arXivDaily arXiv每日学术速递 周一至周五更新
arXiv周末暂无论文更新,休息一下吧,周末愉快~~
arXiv 2304.06343cond-mat.mes-hall
收藏

一种基于单个磁性隧道结实现的类 Huxley-Hodgkin 自旋电子神经元

A spintronic Huxley-Hodgkin-analogue neuron implemented with a single magnetic tunnel junction

Davi R. Rodrigues, Rayan Moukhader, Yanxiang Luo, Bin Fang, Adrien Pontlevy, Abbas Hamadeh, Zhongming Zeng, Mario Carpentieri, Giovanni Finocchio

AI总结:

本文提出用单个磁性隧道结实现无需复位的脉冲神经元,可模拟 Huxley-Hodgkin 和 Leaky-Integrate and Fire 模型,在室温下达到 MHz-GHz 发放频率,兼具紧凑、可扩展、低功耗和 CMOS 兼容优势。

摘要:

脉冲神经网络旨在模拟大脑的特性,以实现类似的并行性和高处理能力。这类神经网络的一个缺陷是仿真计算成本高,而当前的模拟实现方案能效低且不可扩展。我们提出一种基于单个磁性隧道结的器件,用于为脉冲神经网络执行神经元发放,无需任何复位过程。我们利用磁性和热效应这两种物理机制,获得与神经元 Huxley-Hodgkin 模型类似的生物真实脉冲行为。该器件还能够模拟更简单的 Leaky-Integrate and Fire 模型。基于实验参数的数值模拟表明,在室温恒定输入下,发放频率可达 MHz 至 GHz 范围。磁性隧道结的紧凑性、可扩展性、低成本、CMOS 兼容性和功率效率,使其适合在脉冲神经网络的硬件实现中广泛使用。

英文摘要:

Spiking neural networks aim to emulate the brain's properties to achieve similar parallelism and high-processing power. A caveat of these neural networks is the high computational cost to emulate, while current proposals for analogue implementations are energy inefficient and not scalable. We propose a device based on a single magnetic tunnel junction to perform neuron firing for spiking neural networks without the need of any resetting procedure. We leverage two physics, magnetism and thermal effects, to obtain a bio-realistic spiking behavior analogous to the Huxley-Hodgkin model of the neuron. The device is also able to emulate the simpler Leaky-Integrate and Fire model. Numerical simulations using experimental-based parameters demonstrate firing frequency in the MHz to GHz range under constant input at room temperature. The compactness, scalability, low cost, CMOS-compatibility, and power efficiency of magnetic tunnel junctions advocate for their broad use in hardware implementations of spiking neural networks.

机构、总结、中文摘要均由AI分析而来,可能存在错误,请以论文正文为准。

↑