2nm era begins — Why TSMC still sits at the summit, as Samsung and Intel push forward?

The transition from 3nm to 2nm marks the most consequential inflection in semiconductor manufacturing since the industry adopted FinFETs more than a decade ago. This shift is not simply about smaller transistors, but about a structural change in how advanced chips are designed, manufactured, and monetized.

Gate-All-Around architectures, exploding power requirements from AI workloads, and sharply rising capital intensity have fundamentally altered the economics of leading-edge logic.

What makes the 2nm transition different from prior node changes is timing. At advanced nodes, competitive outcomes are now being decided years before meaningful revenue appears. Customers are locking in foundry partners, capacity is being built and tooled, memory ecosystems are aligning around next-generation HBM, and wafer fab equipment spending is rising well ahead of product launches.

By the time 2nm chips reach volume markets in 2026, leadership will not be debated — it will already be embedded in supply chains and reflected in equity valuations.

This article explains why the 2nm race is effectively being decided today, why that outcome still centers on TSMC (TSM) despite visible progress from Samsung (SSNLF) and Intel (INTC), and how investors should interpret the implications for semiconductor stocks before the revenue cycle becomes obvious.

Why this matters now?
The most important investment insight of the 2nm transition is not who claims technical leadership, but when that leadership becomes irreversible. At advanced nodes, competitive outcomes are now decided years before revenue materializes, as customers lock in suppliers, capacity is built, and memory ecosystems align.

By the time 2nm products reach volume markets in 2026, the winners will already be priced into equity valuations. This article explains why the 2nm race is effectively being decided today — and why that matters for investors allocating capital across TSMC, Samsung, and Intel.

Global foundry consolidation and long arc of leadership
According to the table, the global foundry industry has undergone a decisive consolidation over the past two decades, culminating in a level of dominance that directly shapes the 2nm race. The leading foundry’s share rises from 45% in 2005 to an estimated 70% by 2025E, reflecting not cyclical advantage, but sustained execution across multiple node transitions.

Importantly, this consolidation accelerates as nodes become more capital-intensive and technically complex, indicating that leadership at the leading edge has become self-reinforcing rather than contestable.

Samsung’s trajectory illustrates the challenge. Despite heavy investment and periodic technical advances, its foundry share trends downward into the mid-single digits by 2025E. The shrinking “Others” category underscores the collapse of viable alternatives at advanced nodes.

In this context, the 2nm transition is not a reset of competitive balance, but the continuation of a long-running structural shift toward a single dominant manufacturing platform.

-- Dr. Robert Castellano, Semiconductor Deep Dive, USA.