Semiconductor process nodes by company, status, and generation. Leading-edge to mature node landscape.
12nm FinFET for automotive, RF, and embedded memory applications.
FD-SOI process for ultra-low-power IoT and mobile RF applications.
Intel's most advanced node: RibbonFET (GAA) + PowerVia (backside power).
Intel 3 — enhanced Intel 4 with FinFET improvements. Granite Rapids server CPUs.
Intel's first EUV process node. Used for Meteor Lake client CPU.
Samsung 2nm-class GAA nanosheet. Competing with TSMC N2.
Samsung 3nm-class GAA (first in industry at 3nm).
Samsung 4nm-class FinFET. Qualcomm Snapdragon 8 Gen 2/3 lead customer.
SMIC's most advanced node (~7nm-class DUV multi-patterning, no EUV).
TSMC first GAA (nanosheet) node at 2nm-class. Apple A20 lead customer.
N2P — enhanced N2 with backside power rail option (BSPDN).
TSMC 3nm-class (N3) — first-gen FinFET 3nm. Mass production 2022.
N3E — enhanced N3 with improved yield and PPA. Main N3 volume node.
⚠️ SPECULATIVE — Hypothetical 1nm-class future process. Illustrative only.
Status Legend
* Node status and mass production years are demo estimates from public announcements. MP = mass production start year. Terafab nodes are speculative/fictional.