HomePage  |  This day in history  |  Sitemap
Breaking-News >> WorldNews

"Made in America" chips come out, produced at TSMC factories in the US, but packaging is still overseas

[Global Times Special Correspondent in the United States Feng Yaren] According to Reuters and other foreign media reports, artificial intelligence chip giant Nvidia announced last Friday that it would officially release the first Blackwell chip wafer manufactured in the United States. The wafer is produced by TSMC's semiconductor plant in Phoenix, Arizona, USA. Reported that this move marks that the United States is accelerating the reshaping of the high-end chip supply chain against the backdrop of rapid growth in demand for artificial intelligence chips.

“For the first time ever, one of the most important chips is produced in Taiwan’s power plant in the United States,” said Yuan Yonglong, founder and chief executive officer of NVIDIA, when visiting Taiwan’s advanced production plant in Arizona on Friday, which also meets the current U.S. government’s vision of “bringing manufacturing back to the United States.”

According to reports, the Arizona plant plans to mass production of 2nm, 3nm, 4nm and other advanced process chips, these chips are widely used in artificial intelligence, telecommunications and other fields. NVIDIA said that these high-end products will be the core driving force supporting the development of the U.S. AI industry. But it is worth mentioning that the U.S. technology media "Tom's Hardware" website noted that although Blackwell chips have been made in the U.S., but to finally complete the NVIDIA B300 products, they still need to be packaged in the Taiwan plant.

“The United States is further strengthening its domestic semiconductor supply chain.” Japanese and Korean media, such as The Asian Review, The Asian Business Daily, also pay attention to the first NVIDIA-Taiwan chip cycle, and report that the United States has previously attracted chip factories such as Taiwan Electric to the United States by providing large subsidies. During the reign of former U.S. President Biden, Taiwan Electric, with an investment of $65 billion and a subsidy of $6.6 billion, began to build a factory in Arizona. In addition to Taiwan Electric, since the implementation of the 2022 “Chip and Science Law”, the United States has issued billions of dollars in subsidies and subsidies to chip manufacturers, but many companies still complain: their customers are unwilling to pay higher costs for “Made in America” and

According to the Wall Street Journal last month, the U.S. government is working on a new regulation that requires chip manufacturers to produce in the U.S. the number of semiconductors that must be “equal” to the number of chips their customers import from overseas (i.e. 1:1 ratio), and companies that fail to this ratio will face about 100% tariffs. This policy is intended to force chip companies to focus on production back to the U.S. and reduce dependence on the U.S. overseas supply chain.

U.S. Commerce Secretary Howard Lutnick has recently proposed the idea to several chip company executives, believing that it may be a necessary measure to maintain "U.S. economic security." However, industry insiders warned that the implementation of the policy faces challenges. First, the complexity of the supply chain makes tariff calculations extremely complicated; second, the U.S. domestic manufacturing capabilities are shortcomings and cannot cover the production of all high-end chips. Currently, the U.S. government is conducting a trade investigation into the impact of chip imports on national security and is expected to formally announce a new round of chip tariff measures after the investigation is over.

Editor in charge: Guo Bowen



News raw data sources → https://news.sina.com.cn/w/2025-10-21/doc-infuries2980170.shtml

17WorldNews[2025.10.21-09:56] 访问:47
[关闭窗口]  
「Links」 ...
Loading...
Search on site
This day in history
August 2023
Sun
Mon
Tue
Wed
Thu
Fri
Sat
Copyright © 17ljfl.com · World News
The information collected on this site is all from public data information on the Internet, and the authenticity of the query results is for reference only!