The eBeam Initiative conducted its 15th annual Luminaries survey in July 2026 and announced the results at the annual reception held during the SPIE Photomask Technology + EUV Lithography Conference on September 9. A total of 85 industry professionals participated, representing over 45 companies.
Following the annual reception, an expert panel discussed the survey findings, covering topics such as large-sized masks, equipment investment, High-NA EUV, and curved masks. The discussion was moderated by Aki Fujimura, CEO of D2S and co-founder of the eBeam Initiative, and included participants Harry Levinson of HJL Lithography, Naoya Hayashi, Honorary Researcher at DNP, and Glen Scheid of Micron Technology (see Figure 1). This article reports the key insights shared by the panel.
eBeam Initiative (Fujimura): 66% of industry professionals expect the industry to adopt large-format masks, approximately 6 x 12 inches, within the next decade. This survey was conducted prior to TSMC and ASML announcing an industry initiative aimed at establishing a 12-inch mask pilot line by 2031. Why is this shift occurring now?
Levinson: The rationale from the wafer fab side is productivity. High-NA EUV systems are extremely expensive and occupy valuable cleanroom space. Larger masks can expose a larger field of view without introducing the process control complexity of A/B stitching, theoretically increasing productivity by approximately 40%. This advantage is critical when each additional system increases facility and power requirements.
Schindler: The consensus formed around the 12-inch specification is encouraging. A transition in mask size is extremely rare, and this time there are clear productivity gains—exactly what the industry needs.
eBeam Initiative (Fujimura): Harry, you were involved in earlier attempts to standardize 9-inch masks. Why is this time different?
Levinson: The 9-inch proposal was extrapolated from historical trends in DRAM chip size growth. However, target customers had no plans to manufacture chips of that size, so this specification did not address any real-world problems and was ultimately rejected. This time, demand came directly from chip manufacturers. Intel initiated the discussion, and now TSMC, ASML, and other companies are supporting it. The use cases and value are clearly defined.
eBeam Initiative (Fujimura): What does the entire supply chain need to do to achieve the 2031 pilot production line goal?
Hayashi: The entire ecosystem must be ready simultaneously. Some critical links in the mask manufacturing chain have only one supplier. We need to closely monitor the ecosystem and support those areas that could hinder the availability of the complete process when needed.
Scheid: This is a collective schedule, not a set of independent timelines. The more steps required, the lower the overall probability of success. Any supplier delay, especially in single-source环节, will slow down the transition. Redundancy can reduce risk, but this is a small industry, and the investment required may be too large to support two suppliers in every category.
Levinson: There are both financial and technical risks here. Some tools may require larger footprints and different handling methods, and processes such as uniform etching also warrant closer attention. Financial risks should be addressed early, especially for smaller suppliers. Companies planning their fabs according to this timeline are concerned whether all components will be ready on schedule.
eBeam Initiative (Fujimura): What new manufacturing challenges arise when the area of a rectangular mask is doubled?
Hayashi: Spin-coating processes will become more challenging. Photoresist coating, development, cleaning, and drying must remain uniform across larger rectangular surfaces. Longer edges and corners may cause water droplet issues at the corners, requiring thorough evaluation of these processes.
Levinson: It took about 15 years for EUV multilayer coatings to reach a level suitable for mass production. Many lessons can be transferred, but scaling up the size of multilayer blank substrates is not as simple as installing a few larger machines.
Scheid: The larger the pellicle area, the greater the opportunity for particle contamination; according to another survey, contamination and cleaning are already among the primary concerns for EUV mask lifetime. While larger pellicles may benefit fabs, mask factories must still apply uniform chemical treatments to larger surfaces without increasing defects.
eBeam Initiative (Fujimura): Has High-NA EUV shifted from “if it will happen” to “when it will happen”?
Levinson: Yes. Just a year ago, people were questioning whether High-NA even existed. Now it’s already in production at Intel, and multiple companies have publicly released roadmaps—this is a major shift. Samsung has announced mass production on High-NA by 2028, while TSMC plans to adopt it for large-scale manufacturing of advanced nodes starting in 2030. Demand beyond High-NA remains uncertain: device and interconnect engineers must first demonstrate that smaller feature sizes deliver the required performance.
Figure 2: The 2026 eBeam Initiative survey results show an increase in equipment procurement over the next three years.
eBeam Initiative (Fujimura): 68% of industry professionals expect an increase in mask etching system procurement over the next three years (see Figure 2). What are the driving factors behind this result?
Schmidt: Purchases of multi-beam writers and inspection tools are expected to increase over the next three years, reflecting both capacity growth and capability demands. Etching, however, is different. New High-NA mask materials may be incompatible with existing systems and may not operate in parallel with .33 NA materials. Surveys predict that etching equipment purchases will increase over the next three years, not primarily to boost etching capacity, but to enable new capabilities.
Levinson: Meanwhile, requirements for uniformity are becoming stricter. Enhancing uniformity while introducing new materials has made etching a highly active area of R&D.
eBeam Initiative (Fujimura): What about the curved wafer target?
Scheid: The curve mask has already emerged, enabled by multi-beam writers and supported by formats such as MULTIGON. However, within a three-year time frame, it is more difficult to see automated solutions for generating complete layers tailored to curved wafer targets.
Levinson: The advantages demonstrated by curved wafer features are compelling, but backside interconnects and other changes have introduced uncertainty into the timeline. On the mask side, curved technology has arrived and will continue to evolve.
Hayashi: Creating curved masks using a multi-beam writing mechanism is not complex. Certification remains a challenge, but the level of attention this topic has received at this conference indicates that the industry is making progress on it.
The group’s final conclusion was optimistic, but not complacent. The productivity logic for large-format masks is strong; High-NA is now in mass production and also appears on the roadmaps of two other companies, with mask facilities investing in new equipment and curved masks. The truly challenging part will be coordinating the entire ecosystem. The industry’s history of solving common technical challenges may be its most important advantage on the path to 2031.
Reference link:
- 2026 eBeam Initiative Luminaries Survey
- TSMC-ASML Large-Format Mask Initiative
- Breaking News by Aki Fujimura from the 12-Inch Mask Summit