{"title":"Gain cell block architecture for gigabit-scale chain ferroelectric RAM","authors":"D. Takashima, Y. Oowaki, I. Kunishima","doi":"10.1109/VLSIC.1999.797251","DOIUrl":null,"url":null,"abstract":"Summary form only given. A ferroelectric RAM (FRAM), especially a chain FRAM, has great potential for future high-density nonvolatile memory. However, two severe problems inherent to ferroelectric material make it difficult to realize gigabit scale FRAMs; cell polarization decreases drastically in scaled FRAMs, 1) because the cell polarization does not increase by thinning the ferroelectric film, and 2) because the three-dimensional ferroelectric capacitor is difficult to make. Therefore, a sufficient cell signal will not be obtained in 256 Mb FRAMs and beyond. The gain cell approach shown can be a solution for this problem because a large cell signal is obtained even with small cell polarization due to small load capacitance. However, a memory cell using a ferroelectric FET has drawbacks such as fabrication difficulty, poor data retention and read/write disturb. A memory cell composed of a gain transistor, a write transistor, a ferroelectric capacitor and a load capacitor, realizes stable read/write operation. However the memory cell size is very large. The concept of a new gain cell block is proposed. The gain cell block contains two chain cell blocks and a gain unit composed of a gain transistor and a write transistor. The gain unit is shared by the two chain cell blocks. This configuration realizes both a large readout cell signal and a small average cell size.","PeriodicalId":433264,"journal":{"name":"1999 Symposium on VLSI Circuits. Digest of Papers (IEEE Cat. No.99CH36326)","volume":"11 1","pages":"0"},"PeriodicalIF":0.0000,"publicationDate":"1999-06-17","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":"6","resultStr":null,"platform":"Semanticscholar","paperid":null,"PeriodicalName":"1999 Symposium on VLSI Circuits. Digest of Papers (IEEE Cat. No.99CH36326)","FirstCategoryId":"1085","ListUrlMain":"https://doi.org/10.1109/VLSIC.1999.797251","RegionNum":0,"RegionCategory":null,"ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":null,"EPubDate":"","PubModel":"","JCR":"","JCRName":"","Score":null,"Total":0}
引用次数: 6
Abstract
Summary form only given. A ferroelectric RAM (FRAM), especially a chain FRAM, has great potential for future high-density nonvolatile memory. However, two severe problems inherent to ferroelectric material make it difficult to realize gigabit scale FRAMs; cell polarization decreases drastically in scaled FRAMs, 1) because the cell polarization does not increase by thinning the ferroelectric film, and 2) because the three-dimensional ferroelectric capacitor is difficult to make. Therefore, a sufficient cell signal will not be obtained in 256 Mb FRAMs and beyond. The gain cell approach shown can be a solution for this problem because a large cell signal is obtained even with small cell polarization due to small load capacitance. However, a memory cell using a ferroelectric FET has drawbacks such as fabrication difficulty, poor data retention and read/write disturb. A memory cell composed of a gain transistor, a write transistor, a ferroelectric capacitor and a load capacitor, realizes stable read/write operation. However the memory cell size is very large. The concept of a new gain cell block is proposed. The gain cell block contains two chain cell blocks and a gain unit composed of a gain transistor and a write transistor. The gain unit is shared by the two chain cell blocks. This configuration realizes both a large readout cell signal and a small average cell size.