NAND Flash Memory: Boundless Battle Against Capacity and Performance

View at publisher
Authors: Dongku Kang, Takuya Futatsuyama, Jun Hyoung Kim, Suk‐Kang Sung, Cheon An Lee
Cited by: 1
References: 9

Abstract

Since the invention of NAND flash memory in 1980s, it has quickly replaced other non-volatile semiconductor devices due to its unbeatable cost advantages. From small storage devices such as memory cards to large scale server SSDs, NAND flash is widely used to store data. Aggressive scaling technology has been focused on cell area or storage density, however, the recent NAND suffers deteriorated cell area vs. peripheral area because of a slow increase in mobile storage capacity. On the other hand, exploding data in data center requires more storage capacity with higher performance. Furthermore, AI computing architecture needs even higher performance SSDs to store and to reuse generated data from AI computation such as key values. In this paper, NAND flash design challenges are reviewed from those viewpoints with future insights. Evolution of cell size reduction technology and peripheral circuit scaling techniques are reviewed. Then, high performance IO challenges are discussed from both design and transistor device perspectives.

Fields of Study & Subjects

Advanced Data Storage Technologies Distributed systems and fault tolerance Error Correcting Code Techniques

Publication details

Journal / source
2026 IEEE International Memory Workshop (IMW)
Publisher
IEEE
Pages
1–4
Publication date
10 May 2026