⚡ 本页包含 AI 生成的分析内容,仅供参考
该论文介绍了一种采用45nm自对准接触工艺的1Gb NOR闪存,通过消除相邻字线浮栅之间的耦合,解决了工艺节点缩小带来的单元效应恶化问题,实现了5MB/s的编程速度。
Robert L. Melcher, Shahnam Khabiri, Nicholas T. Hendrickson, Andrew D. Proescholdt, David A. Ward, Mark A. Taylor Intel, Folsom, CA When transitioning to the 45nm process node, undesirable cell effects become many times worse and it is difficult to meet product requirements while maintaining sense margin in NOR flash. Fig. 23.3.1 shows the self-aligned-contact (SAC) process architecture, which allows cell-size reduction and improves reliability of the flash cells. The SAC architecture eliminates the coupling between floating gates of adjacent wordlines [1]. This prevents cell Vt shift when programming adjacent wordlines and improves margin for MLC sensing. Program performance of 5MB/s requires the development of a number of circuit techniques. These include larger program and verify bandwidth, increased slew rates of high-voltage nodes, maximized throughput of the program microcode, and reduced control
Johnny Javanifard, Tris Tanadi, Hari Giduturi, Kim Loe,