← All sectors / The AI transformation
111 · Contract electronics & advanced assembly
Whoever actually builds the thing
Curve position
Growth
Binding constraint
Skilled assembly labour in the geographies customers now want.
Almost nobody who designs electronics also builds them. Contract manufacturers assemble at volumes and tolerances that would be uneconomic in house, and the current combination of AI hardware demand and reshoring policy is reshaping who does it and where.
Historically the industry consolidated in Asia on labour cost and supply chain density. Tariffs, geopolitical risk, and customer requirements for regional supply have started to pull some capacity elsewhere at higher cost.
The structural driver is customer risk management rather than economics. Buyers now pay a premium for supply that is not concentrated in one geography, which is a change from three decades of pure cost optimisation.
The technology layer spans surface mount assembly, advanced packaging and system in package, liquid cooling assembly for AI hardware, test and inspection, and the supply chain systems that manage component shortages.
Adoption economics are thin margin and volume driven. Contract manufacturing earns single digit margins, so utilisation and component procurement discipline decide profitability entirely.
The beneficiaries include contract manufacturers with capacity in favoured geographies, equipment makers selling assembly and test lines, component distributors, and the design services firms attached to manufacturing.
The value chain runs from components through assembly and test to the brand. Component procurement is where contract manufacturers make or lose money, more than assembly labour.
The overlooked layer includes assembly equipment makers, test and inspection specialists, component distributors, and the thermal and mechanical suppliers for high power systems.
Competitive dynamics favour scale and geographic breadth. Customers increasingly want the same manufacturer to build in several regions, which advantages the largest firms.
Risks: margins are structurally thin, customer concentration is severe, component shortages destroy quarters, and demand for AI hardware could normalise abruptly.
What to watch: capacity announcements by region, customer concentration disclosures, component lead times, and AI hardware order books.
