Money builds fabs. People run them.
The fabs are coming. The tools are being installed. The question the industry — and policymakers — need to answer with the same urgency they brought to capital investment is this: who, exactly, is going to run them?
By Tina Rupp, National Leader, Semiconductor – Actalent
The CHIPS and Science Act has done something remarkable: by unlocking billions of dollars in federal support, it has made it economically viable for companies to build semiconductor fabs on American soil. With up to 35% total federal support available when combining tax credits and discretionary CHIPS Act funding for qualifying projects, the financial math that once made domestic manufacturing a non-starter has fundamentally changed. There are nearly one hundred manufacturing projects now underway across the United States.
And yet, standing in the middle of this historic industrial revival, I keep hearing the same sentence from clients: “I need more people.”
Every time I hear it, I ask the same question back: “Why? What will they be doing? When do you actually need them to be productive in their role?” The pause that follows tells me everything. The CHIPS Act solved the financial capital problem. The human capital problem is another story.
Facility development outpaces talent
Federal investment can fund construction, tools and equipment. It cannot manufacture the engineers and technicians needed to operate those tools, and that distinction is now coming into sharp relief as facilities approach operational readiness.
There has long been a talent gap problem in semiconductor manufacturing. What the CHIPS Act has done is magnified that existing problem by increasing the scale and urgency of manufacturing needs. Companies that offshore their manufacturing didn’t just move production — they moved the knowledge base, the training pipelines, the career pathways and the institutional memory that takes decades to build. The CHIPS Act provides funding to reassemble the physical infrastructure in a matter of years. It provides no equivalent mechanism for rebuilding the human infrastructure.
The result is a collision that I see playing out in real time with clients across the country: facilities coming online ahead of the workforce capable of running them.
The rude awakening already happening
There is a geographic reality that rarely enters the policy conversation. Semiconductor fabs require enormous amounts of land. Large tracts of affordable land tend to exist in areas with limited existing infrastructure — fewer housing options, fewer services, fewer residents who might fill those estimated 2,000-person headcount growth targets. Companies are not just competing for scarce technical talent nationally; they are asking that talent relocate to communities still being built around the facilities themselves.
Layer onto that a deeper problem: many companies reshoring to the United States are expecting to find the same depth of semiconductor-specific expertise they’ve come to rely on in Taiwan, South Korea, Japan and Europe. They are discovering that expectation was naïve. Those ecosystems were cultivated over generations. The skilled technician workforce, the process engineering bench strength, the institutional knowledge embedded in experienced teams — none of that exists at scale in the United States right now, and it cannot be conjured by a tax credit.
Workforce strategy is not simply an HR function
The companies I work with that are navigating this well share a common trait: they don’t treat talent as a number. They treat it as a sequence of institutional bottlenecks to be solved.
The most effective approach I’ve seen breaks overwhelming headcount targets — those 2,000-person ramps that paralyze talent acquisition teams — into phased, time-bound objectives tied directly to facility milestones. Rather than planning against abstract future numbers, the focus shifts to near-term operational triggers: what needs to be accomplished, by whom, and by when.
That last point is one of the most consistently overlooked in semiconductor workforce planning. Bringing someone on 30 to 60 days before the triggering event — the tool install, the cleanroom qualification, the production ramp — makes the difference between a contributor and a liability on day one. Companies that wait until the last moment to hire are competing in a crowded market for the same candidates, paying premium rates and still getting people who need months of ramp time.
The harder conversation, and the one more clients need to have, is about the root cause behind the hiring request. When a manager says, “I need an engineer,” the real question is: what business bottleneck does that engineer resolve? What work is not getting done today and what is the consequence of it staying undone? Companies that approach talent with that level of operational specificity make better decisions, move faster and get more from their investment in workforce development.
What a real strategy actually requires
Upskilling programs matter. Community college partnerships matter. Veterans’ pathways and apprenticeship models matter. But none of these are sufficient on their own, and none of them operate on a timeline compatible with facilities that are opening now.
A national talent strategy for semiconductor manufacturing has to acknowledge several things simultaneously: pipeline gap is structural and will take years to close; near-term workforce needs require creative solutions including selective high-skilled immigration pathways; offshore and nearshore staffing models can play a legitimate role for non-core functions, freeing domestic talent for the roles that require it most; and companies which treat workforce development as a strategic priority (versus a simple matter of procurement) will be the ones that hit their ramp targets.
Companies currently trying to manage 2,000-person ramps with small internal recruitment teams are going to feel the cost of that choice. The semiconductor industry is margin-obsessed for good reason: missing targets in this environment is existential. The companies that recognize they need strategic talent partners — not just staffing vendors — and make those partnerships early, are building a structural advantage over those still operating as if traditional hiring approaches will scale.
The limiting factor was never the fab
The CHIPS Act proved that domestic semiconductor manufacturing is economically viable again.
The fabs are coming. The tools are being installed. The question the industry — and policymakers — need to answer with the same urgency they brought to capital investment is this: who, exactly, is going to run them?
It is the semiconductor industry’s most important strategic challenge right now. And the clock is already running.
