EUVIMEDEUVIMED entdecken
Crossref

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

IEEE Transactions on Computer-Aided Design of Integrated Circuits and Systems · 2023-07 · 42 (7) · 2136-2148

DOI: 10.1109/tcad.2022.3215071 ↗

Kurz gefasst

Fachartikel in der Zeitschrift IEEE Transactions on Computer-Aided Design of Integrated Circuits and Systems, 2023-07. Verfasst von Di Gao und 7 weiteren Autor:innen. Die Arbeit wurde bisher 13-mal zitiert (Crossref) und nennt 52 Literaturangaben. Gefördert von: Key-Area Research and Development Program of Guangdong Province, NSFC, Zhejiang Provincial NSF.

AbstractBibliografische AngabenZitierenVerwandte Studien

Abstract (Originalsprache)

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.

Förderung

  • Key-Area Research and Development Program of Guangdong Province
  • NSFC
  • Zhejiang Provincial NSF
  • SGC Cooperation Project
  • Deutsche Forschungsgemeinschaft

Zitieren

APA

Gao, D., Yang, Z., Huang, Q., Zhang, G. L., Yin, X., Li, B., Schlichtmann, U., & Zhuo, C. (2023). BRoCoM: A Bayesian Framework for Robust Computing on Memristor Crossbar. IEEE Transactions on Computer-Aided Design of Integrated Circuits and Systems, 42(7), 2136-2148. https://doi.org/10.1109/tcad.2022.3215071

Vancouver

Gao D, Yang Z, Huang Q, Zhang GL, Yin X, Li B, et al. BRoCoM: A Bayesian Framework for Robust Computing on Memristor Crossbar. IEEE Transactions on Computer-Aided Design of Integrated Circuits and Systems. 2023;42(7):2136-2148. doi:10.1109/tcad.2022.3215071

BibTeX
@article{Gao2023,
  author = {Gao, Di and Yang, Zeyu and Huang, Qingrong and Zhang, Grace Li and Yin, Xunzhao and Li, Bing and Schlichtmann, Ulf and Zhuo, Cheng},
  title = {BRoCoM: A Bayesian Framework for Robust Computing on Memristor Crossbar},
  journal = {IEEE Transactions on Computer-Aided Design of Integrated Circuits and Systems},
  year = {2023},
  volume = {42},
  number = {7},
  pages = {2136--2148},
  doi = {10.1109/tcad.2022.3215071}
}
Kostenlos starten →