首页> 外文会议>応用物理学会春季学術講演会 >Impact of solid-state memristor variability on perceptron supervised learning via STDP
【24h】

Impact of solid-state memristor variability on perceptron supervised learning via STDP

机译:通过STDP对固态记忆变异性的影响通过STDP监督学习

获取原文

摘要

Memristors [1] are non-volatile memory elements with an electrically programmable resistance. Bipolar binary memristors have extensive promise in replacing FLASH as the next generation storage/memory medium. Conversely we focus here on bipolar analogue memristors (with a continuously variable resistance), which represent promising candidates for implementation as artificial synapses in hardware-based artificial neural networks towards realisation of power efficient and compact neuromorphic processors [2]. These have impactful applications in the design of autonomous cognitive agents [3] and deep-learning accelerators [4]. However progress in this field has been hindered by practical device issues.
机译:存储器[1]是具有电性可编程电阻的非易失性存储器元件。双极二进制存储器在更换闪存作为下一代存储/内存介质时具有广泛的承诺。相反,我们专注于双极性模数膜(具有不断可变的阻力),这代表了有希望的候选人,用于实现基于硬件的人工神经网络中的人工突触,朝向实现功率高效和紧凑的神经形态处理器[2]。这些对自主认知剂[3]和深度学习加速器的设计有影响力的应用[4]。然而,该领域的进展已经受到实际设备问题的阻碍。

著录项

相似文献

  • 外文文献
  • 中文文献
  • 专利
获取原文

客服邮箱:kefu@zhangqiaokeyan.com

京公网安备:11010802029741号 ICP备案号:京ICP备15016152号-6 六维联合信息科技 (北京) 有限公司©版权所有
  • 客服微信

  • 服务号