A review the global chip landscape makes a compelling argument for full funding of the Chips For America Act

The U.S. Senate this week brought to the floor a grab bag of bipartisan legislation that would authorize well over $100 billion in spending and tax incentives to support advanced technology aimed at competing with China. 

The base legislation is the Endless Frontier Act, a $112 billion high-tech investment plan that would establish a Technology and Innovation Directorate at the National Science Foundation and a White House Office of Manufacturing and Industrial Innovation Policy. Senate Majority Leader Chuck Schumer, the lead sponsor, made the focus of the legislation clear. “We can either have a world where the Chinese Communist Party determines the rules of the road for 5G, artificial intelligence and quantum computing,” he said, “or we can make sure the United States gets there first.” 

While efforts to compete with China enjoy broad support, the massive outlays have both conservative detractors, who ideologically oppose government spending, and liberals who do not want money going to industry.

To contextualize these legislative efforts within the global technology race, I reviewed spending plans around the world on the core component of all advanced industrial technology – semiconductors – an industry the United States invented and could lose its leadership in without government support.

An amendment to the legislative package calls for a $52 billion appropriation to support R&D and manufacturing for semiconductors and the creation of a National Semiconductor Technology Center. American companies still dominate semiconductor design and the equipment and processes that make them. But the U.S. today only accounts for 12 percent of global semiconductor manufacturing, compared to 37 percent in 1990. 

Chip fabrication plants, or fabs, have been moving overseas since the 1990s. As with many other industries, chip companies moved away from holistic, in-house business models to focus only on design and sales, outsourcing manufacturing to less expensive areas overseas. Silicon fabs are now located in Israel, Ireland, and across Asia, where governments offer subsidized construction costs ranging into the tens of billions of dollars.

Today, the world is going chip crazy. The U.S., China, Taiwan, South Korea, Japan, and the European Union are all competing in a global semiconductor manufacturing race. They are driven by a pandemic-induced chip shortage, but also by a desire to secure the chip supply chain within their borders to avoid future tariffs, sanctions, and export controls.

So, let’s travel around the semiconductor world to see what the $52 billion chip manufacturing and innovation appropriation under consideration by Congress is up against.

Just last week, South Korea unveiled an ambitious plan for both government and industry to spend some $450 billion over the next decade to expand its already massive chip design and manufacturing base. The country has a 65% global market share in memory chips, but lags in the higher end logic chips used in AI and other advanced data processing. The government, which refers to chips as “strategic weapons,” also plans to train 36,000 chip experts over the next decade while also attracting foreign makers of chip making equipment. 

Taiwan is the current global chip manufacturing behemoth because of one company, Taiwan Semiconductor Manufacturing Company (TSMC). TSMC is planning to spend $28 billion on capex in 2021 alone, and another $100 billion over the next three years to expand its chip fabs. The company controls more than half of the made-to-order chip foundry market, and dominates high-end components for everything from smartphones to weapons systems. 

Under pressure from Washington, companies are moving to shift some of their manufacturing to America. Last year, TSMC announced plans to build a $12 billion foundry in Arizona. Intel also announced it will invest $20 billion in two new Arizona chip fabs to expand the company’s in-house manufacturing capabilities. Samsung is planning to add a new $17 billion fab to its existing manufacturing site in Austin, Texas. These investments are all reliant on the $52 billion investment support currently being considered by Congress.

In March, the European Union said it is considering spending up to $150 billion to double its current 10% share of the global advanced chips market. EU commissioners are also pushing the idea of building a “EUROFAB.” In Brussels last month, Intel CEO Pat Gelsinger discussed the EUROFAB, and indicated that building it would require a $9 billion EU subsidy. 

Not to be outdone, Japan’s ruling Liberal Democratic Party last week announced the formation of a legislative caucus to determine how to secure the country’s semiconductor supply and boost competitiveness. The caucus will be chaired by former economic minister Akira Amari who declared: “It is not an exaggeration to say that those who control semiconductors will control the world.” 

China is laser-focused on catching up on chips. The country has already made progress on memory and other lower end commodity chips. China is now headed toward becoming the world’s largest chip maker by volume. But the country remains at least five years behind global semiconductor leaders in advanced chips: domestic chip production today, for example, meets only 16% of China’s semiconductor demand.

Making headway on the most advanced chips is an elusive process, as the complex technology behind them has taken decades for others to develop. Nonetheless, since 2014 Beijing has earmarked at least $50 billion for new chip design houses and fabs to achieve chip self-sufficiency. U.S. export controls on Chinese imports of chips and semiconductor design and manufacturing equipment have only heightened this effort. 

Investments in chips through government venture funds is also enormous. Phase I of China’s National IC Fund deployed almost $22 billion. Phase II, which kicked off last year, has about $32 billion in registered capital. More than a dozen local governments have also deployed their own IC funds.

There are two central government megaprojects, one for chip design and one for manufacturing. As of 2020, the megaproject for chip design has deployed some $15 billion in grants. Domestic Chinese venture capital and private equity firms are also incredibly active in semiconductor investments, investing in 408 projects in 2020 and 456 projects in 2019. Despite these efforts, China is not anywhere close to the government’s intended target of local chips constituting 70 percent of the China market by 2025. Industry analysts predict China may fall short of even 20 percent. 

Like so many Chinese government all-hands-on-deck investment plans, no small amount of money has been wasted. Take the case of Hongxin Semiconductor Manufacturing Company, touted as a $19.8 billion showcase project to build the world’s largest silicon wafer fab in Wuhan. Today, the abandoned project is a hulking eyesore of half-finished buildings into which more than $1 billion dollars in government funds was squandered in a spectacular Ponzi scheme.

Then, however, there is Yangtze Memory Technologies Company (YMTC), founded only in 2016 but now mass-producing state-of-the-art memory chips. YMTC has shown that China can move at China-speed in producing the leading workhorse chips underpinning everything from electric vehicles, to smartphones and laptops, to the array of Internet of Things connected products. Though it has eluded U.S. sanctions so far, the company is as focused on protecting itself as it is on producing chips. Two years ago, according to Nikkei, YMTC launched an extensive supply chain review with the goal of eliminating any dependence on the U.S. The effort, which YMTC denies, reportedly involves more than 800 people working full-time. 

As members of Congress mull whether it is necessary for the government to make significant investments in American technology innovation and manufacturing, they may want to gaze across the global semiconductor landscape to guide their thinking.

In a speech last month, TSMC founder Morris Chang expressed concerns whether TSMC’s American investments will be successful under current conditions. U.S. costs are much higher than in Taiwan and, in his view, the U.S. engineering talent pool is lacking. According to Chang, a “short term subsidy cannot make up for long-term operational disadvantage.”