Samsung foundry targets robotics, auto AI chips with Cadence platform

Samsung’s move into physical AI chips via Cadence is a defensive pivot against shrinking advanced-node demand. It will accelerate automotive and robotics SoCs toward 3nm/4nm nodes, pressuring TSMC (Taiwan, China) to open InFO-AIP packaging beyond Apple and pushing UMC and GlobalFoundries to upgrade automotive IP portfolios. With U.S. AI export controls spilling into industrial edge devices, Samsung’s Korea-based EDA-to-fab integration may reduce compliance exposure—but its automotive qualification lags TSMC by over 18 months. NVIDIA and Qualcomm are likely to recalibrate foundry strategies: NVIDIA may deepen Samsung ties for supply diversification, while Qualcomm could fast-track in-house automotive NPUs on mature nodes. Over the next 24 months, industrial AI chips will become the new battleground for foundry margins. Samsung’s success hinges not on technology, but on securing design wins with at least two of the world’s top-five Tier 1 suppliers.