Performance impacts of analog ReRAM non-ideality on neuromorphic computing

Yu Hsuan Lin, Chao Hung Wang, Ming Hsiu Lee, Dai Ying Lee, Yu Yu Lin, Feng Min Lee, Hsiang Lan Lung, Keh Chung Wang, Tseung Yuen Tseng, Chih Yuan Lu

研究成果: Article同行評審

49 引文 斯高帕斯(Scopus)


Resistive random access memory (ReRAM) is often considered as a strong candidate for storing the weights in non-von Neumannneuromorphic computing systems. This paper studies how nonideal memory characteristics, which include programing error, read fluctuation, and retention, affect the inference accuracy of the analog ReRAM neural networks by incorporating memory characteristics extracted from 1-Mb ReRAM into a simulated inference-only neural network. This paper also shows that the different layer in the network can tolerate different amount of such imperfects. We learned four key points: 1) the conductance range of memory with less relative fluctuation is preferred for designing the weight-conductance mapping; 2) the control of programing error is essential for high inference accuracy; 3) retention-induced conductance drift can be fatal to the neuromorphic system. A compensation scheme is proposed in this paper which can effectively recover the inference accuracy; and 4) for multilayer networks, avoiding weight errors in the front layers can help to maintain the inference accuracy by reducing calculation error which may otherwise accumulate and pass down the networks. The concepts and approaches of this paper can also be applied to evaluate other types of nonvolatile memories for artificial neural networks.

頁(從 - 到)1289-1295
期刊IEEE Transactions on Electron Devices
出版狀態Published - 2019 3月

All Science Journal Classification (ASJC) codes

  • 電子、光磁材料
  • 電氣與電子工程


深入研究「Performance impacts of analog ReRAM non-ideality on neuromorphic computing」主題。共同形成了獨特的指紋。