Samsung Reportedly Restarts SF1.4 Development for 2029 Mass Production

Samsung Electronics is reportedly accelerating development of its SF1.4 foundry process again, with mass production now targeted for 2029 as the company prepares to challenge TSMC and Intel in the next generation semiconductor manufacturing race.

According to a report from Korean publication The Bell, Samsung recently shared its SF1.4 development plans with major equipment suppliers including Applied Materials and Lam Research. The partners have reportedly been asked to begin advanced development of manufacturing tools designed around the requirements of Samsung’s future 1.4 nm class process.

The equipment is expected to be installed at NRD K, Samsung’s advanced semiconductor research complex at its Giheung Campus in South Korea. Samsung established the facility to combine research, process development, and product level verification under one location, with approximately KRW 20 trillion planned for investment through 2030.

Samsung originally announced that its 1.4 nm process would enter mass production in 2027. That schedule appeared in the company’s official 2022 foundry roadmap and was reaffirmed during Samsung Foundry Forum 2023. The new report claims the target was later moved to 2029 as Samsung redirected resources toward improving its SF2 and SF2P processes.

“Realizing every customer’s innovations with our partners has been at the core of our foundry service.”
— Siyoung Choi, former President and Head of Samsung Foundry Business

The decision to delay SF1.4 was reportedly intended to give Samsung more time to stabilize yields, optimize manufacturing, and strengthen the production infrastructure behind its 2 nm family. Instead of attempting to accelerate another highly complex node before SF2 reached sufficient maturity, the company appears to have prioritized execution on technology that can generate customer revenue sooner.

That strategy is particularly important because advanced process leadership depends on more than announcing the smallest node name. Samsung must demonstrate competitive yields, predictable capacity, mature design tools, working intellectual property libraries, and reliable customer support before major chip designers will commit expensive products to SF1.4.

The latest report suggests Samsung believes its 2 nm progress is now strong enough to support renewed investment in the following generation. Applied Materials and Lam Research are expected to develop customized equipment based on Samsung’s process requirements, with those tools eventually supporting early SF1.4 research and production development inside NRD K.

High numerical aperture extreme ultraviolet lithography is also expected to play a role. The Bell reports that ASML High NA EUV equipment has already entered NRD K and that Samsung plans to use the technology on selected SF1.4 layers. Samsung previously confirmed that NRD K was designed to receive High NA EUV systems, advanced deposition tools, and wafer bonding infrastructure, although its official announcement did not confirm when the first scanner became operational.

High NA EUV provides greater optical resolution than current EUV equipment, potentially reducing the number of patterning steps required for extremely small features. However, the systems are expensive, physically large, and require new masks, resists, metrology tools, and process integration work. Early access could help Samsung develop the supporting manufacturing ecosystem before SF1.4 reaches commercial production.

Samsung’s reported 2029 target places it behind TSMC’s current A14 schedule. TSMC officially plans to begin A14 production in 2028, with the node expected to deliver up to 15% more performance at the same power, up to 30% lower power at the same speed, and more than 20% additional logic density compared with N2. Previous report on TSMC’s A14 manufacturing expansion in Taiwan examined the company’s large scale fabrication plans and its decision to continue using existing EUV technology rather than immediately adopting High NA equipment. That approach could reduce manufacturing disruption and equipment costs, although it may require more complex patterning techniques.

Intel is approaching the same technology generation through Intel 14A. The node uses second generation RibbonFET gate all around transistors, PowerDirect backside power delivery, and High NA EUV for critical manufacturing layers. Intel claims 14A can provide between 15% and 20% more performance at the same power, between 25% and 35% lower power at the same performance, and up to 30% greater chip density compared with Intel 18A.

Recent industry reporting places Intel 14A risk production in 2028 and volume manufacturing in 2029, which would align it more closely with Samsung’s reported SF1.4 schedule than with TSMC’s earlier A14 target. Intel has already started building customer interest around the process, but reported evaluations involving companies such as Apple remain unconfirmed commercial agreements rather than finalized production contracts. Intel’s manufacturing roadmap has also gained momentum through the start of Intel 18A P risk production. The improved node provides 9% higher performance at the same power or 18% lower power at the same performance compared with Intel 18A, while retaining compatible design rules.

Samsung is reportedly developing future memory manufacturing equipment at NRD K alongside SF1.4. The Bell claims the company is preparing for V12 NAND production around 2030, while Samsung has publicly described plans for memory devices exceeding 1,000 layers and advanced wafer bonding technology at the facility.

Samsung restarting advanced SF1.4 development is an important signal, but the 2029 schedule also shows how far the competitive focus has shifted from roadmap announcements toward manufacturing discipline.

The company was once targeting 2027, which would have placed it near the front of the 1.4 nm class transition. Moving that objective to 2029 gives TSMC a potential production timing advantage, but it may also provide Samsung with the time needed to avoid repeating the yield and customer confidence challenges that affected some of its earlier advanced nodes.

High NA EUV could become one of Samsung’s strategic differentiators. Intel is also adopting the technology aggressively, while TSMC continues extracting more value from conventional EUV. The winner will not necessarily be the company using the newest tool first. It will be the foundry that converts its process into high yielding wafers, competitive costs, mature design support, and dependable production capacity.

The naming also requires perspective. Samsung SF1.4, TSMC A14, and Intel 14A are all described as 1.4 nm class technologies, but their node names do not represent identical transistor dimensions or directly comparable products. Each company uses different transistor structures, power delivery methods, cell designs, density targets, and customer markets.

Samsung therefore does not need to defeat every competitor through one specification. It needs SF1.4 to become commercially credible enough to win major mobile, artificial intelligence, automotive, and high performance computing customers. Renewed supplier engagement is the first step, but yield development and customer adoption will determine whether the process becomes a genuine third option against TSMC and Intel.

Can Samsung close the foundry gap with SF1.4 in 2029, or will TSMC’s earlier A14 production schedule prove too difficult to overcome?