The Semiconductor Engineer Shortage Is a Bigger Problem Than the Chip Shortage

In 2021, the world ran out of chips. The shortage disrupted automobile production, consumer electronics manufacturing, and industrial equipment supply chains for eighteen months. It was a logistics and inventory management failure, exacerbated by pandemic demand shocks, and it resolved as supply chains adjusted and new capacity came online.

The semiconductor engineer shortage is structurally different. It is not resolving.

The global demand for semiconductor engineers — process engineers, device engineers, design engineers, packaging engineers — is growing at approximately 8% annually, driven by the AI buildout, the electrification of transportation, and the expansion of data centre capacity. The supply of graduating semiconductor engineers is growing at approximately 2% annually. This gap has been present for a decade. It is widening.

Why Training More Engineers Is Hard

Semiconductor engineering is not a field you can learn from YouTube tutorials and a bootcamp. A process engineer working on leading-edge fabrication nodes needs a graduate degree in materials science, electrical engineering, or physics, plus three to five years of hands-on fab experience before they are independently productive. The academic pipeline is slow and fixed.

The hands-on experience problem is more severe. Leading-edge fabrication equipment — the extreme ultraviolet lithography machines made by ASML, the etch and deposition systems from Applied Materials and Lam Research — is concentrated in a small number of fabs operated by TSMC, Samsung, and Intel. Training new engineers requires access to that equipment. Access requires employment at one of those fabs. The fabs have more demand for experienced engineers than supply, which means they have limited capacity to train inexperienced ones.

The Geopolitical Dimension

The engineer shortage interacts uncomfortably with geopolitical efforts to reshore semiconductor manufacturing. The CHIPS Act is funding new fab construction in the US. The EU Chips Act is funding new fabs in Germany, Ireland, and the Netherlands. Japan has attracted TSMC and Samsung investment with government subsidies.

None of these initiatives has adequately addressed the question of who will staff the new fabs. TSMC’s Arizona facility is staffed in part by engineers rotated from Taiwan — a solution that works for initial operations and does not scale to a fully independent domestic semiconductor industry. The engineers don’t exist yet. Training them will take fifteen to twenty years from the beginning of a serious national commitment to engineering education. No major chip initiative has announced that commitment.