CNN
—
The U.S. government is doing everything in its power to boost domestic semiconductor manufacturing, pouring billions of dollars into the struggling sector and using all available policy instruments to push Asia and to gain an advantage over the competition.
When the pandemic hit in 2020, companies initially cut orders for these micro building blocks needed for smartphones, computers, cars and many other products. , soared demand for information and communication technologies and the chips that power them. There was a shortage of chips, and car factories were forced to stop production because they could not get chips. This contributed to the skyrocketing prices of new and used cars, a major factor in the painful inflation felt by Americans.
In a statement earlier this year, Commerce Secretary Gina Raimond called the semiconductor shortage a “national security” issue. This is because the US manufacturing industry has made it clear that it depends on imports of semiconductors from abroad. Chips are also used in critical military applications and are required for cybersecurity tools.
The Biden administration and lawmakers rallied in response and passed CHIPS and the Science Act in August. The law includes his $52 billion to boost semiconductor manufacturing in the United States. Of this, $39 billion was allocated to manufacturing incentives, $13.2 billion to R&D and workforce training, and $500 million to international information and communication technology security and semiconductor supply chain activities.
Against this background, several prominent companies are making important announcements. Investment in US manufacturing. Industry giant Taiwan Semiconductor Manufacturing Company (TSMC) is investing at least $12 billion to build a semiconductor manufacturing plant in Arizona, with production he expects to begin in 2024. Ohio’s multi-billion semiconductor manufacturing plant broke ground for a new chip factory last month. And this month, Micron announced that it will invest up to $100 billion over the next 20 years to build a massive semiconductor factory in upstate New York.
Earlier this month, a flurry of tweets from President Joe Biden pledged: “America will lead the way in microchip manufacturing.”
But the United States has a lot of work to do. Currently, US-based fabs account for only 12% of the world’s modern semiconductor manufacturing capacity, according to data from the Semiconductor Industry Association trade group. About 75% of the world’s modern chip manufacturing is now concentrated in East Asia, with the majority of it concentrated in geopolitically vulnerable Taiwan. And even with these new efforts, the US currently doesn’t have the same talent and supply chain pipeline that some Asian markets do to support robust domestic industries.
Complicating matters is that the surge in public and private investment has come at a questionable time as concerns about a global chip supply shortage have abated. Pandemic-related supply chain blockages have eased somewhat, and a weaker economic outlook is hampering demand.
On last week’s earnings call, TSMC CEO CC Wei warned that by 2023 he expects “the semiconductor industry is likely to decline.” ”
Currently, accelerating semiconductor manufacturing in the United States risks leading to overcapacity and oversupply. And with demand weakening, it’s not immediately clear whether government subsidies will be enough to overcome other obstacles the country faces in developing a competitive semiconductor manufacturing hub.
In order to understand the latest US efforts, it is important to clarify where the country stands, not only in relation to the chip industry as a whole, but also to specific valuable geographies.
“The chances of the U.S. increasing its share of global production are very low, even if the U.S. brings more fab capacity online. , and we are building it even faster,” said Scott Kennedy, senior adviser to the Center for Strategic and International Studies.
“But I don’t necessarily think it’s really a big deal,” he added. He points out that measuring manufacturing based on pure volume yields a more realistic and important measure of chip-manufacturing success: low-end chips and cutting-edge, high-end chips can be lumped together. did. “The United States needs to ramp up production of certain types of chips that are directly related to U.S. national security,” he said.
The Biden administration last Friday imposed sweeping new export controls aimed at limiting China’s access to advanced semiconductors made with U.S. equipment, a move targeting the manufacture of advanced weapons systems. showed
On the other hand, “only about 10% to 14% of chips sold” [globally] It comes from a manufacturing facility in the United States, according to a professor at Columbia Business School.” Dan Wang, the US has other strengths. “In terms of design expertise, a lot of it is still in the US.”
Still, the drawbacks are real. “When it comes to foundries, the manufacturing side of semiconductors, the US hasn’t been a major player for many years,” Wang said. It used to be mostly the case, but in the 1980s and he said, manufacturing began to migrate to Asia in the 1990s. “One of the big reasons for this, he said, is that labor costs are low and it’s much cheaper to produce integrated circuits and chips on a very large scale in these parts of the world,” Wang added. . TSMC founder Morris Chang says he costs 50% more to make chips in the US than in Taiwan.
Now, just by having facilities already set up to produce or scale up chip manufacturing, Asia has a huge advantage. Wang said he believes that’s why the United States is “spending a lot of money on companies to set up factories in the United States.” Not just to be more self-sufficient to meet demand, but also “because you have to get these things up and running very quickly in order to be in the race.”
Building a new chip factory is itself an expensive and time-consuming task. Gartner analyst Bob Johnson said, “The modern fab he’s on is about half a million square feet,” requiring “huge clean rooms with massive air handling capabilities.” He added that these massive buildings would need “very strong foundations”. As he said, “You can’t give Fab a vibration. manufacturing process. ”
In addition, a single extreme ultraviolet lithography machine required to map the circuitry of a chip costs about $150 million, and Reuters said, “A state-of-the-art chip factory would require anywhere from 9 to 18 of these machines. It is reported.
In addition, in the manufacture of semiconductors, fluorinated polyimides, Etching gas, chip etching machine, etc. In places such as Taiwan and Fukuoka, Japan, supply chains have developed where the suppliers of these products are located near semiconductor factories. He also has a company or two that produce critical inputs and have long been reliable suppliers to Asian companies. This is not yet the case in places like Arizona and Ohio, where plans to build large chip manufacturing plants are already underway.
We also need a workforce that is willing to work.
The United States lacks both new graduates and seasoned workers with the technical and engineering knowledge necessary to manufacture semiconductors. Kennedy said many people who might have the right experience prefer working in trendy industries instead.
“Even if we had 10 new fabs with world-leading chips today, we wouldn’t have enough people to staff them,” said Kennedy. “It’s not capital, it’s the biggest bottleneck to expanding American fab capacity.”
Intel has a close relationship with Arizona State University and is looking to hire engineers, but it remains to be seen whether the company or other companies building fabs in the US will be able to hire well-trained engineers and technicians. I don’t know. Otherwise, even the billions of dollars put in by the private and public sectors may not be enough to revive semiconductor manufacturing.