AMD, Google, Tesla, and Groq turn to Samsung Foundry as TSMC runs out of capacity - TweakTown

As TSMC's advanced chip production capacity becomes fully booked by major players like NVIDIA, Apple, and Broadcom, companies such as AMD, Google, Tesla, Groq, and BYD are increasingly turning to Sams... Read original →

With TSMC’s sub-3nm capacity locked by NVIDIA, Apple, and Broadcom, AMD, Tesla, and others are turning to Samsung Foundry—not by choice but necessity. This shift triggers a technical ripple: Samsung’s lower yields degrade chip energy efficiency, raising AI compute costs and accelerating co-design of advanced packaging with EUV layers. Geopolitically, U.S. CHIPS Act subsidies tie production to domestic soil, exposing Samsung’s Korean fabs to export control scrutiny, while Intel’s 18A-P emerges as a de-risked alternative for Apple and Google. TSMC will likely fast-track Arizona and Japan expansions, yet its 2nm ramp won’t alleviate near-term shortages. Over the next 12–24 months, the foundry market will bifurcate: premium clients monopolize TSMC’s leading nodes, while mid-tier demand flows to Samsung and Intel—permanently fragmenting the advanced semiconductor supply chain.