’s advanced semiconductor capabilities, including CoWoS, to its shores. Failure to handle these negotiations with delicacy and mutual respect could result in the U.S. missing a seminal opportunity.
As TSMC expands its investments in the U.S., the nexus of cutting-edge CoWoS technology, diplomatic finesse, and corporate interests is shaping a transformative geopolitical landscape that demands close attention.
Margaret Siu is a JD candidate at Harvard Law School, where she studies U.S.-East Asian economic statecraft. As a Marshall Scholar, she earned her M.Sc. degrees from the University of Oxford and the London School of Economics and Political Science. She is on the executive board of the Harvard International Law Journal, the board of advisors for the Oxford Silk Road Society, and is the founder of Apricity Magazine, an international arts magazine. Previously, she worked for the Center for Strategic and International Studies (CSIS) and Taiwan Semiconductor Manufacturing Company’s (TSMC) Corporate and Compliance Legal Division. Read more
Max Baucus
Former U.S. Ambassador to China
Get Premium Access
Necessary cookies are absolutely essential for the website to function properly. This category only includes cookies that ensures basic functionalities and security features of the website. These cookies do not store any personal information.
Performance cookies are key in allowing web site screens and content to load quickly on all types of devices.

Cookie Description
_gat This cookies is installed by Google Universal Analytics to throttle the request rate to limit the colllection of data on high traffic sites.
YSC This cookies is set by Youtube and is used to track the views of embedded videos.

Preference cookies are used to store user preferences to provide them with content that is customized accordingly. These cookies also allow for the viewing of embedded content, such as videos.

Cookie Description
bcookie This cookie is set by linkedIn. The purpose of the cookie is to enable LinkedIn functionalities on the page.
lang This cookie is used to store the language preferences of a user to serve up content in that stored language the next time user visit the website.
lidc This cookie is set by LinkedIn and used for routing.
PugT This cookie is set by pubmatic.com. The purpose of the cookie is to check when the cookies were last updated on the browser in order to limit the number of calls to the server-side cookie store.

Analytics cookies help us understand how our visitors interact with the website. It helps us understand the number of visitors, where the visitors are coming from, and the pages they navigate. The cookies collect this data and report it anonymously.

Cookie Description
__gads This cookie is set by Google and stored under the name dounleclick.com. This cookie is used to track how many times users see a particular advert which helps in measuring the success of the campaign and calculate the revenue generated by the campaign. These cookies can only be read from the domain that it is set on so it will not track any data while browsing through another sites.
_ga This cookie is installed by Google Analytics. The cookie is used to calculate visitor, session, campaign data and keep track of site usage for the site’s analytics report. The cookies store information anonymously and assigns a randomly generated number to identify unique visitors.
_gid This cookie is installed by Google Analytics. The cookie is used to store information of how visitors use a website and helps in creating an analytics report of how the wbsite is doing. The data collected including the number visitors, the source where they have come from, and the pages viisted in an anonymous form.
_omappvp The cookie is set to identify new vs returning users. The cookie is used in conjunction with _omappvs cookie to determine whether a user is new or returning.
_omappvs The cookie is used to in conjunction with the _omappvp cookies. If the cookies are set, the user is a returning user. If neither of the cookies are set, the user is a new user.
GPS This cookie is set by Youtube and registers a unique ID for tracking users based on their geographical location

Advertisement cookies help us provide our visitors with the most relevant ads and marketing campaigns.

Cookie Description
__qca This cookie is associated with Quantcast and is used for collecting anonymized data to analyze log data from different websites to create reports that enables the website owners and advertisers provide ads for the appropriate audience segments.
_fbp This cookie is set by Facebook to deliver advertisement when they are on Facebook or a digital platform powered by Facebook advertising after visiting this website.
everest_g_v2 The cookie is set under eversttech.net domain. The purpose of the cookie is to map clicks to other events on the client’s website.
fr The cookie is set by Facebook to show relevant advertisments to the users and measure and improve the advertisements. The cookie also tracks the behavior of the user across the web on sites that have Facebook pixel or Facebook social plugin.
IDE Used by Google DoubleClick and stores information about how the user uses the website and any other advertisement before visiting the website. This is used to present users with ads that are relevant to them according to the user profile.
mc This cookie is associated with Quantserve to track anonymously how a user interact with the website.
personalization_id This cookie is set by twitter.com. It is used integrate the sharing features of this social media. It also stores information about how the user uses the website for tracking and targeting.
PUBMDCID This cookie is set by pubmatic.com. The cookie stores an ID that is used to display ads on the users’ browser.
TDCPM The cookie is set by CloudFare service to store a unique ID to identify a returning users device which then is used for targeted advertising.
TDID The cookie is set by CloudFare service to store a unique ID to identify a returning users device which then is used for targeted advertising.
test_cookie This cookie is set by doubleclick.net. The purpose of the cookie is to determine if the users’ browser supports cookies.
uid This cookie is used to measure the number and behavior of the visitors to the website anonymously. The data includes the number of visits, average duration of the visit on the website, pages visited, etc. for the purpose of better understanding user preferences for targeted advertisments.
uuid To optimize ad relevance by collecting visitor data from multiple websites such as what pages have been loaded.
uuidc This cookie is used to stores information about how the user uses the website such as what pages have been loaded and any other advertisement before visiting the website. This data is used to provide users with relevant ads.
VISITOR_INFO1_LIVE This cookie is set by Youtube. Used to track the information of the embedded YouTube videos on a website.

source

Facebook Comments Box

Trả lời

Email của bạn sẽ không được hiển thị công khai. Các trường bắt buộc được đánh dấu *