← 返回论文列表 📄 下载原文 PDF  ISSCC 2018 · 30.4
ISSCC 2018Session 30 · EMERGING MEMORIESOther60nm crystalline oxide semiconductor

A 20ns-Write 45ns-Read and 1014-Cycle Endurance Memory Module Composed of 60nm Crystalline Oxide Semiconductor Transistors

⚡ 本页包含 AI 生成的分析内容,仅供参考

📋 论文概要

本文提出了一种基于60nm结晶氧化物半导体晶体管的存储模块,实现了20ns写入、45ns读取和10^14次循环的高耐久性,旨在为人工智能应用提供低功耗高速的LSI。

💡 主要创新点

工艺节点
60nm crystalline oxide semiconductor
重要性
发表年份
ISSCC 2018

🏷 关键词

氧化物半导体非易失性存储器高耐久性低功耗人工智能

📄 原文摘要

Takahiko Ishizu1, Tomoaki Atsumi1, Yoshinori Ando1, Daisuke Matsubayashi1, Kiyoshi Kato1, Takashi Okuda1, Masahiro Fujita2, Shunpei Yamazaki1 Semiconductor Energy Laboratory, Atsugi, Japan 2 University of Tokyo, Tokyo, Japan 1 Development of LSI targeting artificial intelligence (AI) has accelerated, some chips have been used and are commercially available in a number of applications. LSI capable of performing arithmetic operation for deep learning, etc., at low power and high speed is crucial for achieving more sophisticated AI. Power consumption is increasing significantly owing particularly to the practical use of AI, and power reduction techniques are urgently necessary. One way of reducing LSI power consumption is by power gating, where the LSI is powered off during standby operations. Embedded memories capable of retaining data for a long time are required for power gating [1-4]. This embedded

👥 作者与机构

Shuhei Maeda1, Satoru Ohshita1, Kazuma Furutani1, Yuto Yakubo1,

分类:Other · 年份:ISSCC 2018