Samsung Delays Advanced Chipmaking to 2030, Skips High-NA EUV Until 1nm

What happened

Samsung Electronics announced a significant revision to its semiconductor manufacturing roadmap at the 2026 Next-Generation Lithography + Patterning Conference in Korea. The company pushed back its SF1.4 (1.4nm-class) process from 2027 to 2029 and will now spend the next several years refining its 2nm-class SF2 family, including variants SF2P, SF2X, SF2A, and SF2Z with backside power delivery. Samsung's first 1nm-class SF1A process is now targeted for around 2030. The company also clarified that it does not plan to use High-NA EUV lithography for 2nm or 1.4nm production. Instead, this most advanced chip-printing technology-featuring a 0.55NA numerical aperture-will become necessary only from the 1nm-class generation and below, around 2030. Samsung VP of Technology ChangMin Park explained that the firm had originally expected to use High-NA EUV machines for volume production at the 2nm and 1.4nm nodes, but the technology's lack of refinement led to the delay. High-NA EUV enables smaller circuit sizes through single-pattern printing, avoiding the costs, complexity, and defects associated with multi-patterning required by traditional low-NA EUV machines at 13.5 nanometers wavelength.

Why it matters

Intel competes directly with Samsung in the foundry business, where both are racing to catch TSMC's manufacturing lead and attract customers like AMD, NVIDIA, and Qualcomm. Samsung's roadmap delay reduces near-term competitive pressure on TSMC, potentially strengthening the Taiwanese foundry's market position and giving it more time to secure long-term wafer contracts. For Intel, which is working to establish its 18A and 14A nodes as viable alternatives, Samsung's slower cadence could create an opportunity to close the technology gap if Intel executes well on its own advanced-process roadmap. The delayed adoption of High-NA EUV also signals industrywide challenges with the technology's maturity. TSMC has already signalled it is exploring alternatives like photomask pellicless technology instead of rushing to High-NA EUV, suggesting the equipment may not deliver sufficient yield or cost advantages yet. Intel must carefully evaluate its own High-NA strategy, balancing the technology's potential benefits against commercial viability and customer demand timelines.

Bigger picture

Leading-edge semiconductor manufacturing has become extraordinarily expensive, and all major foundries are grappling with yield challenges and utilization rates at advanced nodes. Samsung's decision to extend the life of its SF2 family rather than racing through node names reflects a broader industry shift toward stabilizing production and improving commercial viability before committing customers to major transitions. High-NA EUV lithography represents a critical technology for future chipmaking, but its adoption is proving slower than equipment maker ASML and chipmakers initially anticipated. The technology allows manufacturers to print smaller circuits in a single step, reducing defects and complexity, but integration challenges are delaying deployment across the industry. Samsung remains one of only three companies capable of manufacturing advanced logic chips at scale, alongside TSMC and Intel. Its roadmap decisions influence long-term wafer pricing, production capacity availability, and competitive dynamics for chip designers including AMD, NVIDIA, Qualcomm, and others who rely on external foundries. The foundry market is consolidating around fewer viable options at the leading edge, making each player's technology progress strategically important.

What to watch

Monitor Intel's progress on its 18A and 14A process nodes, which are critical to the company's foundry ambitions and ability to compete with TSMC and Samsung. Watch for announcements about Intel's own High-NA EUV adoption timeline and whether it follows Samsung's more cautious approach or attempts to gain advantage through earlier deployment. Track foundry customer wins and design tape-outs, particularly from AMD, NVIDIA, and Qualcomm, which will signal which manufacturer is successfully balancing technology advancement with commercial viability. Pay attention to TSMC's roadmap updates and whether it continues pursuing alternatives to High-NA EUV or eventually commits to the technology. Finally, watch Samsung's 2nm-class yield improvements and customer adoption, as successful stabilization of SF2 could position the company to compete more effectively when it does reach 1nm production around 2030.