FeRAM using Anti-ferroelectric Capacitors for High-speed and High-density Embedded Memory

S. -C. Chang,N. Haratipour,S. Shivaraman,C. Neumann, S. Atanasov, J. Peck,N. Kabir, I. -C. Tung, H. Liu, B. Krist,A. Oni,S. Sung, B. Doyle, G. Allen, C. Engel, A. Roy, T. Hoff,H. Li,F. Hamzaoglu, R. Bristol,M. Radosavljevic, B. Turkot,M. Metz,I. Young, J. Kavalieros,U. Avci

2021 IEEE International Electron Devices Meeting (IEDM)(2021)

引用 12|浏览7
暂无评分
摘要
This paper demonstrates industry-best hafnium-based FeRAM performance and reliability by showing (i) read/write speed scaled down to ~2ns, (ii) read/write endurance beyond 10 12 cycles, and (iii) tail-bit variations of scaled capacitors working at $4\sigma$ across a 300mm wafer at elevated temperature, by switching anti-ferroelectric (AFE) capacitors at −1.6V and 1.2V. Furthermore, a physics-based multi-domain compact circuit model is developed for AFE capacitors to describe FeRAM operations. Array-level circuit simulations show that FeRAM is less vulnerable to disturb through parasitic capacitor coupling due to the small amount of polarization charge change $(\Delta P)$ relative to its high remanent polarization $(P_{r})$ . Finally, high yield in a capacitor-array with no significant degradation in retention well over 10s and a healthy memory window (MW) under 1ms disturb 20% of $V_{write}$ at elevated temperature is shown, paving way for AFE-based FeRAM toward the next generation high-speed and high-density embedded memory.
更多
查看译文
关键词
next generation high-speed embedded memory,read-write endurance,read-write speed,industry-best hafnium-based FeRAM performance,switching anti-ferroelectric capacitors,AFE-based FeRAM,healthy memory window,capacitor-array,high remanent polarization,polarization charge change,parasitic capacitor coupling,array-level circuit simulations,AFE capacitors,physics-based multidomain compact circuit model,scaled capacitors,tail-bit variations,reliability,high-density embedded memory,anti-ferroelectric capacitors,size 300.0 mm,voltage -1.6 V,time 10.0 s,time 1.0 ms,voltage 1.2 V
AI 理解论文
溯源树
样例
生成溯源树,研究论文发展脉络
Chat Paper
正在生成论文摘要