In-MRAM Computing Elements with Single-Step Convolution and Fully Connected for BNN/TNN

2021 IEEE International Conference on Integrated Circuits, Technologies and Applications (ICTA)(2021)

引用 1|浏览2
暂无评分
摘要
In-memory computing shows great energy-efficient attraction in non-von Neumann architecture. Computing in spin-transfer-torque magnetic RAM (STT-MRAM) is expected to achieve both ultra-low-power and interconnection delay reduction. In this work, a 4T2M MRAM bit-cell and associated peripheral circuits are proposed, which can perform XNOR operation and ternary operation extension in-memory. At the s...
更多
查看译文
关键词
Integrated circuit technology,Energy consumption,Convolution,Conferences,Neural networks,Random access memory,Integrated circuit interconnections
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