BRoCoM: A Bayesian Framework for Robust Computing on Memristor Crossbar

Di Gao, Zeyu Yang, Qingrong Huang, Grace Li Zhang, Xunzhao Yin, Bing Li, Ulf Schlichtmann, Cheng Zhuo

Research output: Contribution to journalArticlepeer-review


Memristor crossbar arrays are considered to be a promising platform for neuromorphic computing. To deploy a trained neural network (NN) model on memristor crossbars, memristors need to be programmed to the corresponding weight values. In fact, due to device-based process variation and noise, deviations of the stored weights from the trained weights are inevitable, thereby causing the degradation of the actual inference performance. This article proposes a unified Bayesian inference-based framework, BRoCoM, which connects device nonidealities and algorithmic training together for robust computing on memristor crossbars. BRoCoM is able to incorporate different levels of nonidealities into prior weight distribution, and transform robustness optimization to Bayesian NN (BNN) training, the weights of NNs are optimized to accommodate uncertainties and minimize inference degradation. Experimental results confirm the capability of the proposed BRoCoM to achieve stable inference performance while tolerating the nonideal effects of process variation and noise.

Original languageEnglish
Pages (from-to)2136-2148
Number of pages13
JournalIEEE Transactions on Computer-Aided Design of Integrated Circuits and Systems
Issue number7
StatePublished - 1 Jul 2023


  • Bayesian neural network (BNN)
  • memristor crossbar array
  • neuromorphic computing
  • system robustness


Dive into the research topics of 'BRoCoM: A Bayesian Framework for Robust Computing on Memristor Crossbar'. Together they form a unique fingerprint.

Cite this