Reliable In-Memory Neuromorphic Computing Using Spintronics
Reliable In-Memory Neuromorphic Computing Using Spintronics
复制标题
使用自旋电子学进行可靠的内存中神经形态计算
DOI：
10.1145/3287624.3288745
复制
发表时间：
2019
期刊：
影响因子：
--
通讯作者：
M. Tahoori
中科院分区：
文献类型：
--
作者：
Christopher Münch;R. Bishnoi;M. Tahoori
最近，自旋传输扭矩随机访问记忆（STT-MRAM）技术引起了直接实现神经网络的广泛关注，因为它提供了几种优势，例如接近零泄漏，高耐力，良好的可扩展性，小脚印和CMOS兼容性。该技术中的存储设备是磁性隧道连接点（MTJ），是使用需要新的制造材料和过程的磁层开发的。由于制造步骤和材料的复杂性，MTJ细胞受到各种失效机制的影响。结果，基于该技术的神经形态计算体系结构的功能受到严重影响。在本文中，我们开发了一个框架来分析几个MTJ缺陷存在下神经网络推断的功能能力。使用此框架，我们已经证明了所需的内存数组大小，这些大小是可以忍受给定的缺陷量以及如何通过禁用网络部分来主动降低该开销的必要数量。
Recently Spin Transfer Torque Random Access Memory (STT-MRAM) technology has drawn a lot of attention for the direct implementation of neural networks, because it offers several advantages such as near-zero leakage, high endurance, good scalability, small foot print and CMOS compatibility. The storing device in this technology, the Magnetic Tunnel Junction (MTJ), is developed using magnetic layers that requires new fabrication materials and processes. Due to complexities of fabrication steps and materials, MTJ cells are subject to various failure mechanisms. As a consequence, the functionality of the neuromorphic computing architecture based on this technology is severely affected. In this paper, we have developed a framework to analyze the functional capability of the neural network inference in the presence of the several MTJ defects. Using this framework, we have demonstrated the required memory array size that is necessary to tolerate the given amount of defects and how to actively decrease this overhead by disabling parts of the network.
影响因子：
2.4
作者：
Nan Zheng;P. Mazumder
通讯作者：
Nan Zheng;P. Mazumder
DOI：
10.1109/tvlsi.2016.2630315
发表时间：
2017-04
影响因子：
2.8
作者：
R. Bishnoi;Fabian Oboril;M. Tahoori
通讯作者：
R. Bishnoi;Fabian Oboril;M. Tahoori
VAET-STT：基于 STT-MRAM 的存储器的变化感知估计工具
DOI：
10.23919/date.2017.7927221
发表时间：
2017
期刊：
Design, Automation & Test in Europe Conference & Exhibition (DATE), 2017
影响因子：
--
作者：
S. Mohanachandran Nair;R. Bishnoi;M. S. Golanbari;F. Oboril;M. B. Tahoori
通讯作者：
M. B. Tahoori