Imagine running a bakery where the queue outside stretches around the block, the till is overflowing, and every loaf is spoken for weeks in advance. Your biggest problem is not that unsold bread is going stale. It’s that you cannot find enough bakers, and the ones you have keep getting poached by the baker across the road.
That, in a nutshell, is the strange bind the world’s biggest chip companies find themselves in. Business has rarely been better. Yet the thing holding them back is not weak demand or shaky finances. It is people. There simply are not enough trained hands to go around, and the ones who exist are being fought over like star footballers in a transfer window.
The latest sign of this came from ASML, the Dutch firm that makes the extraordinary machines used to print circuitry onto silicon. ASML confirmed it plans to offer eligible staff a bonus worth 20,000 euro, roughly 22,838 US dollars, if they stay with the company between 2027 and 2030.
The reward will come as a conditional stock grant starting in January 2027, with the finer details still being worked out. In plain terms, ASML is paying its own people to not leave.
Why a company this successful has to bribe its own staff
On paper, ASML looks untouchable. It is Europe’s most valuable company by stock market value. It reported a net income of 2.92 billion euro this quarter. Its flagship lithography machines, the ones that make advanced chips possible, are almost completely sold out through 2027.
The company has raised its revenue outlook for the year to somewhere between 43 and 45 billion euro, and it wants to lift production of its most advanced systems by 30 percent in 2027.
So why the nervousness? Because none of that matters if the specialists who build, install, and maintain these machines walk out of the door.
ASML is not alone here. Samsung Electronics, TSMC, and SK Hynix have all been handing out extra pay to keep workers happy. When earnings are strong across an entire industry, every company is expanding at once.
And when everyone expands at once, they all reach for the same small pool of talent at the same moment. The result is a bidding war, and retention bonuses are simply the industry’s way of saying please do not go and work for our rival.
The maths that keeps chip bosses awake
The scale of the shortage is genuinely eye-watering. Consultancy Deloitte estimates the semiconductor industry will need more than one million additional skilled workers by 2030. That works out at over 100,000 fresh recruits every single year.
To put that in perspective, fewer than 100,000 students enrol in electrical engineering and computer science courses across the entire United States each year. The industry wants to hire almost that many people annually, and it is only one of many industries fishing in the same pond.
A survey of executives found that 77% already consider the talent shortage critical, with another chunk expecting it to turn severe within three years. This is not a problem looming on the horizon. It has already arrived.
Why can’t they just train more people?
This is the obvious question, and the answer explains why money alone will not solve it.
First, these are not jobs you can learn in a weekend. Building and running a lithography machine is closer to precision surgery than factory work.
The people who understand them accumulate years of hard-won, hands-on knowledge that cannot simply be downloaded or rushed. Industry estimates suggest it can take 18 months or more just to bring a new hire up to useful speed.
ASML chief executive Christophe Fouquet has hinted that lithography is uniquely hard to copy, noting that only Japan and Europe have ever managed to build the technology at all. The know-how, in other words, sits in a very small number of heads.
Second, there is a quiet time bomb built into the workforce. A large share of the most experienced engineers and technicians are approaching retirement. Europe alone could lose close to 30% of its chip workforce to retirement by the end of the decade.
When those people leave, they take decades of institutional memory with them, the kind of instinct that tells a technician something is wrong before an alarm ever sounds.
Fouquet has argued that Europe holds a quiet advantage here, because its workforce has stayed rooted locally rather than scattering, as he put it in an interview with the business school IMD, the talent base is “still very localized.” That rootedness is a strength only for as long as those people stay.
Third, the education pipeline has drifted in the wrong direction. Over the years, universities and schools poured energy into software and computer science, the glamorous side of tech, while the more physical, hands-on disciplines that chipmaking depends on were quietly neglected. There is now a mismatch between what young graduates are trained to do and what the industry desperately needs.
The demand for equipment technicians, roles that need solid vocational training rather than a master’s degree, has been largely overlooked by traditional academic routes.
There is a sharper reason to worry about people leaving, too. When an engineer walks out, the knowledge walks with them. There have been reports that former ASML engineers in China have partly reverse-engineered the company’s technology, a reminder that in this industry a single departing specialist can carry secrets worth billions.
The AI factor
Sitting on top of all this is the artificial intelligence boom. The enormous spending on AI data centres has sent demand for advanced chips soaring, which in turn has driven demand for the machines ASML makes. AI is the reason ASML’s order books are so full.
Fouquet has admitted he did not see the boom coming, and reckons the resulting chip shortage will persist for the next three to five years. The demand is also, indirectly, making the labour crunch worse, because every chipmaker is racing to expand at the same time to feed the same AI-hungry customers.
There is a certain irony here. The technology being built to automate human work is itself blocked by a shortage of humans skilled enough to build it.
What lies ahead for ASML and its rivals
So where does this leave companies like ASML?
In the short term, expect more of the same. Retention bonuses, stock grants, and sweeteners will keep flowing as long as the talent war rages. For workers with the right skills, this is a good moment to be alive. They hold real bargaining power, perhaps more than at any point in the industry’s history.
But bonuses are a sticking plaster, not a cure. A 20,000 euro grant might persuade someone to stay another three years. It does not create a single new engineer.
The deeper fix has to come from building the pipeline, and that means working with universities and vocational colleges, investing in apprenticeships, opening the door to people from outside the traditional degree route, and making these careers visible and attractive to school leavers who have never considered them.
Governments are pushing hard too. The United States, through its CHIPS Act, and Europe, through its own Chips Act, have poured public money into building domestic chip industries.
But spreading chip production across more countries actually makes the labour problem worse, not better, because you now need skilled teams in far more locations than before. Money can build a factory in eighteen months. It cannot conjure a seasoned workforce to run it in the same time.
The likely path ahead is a decade of steady, grinding effort to widen the talent pool, paired with heavy investment in automation to squeeze more output from the people already in place. Companies that get both right will pull ahead.
Those that rely only on chequebooks to hold onto staff may find, sooner or later, that there is always a rival willing to write a slightly bigger cheque.
For now, ASML’s message to its people is simple and slightly telling. We are doing wonderfully. Please, whatever you do, do not leave.