Nexstrom Scales Two-Dimensional Materials for Next-Generation Chip Production

The Silicon Ceiling

Silicon transistors have been shrinking for half a century, but the physics are catching up. At sub-3-nanometre nodes, controlling electron flow becomes a materials problem: leakage currents climb, power efficiency drops, and thermal management turns into an engineering nightmare. Nexstrom, a Singapore-based semiconductor equipment startup, is betting that the answer lies not in pushing silicon further but in replacing it with materials just a few atoms thick.

The company has now raised $15 million in total funding, including a $12 million seed round announced this month, to commercialise equipment that grows transition-metal dichalcogenides (TMDs) directly on 300mm wafers. Investors in the seed round include Xora Innovation, Foothill Ventures, and SEEDS Capital. Nexstrom's co-founder and chief scientist, Lance Li, previously led post-silicon electronics research at TSMC and has been working on two-dimensional materials since 2012.

TMDs are not truly two-dimensional in the geometric sense; they are simply so thin - typically just a few atomic layers - that they fall into the family of materials labelled "2D" to emphasise their extreme thinness. That thinness is precisely what makes them attractive to chipmakers searching for ways to continue scaling transistor density as silicon approaches its physical limits.

Bridging Lab and Fab

The gap between laboratory demonstrations and industrial production is where most advanced materials falter. Researchers can produce small, high-quality samples of TMDs in controlled environments, but scaling that process to cover a full 300mm wafer - the standard size in commercial chip fabrication - has remained elusive. Nexstrom's product, called North Star, is designed to address that gap by growing TMDs uniformly across entire wafers in a process compatible with existing foundry workflows.

According to Nexstrom, the company has systematically scaled its process from 2-inch to 6-inch wafers and plans to complete qualification on 8-inch wafers by the end of October 2026. A 12-inch (300mm) system has already been installed, and the company is now focused on delivering material quality that meets the specifications its industry partners have set. Nexstrom has not disclosed which companies are testing its samples, but it says several firms in the semiconductor supply chain are evaluating its wafers and materials.

The challenge is not merely one of size. Uniformity, defect density, and repeatability - the metrics that determine whether a process can be trusted in volume manufacturing - all become harder to control as wafer size increases. At Opentechwire, we have tracked similar scaling bottlenecks in other advanced materials, from graphene to III-V compounds, where laboratory promise rarely translates into foundry-ready processes within a decade.

A Crowded but Fragmented Field

Nexstrom is not the only entity pursuing two-dimensional semiconductors. TSMC, Intel, and Belgium's IMEC are all developing integration pathways for 2D materials in transistor architectures. Equipment and materials suppliers including AIXTRON and CDimension are working on adjacent parts of the manufacturing chain. In a notable demonstration, TSMC, ASML, and IMEC recently showed how 2D-material transistors could be integrated into a 300mm process flow, signalling that the industry is taking the technology seriously.

Yet the ecosystem remains fragmented. Equipment makers, materials suppliers, and chipmakers are all experimenting with different approaches, and no single process has emerged as the standard. Nexstrom's strategy is to position itself as a supplier to foundries rather than competing with them directly. The company plans to sell its equipment or collaborate with partners such as TSMC, Samsung, and Intel, rather than attempting to build its own fab.

This is a pragmatic choice. The capital intensity of semiconductor manufacturing means that startups entering the foundry business face insurmountable barriers. By focusing on a single, difficult step in the process - growing TMDs at scale - Nexstrom can carve out a niche without requiring the billions of dollars needed to build a fab or the decades of process integration knowledge that the incumbents possess.

Timeline and Technical Hurdles

Nexstrom expects its equipment to be ready for commercial production between 2030 and 2035. That timeline reflects the reality of semiconductor adoption cycles: even after a process is proven, qualification, integration, and ramp-up can take years. The company plans to use the new funding to improve process control and metrology, expand its engineering team, and support qualification programmes with potential customers.

The technical hurdles are non-trivial. Growing TMDs uniformly across a 300mm wafer requires precise control of temperature, precursor flow, and substrate preparation. Defects that would be tolerable in a research setting become yield killers in production. Edge effects, where material properties vary near the wafer's perimeter, must be minimised. And all of this must be achieved in a process that can be repeated thousands of times without drift.

Moreover, integrating 2D materials into existing transistor architectures is not a simple substitution. Contact resistance, interface engineering, and compatibility with gate dielectrics all require optimisation. The industry has spent decades refining silicon processes; replicating that level of maturity with new materials will not happen quickly.

The Data Centre Imperative

The push for post-silicon materials is being driven in part by the relentless growth of data centre demand. Training and inference workloads for AI models require chips that can deliver more compute per watt, and silicon's diminishing returns are becoming a constraint. Foundries are under pressure to deliver next-generation nodes that can meet those requirements, and 2D materials are one of several pathways being explored.

Other approaches include gate-all-around (GAA) transistors, backside power delivery, and chiplet architectures. Each addresses a different aspect of the scaling problem, and it is likely that future nodes will combine multiple innovations rather than relying on a single breakthrough. The question for Nexstrom is whether its technology can become part of that package before the window of opportunity closes.

Risk and Realism

Nexstrom's timeline to 2030-2035 is both ambitious and realistic. It acknowledges the years required to move from prototype to production, but it also assumes that the industry will continue to prioritise 2D materials over that period. If alternative pathways prove more practical, or if silicon scaling finds new life through innovations such as new lithography techniques or novel doping strategies, demand for TMDs could soften.

The company's reliance on partnerships with foundries also introduces execution risk. Qualification cycles are long, and foundries are notoriously conservative about adopting new materials or processes that could jeopardise yield. Nexstrom will need to demonstrate not just technical capability but also reliability and cost-competitiveness.

Still, the company has assembled a credible team and secured backing from investors with experience in deep tech. Li's background at TSMC provides both technical credibility and industry connections, and the company's stepwise scaling approach - from 2-inch to 6-inch to 8-inch to 12-inch - suggests a methodical engineering culture rather than a rush to hype.

At Opentechwire, we see Nexstrom as part of a broader shift in semiconductor innovation away from pure lithography scaling towards materials and architecture innovation. Whether TMDs become a mainstream technology or remain a niche solution will depend on dozens of factors, many of them outside Nexstrom's control. But the company's focus on a specific, difficult problem - growing 2D materials at scale - gives it a clearer path than many of the startups trying to reinvent the entire chip stack.