3D NAND has continued to evolve through vertical and lateral scaling, with more than 300 layers now being demonstrated. However, as scaling becomes increasingly complex, simply adding more layers is no longer sufficient. Cost, process complexity, device performance, and reliability are becoming critical challenges.

The key challenges lie in three areas.

First, HARC etching and process integration.

As the stack becomes higher, high-aspect-ratio etching becomes more challenging and costly. HARC process integration and process merging are therefore essential to reduce process complexity and maintain cost competitiveness.

Second, cell stack formation and cell current.

Higher aspect ratios make uniform and conformal film formation increasingly difficult. At the same time, longer channels lead to higher resistance and lower cell current. Advanced ALD processes, grain-size engineering, trap passivation, and emerging approaches such as MILC (Metal Induced Lateral Crystallization) are being explored to maintain cell performance. For MILC, controlling grain size variation and residual metal contamination remains a key challenge

Third, film stress and manufacturability.

As the stack becomes thicker and integration becomes more complex, film stress can cause wafer warpage, pattern collapse, cracks, and handling issues. Stress engineering, including backside process optimization, layout design, and improved wafer handling, is becoming an essential part of 3D NAND development.

Beyond conventional scaling, new structures such as isolated trap layer, air-gap cells, Multi-Site Cell (MSC), and trench-type cells are being explored to increase bit density without relying solely on additional layers. Meanwhile, emerging technologies such as FeNAND and HBF (High Bandwidth Flash) are expanding the possibilities of NAND beyond conventional charge-based memory and traditional storage applications.

Ultimately, the next stage of 3D NAND development will require balanced optimization of cost, performance, reliability, and manufacturability, supported by advances in integration, structure, and application.

The future of NAND will depend not only on scaling, but also on advances in integration, structure, and application.