Samsung Electro-Mechanics has reportedly secured a leading position in the supply chain for advanced substrates used in the Groq 3 language processing unit (LPU), a next-generation inference chip integrated into Nvidia’s upcoming Vera Rubin AI platform.

Artificial intelligence is entering a capital-intensive phase, with industry executives warning that Asia’s next generation of factories will be defined less by breakthroughs in algorithms than by constraints in energy, infrastructure, and capital allocation.

Taiwan’s listed AI hardware companies collectively generated $69.7 billion in March 2026 revenue across 13 supply-chain segments — up 63% year-over-year — offering the most comprehensive single-month snapshot yet of where global AI infrastructure spending is actually flowing. The table below covers 49 companies from TSMC’s silicon foundry all the way down to the rail kits that slide servers into racks. Read together, the numbers tell a story that goes well beyond any single company’s earnings call.