Nexstrom Raises $12M to Bring Ultra-Thin Materials to Chip Fabs
Singapore semiconductor startup Nexstrom has raised a $12 million seed round to bring transition-metal dichalcogenides, or TMDs, into commercial chip factories. Total funding to date is $15 million, according to TechCrunch.
Nexstrom was incubated by Xora Innovation. The seed round drew participation from Xora, Foothill Ventures and SEEDS. The company is led by CEO Phoebe Tan. Co-founder and chief scientist Lance Li previously led research on post-silicon electronics at TSMC.
The technical bet centers on equipment and process integration, not chip design. Nexstrom is building hardware and process recipes intended to let foundries manufacture TMDs directly on wafers. TMDs are semiconductors only a few atoms thick. Its product, called North Star, is designed to grow TMDs directly on 300mm wafers, the large discs used as the standard base in volume chipmaking.
Nexstrom said it aims to sell its technology to or partner with chipmakers including TSMC, Samsung and Intel, rather than compete with them as a manufacturer. The company is based in Singapore. The city-state attracted over S$30 billion in semiconductor investments from 2022 to 2025, according to EDB.
The broader context here is the gap between materials research and factory approval. The relevant claim is not that TMDs can be made in a lab. It is that they can be grown directly, uniformly and cleanly on a 300mm wafer inside a tool a foundry would actually accept. Direct growth avoids transfer steps, where a film made elsewhere is moved onto the production wafer. That matters because transfer introduces defects, contamination risk and handling complexity that production lines resist.
In my view, the business model is as important as the materials science. Selling tools and processes to incumbents lowers capital risk. It also places evaluation where foundries prefer it, at the level of defect density, repeatability, thermal compatibility and maintenance. Fabs qualify tools, not papers.
Worth flagging is what remains to be established in customer trials. Throughput, process window, precursor management, cross-contamination controls, metrology, and insertion into existing front-end flows will decide whether North Star can move from evaluation to pilot. Those are normal unknowns at seed stage. They are also the exact criteria process engineers will use to judge the system. No seed announcement answers them.
Looking at what this enables if the integration work succeeds, the payoff is optionality. A stable TMD module that fits inside established 300mm flows would give device and integration teams another structure to work with, including stacked and heterogeneous configurations, without requiring a change in substrate or factory infrastructure. Incremental insertion has historically been the easier path into production.
My own expectation, after watching several materials move from laboratory samples to manufacturing, is for a slow then fast pattern. Early progress looks like extended yield learning and tool iteration. Later progress looks sudden once reliability data accumulates. A $12 million round funds runs, characterization and customer cycles. It does not shortcut that learning. It makes it possible to start.
