Samsung's 1.4nm Delayed to 2029: When Process Limits Start to Bite

The Event

Samsung's originally planned 2027 mass production of 1.4 nanometer process has been revealed to be delayed until 2029. This is not the first time. Over the past year, Samsung has publicly claimed to be "still developing" 1.4 nanometers, but internally has already faced a double squeeze from yield rates and costs.

Meanwhile, TSMC's 3 nanometer is already mass-supplying Apple's A18 chips, and 2 nanometer is reportedly in trial production. The gap between Samsung and TSMC is no longer "lagging by one process generation," but an increasingly insurmountable chasm of time and cost.

Surface vs. Structure

On the surface, this appears to be a failure in Samsung's process R&D—yield problems, EUV (extreme ultraviolet lithography) equipment bottlenecks, talent drain.

But the deeper pattern is: chip process advancement no longer follows the simple extrapolation of Moore's Law.

From 65 nanometers to 28 nanometers, the industry spent 6-7 years and invested tens of billions of dollars to achieve it. But from 14 nanometers to 7 nanometers, the timeline lengthened and costs doubled. Now, from 7 nanometers to 1.4 nanometers, the investment scale has already exceeded what a single manufacturer can bear.

Why Are Returns Diminishing?

Physics is hitting a ceiling.

When transistor sizes approach atomic scales (1 nanometer is already the width of a dozen silicon atoms), the following problems become impossible to avoid:

1. Quantum Tunneling: Electrons pass directly through the gate; logic becomes unreliable 2. Leakage Current: Power consumption spikes sharply; cooling costs become the primary expense 3. Manufacturing Precision: The wavelength of EUV light itself is the limit; smaller patterns cannot be created 4. Yield Collapse: Tens of billions of transistors exist on each chip; any single manufacturing error is a defective product

What's the result? Manufacturing costs don't rise linearly; they spike exponentially.

Industry estimates suggest that the per-unit cost of 1.4 nanometers could be 4-5 times that of 3 nanometers, while performance gains are only 15-20%. This means that even if Samsung successfully mass-produces it, it will be difficult to convince customers to adopt this expensive process.

Apple's Dilemma

Apple has long relied on TSMC as the exclusive supplier of its most advanced processes. Between 2025-2026, TSMC's 3 nanometer capacity is already strained; if 2 nanometer launches on schedule, Apple still needs multiple sources to mitigate supply risk.

This is also why Apple has been extending olive branches to Samsung, Renesas, and others in recent years, requiring them to at least reach 3 nanometer or near-3 nanometer capabilities. But Apple won't demand 1.4 nanometers—because for iPhone processors, the performance bottleneck is no longer the chip itself, but software and battery.

The Industry's Quiet Revolution

The true significance of this news is: the focus of semiconductor progress is shifting.

No longer "how small," but "how fast," "how power-efficient," and "how easy to manufacture."

• TSMC is investing heavily in Chiplet and 3D stacking technology, rather than simply chasing smaller process numbers
• Qualcomm and Apple are designing specialized processors themselves, no longer pursuing generalized process leadership
• Startups are innovating on conventional processes with novel architectures and gaining performance advantages instead (such as DeepSeek's inference architecture)

Samsung's 1.4 nanometer delay symbolizes a turning point: when the cost of chasing the frontier exceeds the benefits, the entire competitive ruleset gets redefined.

Lessons for Investors

If you're evaluating chip manufacturers' futures, stop asking "when will the next generation process launch." Ask instead:

1. The Economics of the Process: Is the cost-to-performance ratio of this node reasonable? 2. Actual Customer Needs: Does the market truly need this process, or is the manufacturer just indulging itself? 3. Can Architecture Innovation Compensate for Slowing Process Advancement?

Samsung's 1.4 nanometer delay looks like a loss on the surface; but if Samsung invests the saved $5 billion into Chiplets, ASIC design, or novel cooling solutions, it could actually bounce back in the next round of competition. Conversely, clinging to the commitment of "chasing the smallest process numbers" will only exhaust itself on an ever-steeper treadmill.