NNewsGPT ← Home
CN

Samsung's 4nm Chip Capacity Sold Out, 8nm Running Near Full; Selectively Accepting Orders

CN1 hr ago

Samsung Electronics' foundry business has adjusted its supply strategy, prioritizing existing clients and selectively accepting new orders, according to industry insiders cited by ChosunBiz. Analysts attribute this shift to the explosive growth in the artificial intelligence market, which is fundamentally altering demand for foundry services. Previously, advanced manufacturing processes were primarily sought for smartphone application processors (APs). However, demand has now pivoted towards AI accelerator chips, application-specific integrated circuit (ASIC) chips, and high-performance computing (HPC) chips.

Samsung's 4nm process capacity is reportedly almost entirely sold out, with production slots for next year also fully booked. Furthermore, some of its 8nm process capacity is operating at near full load. This situation indicates a significant demand surge for advanced semiconductor manufacturing, driven by the burgeoning AI sector.

AI Analysis

The semiconductor foundry market is experiencing a structural shift, with demand for advanced nodes like 4nm and 8nm being significantly influenced by the rapid expansion of AI applications. This surge in demand for AI-specific chips, such as accelerators and ASICs, is reallocating capacity away from traditional markets like smartphone APs. Samsung's foundry, by prioritizing existing clients and selectively accepting new orders, is navigating this high-demand environment by optimizing its resource allocation. This strategy aims to maximize revenue and solidify relationships with key partners in a period of intense competition and technological evolution. The long-term implications involve potential bottlenecks for emerging players and a continued emphasis on scaling advanced manufacturing capabilities to meet future AI-driven computational needs.

AI-generated to prompt reflection — not editorial opinion, not advice, not a statement of fact. How this works.

Compiled by NewsGPT from 36Kr (CN). Read the original for full details.