Collaborative Research: BRAM: Balanced RAnk Modulation for data storage in next generation flash memories

集成，混合和复杂系统蒋安晓江，德克萨斯州工程实验站加州理工学院的Jehoshua Bruck，GA技术研究公司-GA理工学院的Paul E.Hasler合作研究：BRAM：用于下一代闪存数据存储的平衡秩调制这项研究的目标是开发一种新的数据存储技术，用于下一代闪存，大大提高了寿命、可靠性和效率。这种方法是使用闪存单元级别的等级来表示数据，而不是使用单元级别的绝对值。研究了新的技术，包括消除单元过度编程、减少甚至消除块擦除、数据修改、负载平衡和纠错。该研究将新的信息论方法与硬件设计紧密结合。智力价值：本研究的智力价值包括为新的数据存储技术奠定理论基础，并为闪存设计新的硬件架构。秩调制方案消除了电荷过度注入的风险，并减少了块擦除，这是闪存进步的两个主要障碍。理论分析和代码设计有可能推动信息论的发展。对硬件的研究，包括新的电路元件和测试技术，有可能推动存储器设计。广泛的影响：电子存储器是一种广泛使用的存储介质，与磁性和光学介质一样。闪存占非易失性电子存储器市场的90%。实现下一代闪存的技术有可能产生重大的社会和经济影响，有可能使工业生产、科学研究和大量闪存用户受益。这项研究通过开发新的课程，让学生，包括来自代表性不足群体的学生，参与高级研究，并传播新知识，将研究与教育结合起来。

Integrative, Hybrid and Complex SystemsAnxiao Jiang, Texas Engineering Experiment StationJehoshua Bruck, California Institute of TechnologyPaul E. Hasler, GA Tech Research Corporation - GA Institute of TechnologyChristopher M. Twigg, SUNY at BinghamtonCollaborative Research: BRAM: Balanced RAnk Modulation for Data Storage in Next Generation Flash MemoriesThe objective of this research is to develop a new data storage technology for next-generation flash memories with substantially improved longevity, reliability and efficiency. The approach is to use the ranks of flash-cell levels to represent data, instead of using the absolute values of the cell levels. Novel techniques are studied, including elimination of cell over-programming, reduction or even elimination of block erasures, data modification, load balancing, and error correction. The research closely combines new information-theoretical methods and hardware design.Intellectual Merit: The intellectual merit of this research includes laying the theoretical foundation for a new data-storage technology and designing a new hardware architecture for flash memories. The rank-modulation scheme eliminates the risk of charge over-injection and reduces block erasures, which are two major barriers to advances in flash memories. The theoretical analysis and the code designs have the potential to advance information theory. The study of hardware, including novel circuit components and testing techniques, has the potential to advance memory design.Broader Impacts: Electronic memories are a widely used storage media, along with magnetic and optical media. Flash memories account for 90 percent of the non-volatile electronic memory market. Techniques that enable next-generation flash memory have the potential to make significant societal and economic impacts, with the possibility of benefiting industrial production, scientific research, and the large population of flash-memory users. This research integrates research and education by developing new courses, engaging students in advanced research, including students from under-represented groups, and disseminating new knowledge.