Samsung SF1.4 production scheduled for 2029 as foundry roadmap expands

The delay of 1.4nm volume production to 2029 signals a shift toward stability and yield optimization over aggressive node shrinking. For the streaming ecosystem, this roadmap dictates the efficiency gains available for next-generation server-side encoding and high-performance edge computing hardware. By expanding the 2nm family first, Samsung aims to avoid the low-yield issues that hampered its early 3nm Gate-All-Around rollout. This conservative cadence gives competitors TSMC and Intel Foundry a window to capture market share in the high-performance computing segment. Watch for specific yield data from the SF2P+ node in 2027 as a bellwether for Samsung's 1.4nm readiness.

Samsung Foundry's SF1.4 timeline faces direct competitive pressure from TSMC, which is targeting the same 1.4nm-class generation with its A14 process. At its North America Technology Symposium in April 2025, TSMC disclosed A14 with second-generation GAA nanosheet transistors and NanoFlex Pro design-technology co-optimization, promising 10% to 15% speed improvement and 25% to 30% power reduction versus its N2 node. TSMC plans A14 mass production in 2028, a full year ahead of Samsung's SF1.4 target, and Chairman C.C. Wei confirmed at a July 2026 investor conference that A14 production scale will exceed the 2nm ramp, with A13 and A12 processes following in 2029. That cadence gives TSMC a structural lead in attracting high-performance computing and AI accelerator customers who need the earliest access to sub-2nm density gains.

The business implications extend to backside power delivery, a capability Samsung has not yet confirmed for SF1.4. TSMC's initial A14 version omits backside power delivery, but a variant with Super Power Rail backside power is planned for 2029, directly overlapping Samsung's SF1.4 window. Bloomberg reported that TSMC intends A14 to serve AI customers including Nvidia starting in 2028, positioning the foundry to capture data-center and edge-compute workloads that also feed streaming infrastructure. Intel Foundry, meanwhile, has not publicly committed to a 1.4nm-class node, leaving the two Asian foundries as the primary contenders for leading-edge HPC silicon through the end of the decade.

On the technical front, TSMC's A14 specifications set a benchmark that Samsung's SF1.4 must match or exceed to remain competitive. TSMC's own symposium materials cite more than 20% logic density improvement over N2 and evolution of its NanoFlex standard-cell architecture to NanoFlex Pro, enabling finer-grained transistor-level trade-off exploration for chip designers. Samsung's emphasis on Design Technology Co-Optimization across its SF2 family mirrors this approach, but the one-year gap in production timing means Samsung's HPC and streaming-hardware customers may need to plan around TSMC's earlier availability when selecting a foundry partner for next-generation encoding and edge-compute silicon.