Samsung Foundry
Samsung Electronics Foundry Division is a South Korea chip foundry · $12.5B revenue · 20,000 employees.
Samsung Electronics Foundry Division is a South Korea chip foundry · $12.5B revenue · 20,000 employees.
Basic
Samsung Electronics Foundry Division (founded 2017) is a semiconductor foundry headquartered in South Korea. Led by Kyehyun Kyung. Leading nodes: advanced processes.
Deep
Samsung Electronics Foundry Division operates from South Korea with approximately 20,000 employees and $12.5B in annual revenue. Founded in 2017. Leading process nodes: multiple advanced nodes. BenchGecko tracks Samsung Foundry on the /foundries page with capacity, node portfolio, client mix, and geography data.
Expert
Samsung Electronics Foundry Division · revenue $12.5B, employees 20,000, HQ South Korea. Process leadership: . Foundry economics turn on yield, utilization rate, and node leadership timing. AI-accelerator demand has restructured client mix toward a small number of large customers (NVIDIA, AMD, Google, Amazon), which concentrates revenue but also exposes to buyer concentration risk. Capex cycles for leading-edge nodes ($30B+ per fab for N2 generation) set the multi-year competitive landscape.
Depending on why you're here
- ·Samsung Foundry: advanced nodes
- ·HQ South Korea · $12.5B revenue · 20,000 employees
- ·Founded 2017 · tracked on /foundries/samsung-foundry
- ·Samsung Foundry fabricates the chips your AI provider rents you
- ·Node advances from Samsung Foundry show up in the next GPU generation's spec sheet
- ·Cross-reference /foundries/samsung-foundry with /hardware to see chip-to-foundry mapping
- ·Samsung Foundry revenue $12.5B
- ·Capex cycles for leading-edge nodes shape competitive position for 3-5 years
- ·AI-accelerator demand drives margin · HBM + advanced packaging are the new bottlenecks
- ·Samsung Foundry physically makes the chips that run AI
- ·Located in South Korea
- ·Without foundries, there is no AI hardware · they're the factory layer
Samsung Foundry is one of fewer than 10 foundries that matter in AI. Their node timeline IS the AI hardware timeline.