Just as Taiwan Semiconductor Manufacturing (TSM) entered mass production of its 2nm process this year with wafer prices reportedly reaching as high as $30,000, a Japanese semiconductor startup backed by significant government support is quietly positioning itself to challenge the industry giant through pricing advantages. Rapidus, the company in question, plans to begin mass production of 2nm chips in the second half of 2027, with pricing set at 3 million to 3.5 million yen per wafer (approximately $18,500 to $21,500), representing a roughly 30% discount to TSMC’s quoted prices.
However, in the foundry business — a winner-take-all arena — is price truly the deciding factor?
TSMC’s Moat Runs Far Deeper Than It Appears
Consider the numbers. In the first quarter of 2026, the world’s top 10 foundries generated combined revenue of $47.953 billion, up 31.7% year-over-year. Of that total, TSMC alone captured $35.9 billion, increasing its market share from 70.4% in the previous quarter to 72.3%. Even more striking, TSMC commands approximately 95% of the “leading-edge node” market (3nm and below). Despite years of effort from established players like Samsung and Intel, the gap has only widened.
TSMC’s financial performance is equally impressive: first-quarter revenue grew 40.6% year-over-year, with gross margin reaching 66.2% and net profit surging 58.3%. The company remains highly confident in achieving over 30% revenue growth for the full year.
Such profitability and growth momentum cannot be shaken by “low prices” alone. When semiconductor design companies choose their manufacturing partners, price is never the primary consideration — yield rates, reliability, production capacity, and technological ecosystem are the true decision-making criteria. NVIDIA CEO Jensen Huang has repeatedly praised TSMC, calling it a “world-class foundry.” This trust has been earned through decades of consistent, reliable delivery.
Rapidus’s Real-World Challenges
Founded in 2022 with backing from eight Japanese conglomerates including Toyota, Sony, SoftBank, and Fujitsu, Rapidus has received cumulative R&D support of approximately 2.354 trillion yen (about $15.4 billion) from the Japanese government. The company currently operates only a single pilot production line, plans to produce test chips for customers by the end of 2026, and has an initial monthly production capacity of just 6,000 wafers, with a target of expanding to approximately 25,000 wafers per month by 2028.
In contrast, TSMC’s 2nm capacity expansion is proceeding rapidly, with monthly capacity expected to reach 70,000 to 75,000 wafers by the end of 2026. On the customer front, six major clients — including AMD, Apple, MediaTek, and Qualcomm — have already locked in 95% of N2 process orders. While Rapidus has reportedly engaged with more than 60 potential customers and has confirmed Fujitsu as its first commercial client, competing with TSMC in terms of production scale and customer trust remains an immense challenge.
Final Thoughts
Rapidus CEO Atsushi Koike has made it clear that “we cannot lose to TSMC on pricing.” While this ambition is certainly aggressive, competition in the semiconductor industry has never been solely about price. TSMC holds a first-mover advantage of more than a year in 2nm mass production, boasts massive production capacity, and maintains deep ties with the world’s top customers. Even if Rapidus begins mass production as scheduled in 2027, it will face an industry giant that has built comprehensive moats across cost, technology, scale, and ecosystem.
For investors, Rapidus’s low-price strategy is worth watching, but it is far from sufficient reason to adopt a bearish view on TSMC. In the marathon of semiconductor manufacturing, a late-starting, smaller-scale challenger needs far more than pricing advantages to make a meaningful impact.
