AI models keep getting bigger, but the limiting factor may be people, not transistors. A new McKinsey and SEMI Foundation analysis projects the U.S. could be missing as many as 157,000 semiconductor workers by 2030. That shortfall lands just as the country rushes into its fastest factory buildout yet to feed AI's demand for memory and logic chips.
The talent pool is thin. Each year, just 3% of U.S. grads in engineering jobs end up in semiconductors, and 73% of chip employers say engineering hires are hard to fill. TSMC and Intel are already turning out advanced logic from new Arizona fabs, and Samsung is next in line with advanced logic production in Taylor, Texas later this year. The sticking point is staffing.
In Austin, Samsung's semiconductor division EVP Jon Taylor said, "I'm concerned." "We just don't see that there's enough technical people in the pipeline."
Samsung expects to begin ramping advanced logic in Taylor this year in the initial facility of two new fabs, part of a $35 billion Texas expansion projected to create about 3,500 jobs. "We're hiring engineers, we're hiring technicians, we're hiring people in supply chain," Taylor said. "Everybody wants and needs the same thing, and it's a bit of a race against time right now as everything is starting to come online." A July 2026 photo shows Samsung's facility in Taylor, Texas that manufactures advanced logic chips.
Much of the know-how still sits in Asia. "If you look at where advanced technology currently resides, it is in Asia," Taylor said. Samsung and SK Hynix are bringing over South Korean employees temporarily to jumpstart new U.S. facilities. Samsung also flew U.S. staff to South Korea for months of hands-on training, echoing TSMC's approach in Arizona in 2021. ASML is pitching in with a new U.S.-based training center.
Bringing in foreign experts on H-1B visas is costly and slow, so companies are trying to grow talent at home. One sign of the gap today: Idaho-based Micron has been conducting rapid-fire hiring sessions at South Korea's technical universities, where it interviews students and issues offers for roles at its Asia fabs in a single day.
Pay does not make it easier. U.S. chip roles typically range from $127,000 to $187,000, with senior positions topping $238,000, the SEMI Foundation reports. In South Korea, where 12-hour shifts are common and chip workers are especially prized, strike threats at Micron, Samsung, and SK Hynix helped push bonuses above $500,000.
Memory makers are adding capacity fast amid a global shortage that has Nvidia and Advanced Micro Devices working to lock in supply. The epicenter right now is South Korea, where SK Hynix will start to bring the world's largest memory campus online in February.
When industries face growing pains, patient investors focus on steady, long-term positioning. Join Briefs Finance CEO Jaspreet Singh on September 29th for a FREE live investor workshop, How to Profit From A Dollar That's Losing its Value, where he shows how we're spotting investment opportunities as the dollar falls. Save your spot.
Micron's advanced memory production is presently entirely in Asia; however, the company is constructing the first U.S. advanced memory fab in Boise, Idaho, slated to open next year, and it has a second site underway in Clay, New York. SK Hynix is likewise putting $4 billion into its first U.S. facility located at Purdue Research Park in West Lafayette, Indiana, concentrating on packaging instead of front-end manufacturing. As a result, SK Hynix and Micron will vie for a scarce U.S. memory talent pool.
Reuters reported Wednesday that SK Hynix is discussing, for the first time, making its memory chips in the U.S. with Intel, potentially by leasing a portion of Intel's Ohio fab. At a July event, Commerce Secretary Howard Lutnick pressed SK Hynix and Samsung to set up U.S.-based front-end memory factories. Taylor underscored the bigger constraint: "Whether it's memory or whether it's logic, I still will be pressed to find all the resources to be able to run those factories."
"We can secure a number of high-quality employees within a few years, based on the collaboration with the local community," SK Hynix CEO Kwak Noh-Jung said at the Indiana groundbreaking. "It's really the critical issue." A related video drives home the scale: "SK Hynix's $720 billion bet to end the memory shortage."
Universities and community colleges are racing to close the gap, investing millions in chip training. Purdue launched semiconductor degrees in 2022, and about 2,500 students take chip-related classes each semester. Its Birck Nanotechnology Center gives students time on tools similar to what SK Hynix will use nearby. "If you want young people specifically prepared in semiconductors, this is the one place you can come right now for people who have actually majored in and achieved a degree in that," said Mitch Daniels, who serves as Purdue's interim president and is a former Indiana governor. "And our job fairs are always overbooked."
ASU turned a former Motorola fabrication plant into the MacroTechnology Works cleanroom and, with $200 million from Applied Materials, launched a new center. "We're already the largest producer of college graduates for the Intel Corporation globally, and so we're trying to now become that for TSMC," ASU president Michael Crow said.
In 2024, Intel kicked off an apprenticeship program in Arizona and pledged $50 million in scholarships spread across more than 80 Ohio institutions located near its under-construction fab. It also introduced a wider semiconductor education pathways program in 2025 that begins at the K-12 level. Samsung began its workforce initiative in 2023, donating millions to the University of Texas, Texas A&M, and the University of Illinois to seed new chip engineering programs.
Last year it put $1 million toward creating technical laboratories at Taylor High School, and it recruits over 100 interns annually. Since the CHIPS Act in 2022, upwards of 80 community colleges have started or enlarged semiconductor programs, aided by a $200 million federal fund to boost the domestic chip workforce.
Here is the bottom line. Even as AI labs talk about slowing models for safety, the hardware side already has a speed cap that comes down to people who can run fabs. Companies are importing expertise and ramping training, but the hiring math is tight. If staffing lags, factory ramps can slip, which can shape the timing and availability of the chips that power AI.
Amid constant change, keeping a disciplined plan can safeguard and grow savings. Our CEO Jaspreet Singh is hosting a FREE live investor workshop, How to Profit From A Dollar That's Losing its Value, on September 29th. Sign up free to join him live.
