U.S. Chip Expansion Needs 150,000 More Workers

Shari Liss says the United States can expand its semiconductor manufacturing footprint, but the workforce is already one of the industry’s main constraints. Asked by Caroline Hyde whether the US would be able to bring more chip manufacturing and design back onto American soil if geopolitical tensions increased, rather than remaining so reliant on Taiwan, Liss answered directly: “We are.”

Her qualification was the workforce. Speaking as SEMI’s vice president of global workforce development and initiatives, Liss said the CHIPS-related investments now underway are creating “incredible growth” in US semiconductor activity, and that growth will require “another 150,000 or so people” in the work environment. The response she described is not a single national training program, but a set of programs being built around the country to meet regional needs.

The central issue, as Ed Ludlow framed it, is not whether there are plans on paper to build more fabs or expand foundry capacity in the United States. It is whether the US has enough skilled people to make those fabs “run” and “hum” in an industry with “brutal economics.” Ludlow pointed to Taiwan as a place with deep semiconductor manufacturing capacity and talent, then asked whether the US has the skills base to operate the facilities it plans to build.

Liss said the US is trying to rebuild capability after a period in which fewer domestic programs existed because the country was not manufacturing as much. “We didn’t have as many programs established in the US anymore because we weren’t manufacturing here,” she said. With new investments now arriving, programs are launching across the country.

One specific mechanism she cited is a $200 million workforce investment under the CHIPS Act, administered through the National Science Foundation in concert with the Department of Commerce. Liss described the goal as building “a national infrastructure around workforce development,” with regional nodes funded to develop the capabilities needed locally and then connect those efforts into a broader national system.

Bloomberg’s on-screen graphics during the exchange showed intraday declines for Advanced Micro Devices, Arm Holdings, ASML, Broadcom, Intel, Micron, Nvidia, and TSMC ADR. The charts were not explained in the discussion and did not assign a cause to the moves; they functioned as a market backdrop while the hosts and Liss discussed the workforce constraint behind US semiconductor expansion.

Ludlow pressed Liss on what kinds of workers the US actually needs, invoking Tim Cook’s comparison that China could fill a sports stadium with tooling engineers while America would struggle to fill a meeting room. Liss’s answer was broad: the shortage is not confined to a single elite engineering category.

The industry needs technicians and operators to work on fab floors. It needs electrical engineers, mechanical engineers, and chemical engineers. It needs researchers and PhDs. Liss also included marketing and finance talent, arguing that the semiconductor industry’s workforce requirements span the full business, not only the cleanroom.

The difficulty, in her telling, is partly educational capacity and partly demand generation among students. She said the challenge in the US is “the image and awareness of our industry with students.” SEMI’s work, as she described it, includes trying to get students excited about the industry or even aware that it exists as a career path.

Her example was the invisibility of chips in everyday life. Students carry phones, iPads, and computers. Cars depend on chips. Appliances depend on chips. Yet Liss said many people, including parents, do not connect those daily technologies to semiconductor work or understand that the industry sits underneath them: “Everything we use all day long, and most people don’t know that. Parents don’t all know that. You know, like, so it’s an educating of the country, really.”

That phrasing captured a larger point: the workforce problem is not only a question of reskilling workers already adjacent to the sector. It is also a problem of national visibility. If students do not see semiconductors as a place to work, the pipeline remains thin no matter how many fabs are announced.

Hyde raised a second tension in the US manufacturing push: some of the expansion depends on domestic players such as Intel, but much of it also involves attracting foreign manufacturers, especially TSMC, to build in the United States. That raises a practical question about workforce transfer. Has the US relied on TSMC and other foreign talent coming in to teach and run these operations, or can it train the needed workers through its own education system?

Liss did not reject the importance of foreign talent. “There’s certainly been a mix of both,” she said. The semiconductor industry is global, “intense,” “intricate,” and “complicated,” and she said it needs talent from everywhere. She also emphasized that countries and companies continue to learn from one another because different parts of the global semiconductor ecosystem have different strengths.

But her policy goal was clear: build enough workforce capacity in the United States, among US citizens, to fill US jobs created by the domestic manufacturing push. That does not mean severing the industry from global expertise. It means reducing the gap between the fabs the US wants to build and the number of Americans prepared to work in them.

The SEMI membership graphic shown during the segment underscored the scale and international character of the organization Liss was speaking from: more than 3,000 member companies, 1.3 million individual members, representation across more than 94 countries, and more than 1,000 standards. Her argument sat inside that global frame. Semiconductor manufacturing cannot be treated as a purely domestic craft, but the US expansion will fail if it cannot create a domestic labor base large enough to support it.